upload 101 to 200

file101.txt

@@ -0,0 +1,259 @@
+
+	
+	
+		
+IntroductionAir traffic demand is projected to increase significantly in the upcoming years [1].In order to meet the forecasted levels, the human workload associated with conflict detection and resolution must be reduced to assure system safety and performance.Automated separation assurance systems are proposed as a way to efficiently separate aircraft in highly dense traffic situations up to two to three times current levels.There are numerous algorithms proposed to provide separation assurance in the future air traffic system [2].With any automated resolution tool, the type of resolution selected is based on some cost function.Two costs associated with conflict resolution maneuvers are delay and fuel burn.These two costs are closely correlated, but not necessarily the same.There have been several previous studies of conflict resolution algorithms in which the preferred resolution is selected based on delay [3,4].These studies demonstrate that the algorithms are robust to high traffic demand and can find resolutions that have low average delay, but they have not been examined for fuel efficiency.The purpose of the current study is to compare the system performance of a conflict resolution algorithm in realistic traffic scenarios when selecting resolutions based on minimum delay to system performance when selecting resolutions based on minimum fuel burn.The total costs in terms of fuel burn and delay are compared between the two resolution selection criteria.Also, the total number of conflicts is compared to determine if there is any adverse impact on safety.Finally, an examination of the different types of resolutions selected in the two cases is performed to understand how the selection criteria may affect specific resolution scenarios.The paper is organized as follows: a brief overview of the simulation setup is provided in the next section and is followed by the equations that govern fuel burn.The results are then presented along with a discussion of their implications, concluding with a summary of the findings and a recommendation for future work.
+Simulation SetupIn this study the Airspace Concept Evaluation System (ACES) is used to simulate the National Airspace System (NAS) in a fast-time simulation.The conflict resolution algorithm evaluated is the Advanced Airspace Concept (AAC) Autoresolver [4,5].For this simulation ACES computes the delays and the fuel burn values used by the Autoresolver to select preferred resolutions.
+Test BedACES is a fast-time, agent-based simulation of the NAS that uses four-degree-of-freedom equations based on the Base of Aircraft Data (BADA) to generate aircraft trajectories [4].ACES was developed specifically to provide a general purpose environment for evaluating future air traffic management and control concepts, including automated resolution algorithms.Essential to the simulation of resolution algorithms is the ability to generate 4D trajectories.In ACES these trajectories begin at the departure fix and end at the arrival fix.By using aircraft-type-specific performance data together with guidance and navigation models, the ACES trajectory engine can generate representative trajectories for many aircraft.In the ACES simulation environment these aircraft trajectories are entirely deterministic; aircraft conflicts can be predicted with perfect accuracy and resolution trajectories are guaranteed to be followed precisely by the simulated aircraft.In addition to deterministic aircraft trajectories, simplifications were made in the modeling and execution of the experiment.Negotiation of resolution trajectories between aircraft operators and/or the air navigation service provider were not modeled.Neither were data link transmission delays or pilot-action delays.Once a resolution trajectory was determined by the automation it was executed immediately and precisely [3].
+Test Article: AAC AutoresolverThe AAC Autoresolver is a strategic conflict resolution algorithm designed to deconflict aircraft that are predicted to lose separation more than two minutes in the future.For this study, every minute of simulation time the future trajectories of all the aircraft are computed and processed to determine if there are any predicted losses of separation where two aircraft come within 5 nautical miles horizontally and 1,000 feet vertically of one another.The Autoresolver receives a list of aircraft conflict pairs ordered by predicted time to first loss of separation.For each conflict in the conflict list, the Autoresolver follows an iterative approach for resolution.Taking into account characteristics such as aircraft type, speed and airspace boundaries, the resolver calculates a future route composed of waypoints, speeds and altitudes which may possibly resolve the conflict.Figure 1 shows the types of future routes attempted by the Autoresolver grouped by whether they are horizontal, vertical, or speed resolutions.This future route is then sent to a trajectory engine that computes a trial resolution trajectory based on this route.In order for the resolution to be viable, it must resolve the primary conflict, be free of predicted losses of separation with the primary aircraft in the conflict, as well as any other aircraft in the simulation for a specified period of time.If these conditions are met, the Autoresolver has successfully generated a candidate resolution trajectory and stores it.If the resolution is not free of primary or secondary conflicts, the Autoresolver computes a new resolution route and checks if it is successful.For each resolution type this iteration is continued until a successful resolution is found or all possibilities of that type have been tried.For each successful resolution, both the associated delay and the fuel burn are calculated.A common spatial point on the original trajectory and the resolution trajectory is found.To calculate the delay, the time on the original trajectory at the common point is subtracted from the time on the resolution trajectory at the common point.Similarly for the fuel burn, the weight of the aircraft at the common point for the resolution trajectory is subtracted from the aircraft weight for the original trajectory.A discussion of how the aircraft weight is calculated and converted to fuel burn is given in a subsequent section.The resolver will generate up to seven successful resolutions per aircraft in conflict for a total of fourteen available between the two aircraft.In this study, the algorithm selects a resolution from among the set of successful resolutions using either the minimum delay or the minimum fuel burn criterion, depending on how the algorithm is configured.The selected resolution is then sent to ACES for implementation.Further discussion regarding the design of the algorithm and the types of resolutions that are generated is presented in [4,5].
+ProcedureTo illustrate the differences between selecting conflict-resolution maneuvers based on delay and selecting resolution maneuvers based on fuel burn, a test plan was developed to isolate this variable.Two cases were simulated for each specific scenario: one with resolution selection based on minimum delay and one with resolution selection based on minimum fuel burn.
+Demand SetFlight operations over a 36-hour period were simulated based on Aircraft Situation Display to Industry (ASDI) data recorded March 8, 2007.ASDI data comes from the FAA's Enhanced Traffic Management System (ETMS) and contains information about flights controlled by air traffic control.The data set included 62,970 flights, their associated routes, and their departure times.This dataset had mixed aircraft types representing the current fleet mix.The specific day was selected because it represented a "low weather," high volume day in the NAS.
+Simulated AirspaceFor this study, the Autoresolver provided conflict resolution services for a single Air Route Traffic Control Center at a time.Three centers were used to analyze the algorithmic performance with different types of air traffic flows.These centers were Indianapolis (ZID), Minneapolis (ZMP) and Atlanta (ZTL).The simulation included all types of air traffic for each center: departures, overflights and arrivals for air carrier, business and general aviation.Each of the demand sets provided thirty six hours of simulated air traffic transitioning through the selected airspace.Two simulations were run per center, one for fuel burn and one for delay for a total of six simulations.Although the data set used in the simulations consisted of 62,970 flights, the number of flights that passed through each center differed because of differences in the size and layout of the airspace along with the traffic volume and composition.Table 1 shows the experiment test matrix and number of flights that passed through each center.
+Center
+Fuel Burn EquationsBy default, the AAC Autoresolver selects the preferred conflict resolution based on minimum delay.For this study, the algorithm was modified to allow for selection of the preferred resolution based on minimum fuel burn.Since the computation of fuel burn is critical to the results presented here, the equations used to calculate this fuel burn will now be discussed.The fuel burn required for a resolution for this simulation is computed by ACES using aircraft-specific coefficients selected from the Base of Aircraft Data [7].The BADA is comprised of the performance and operating procedure coefficients of 295 aircraft types.These coefficients encompass those that are used to calculate thrust, drag, and fuel flow along with those used to specify nominal cruise, climb and descent speeds.The BADA fuel model uses the thrustspecific fuel consumption, η, measured in kilogram/minute/kilonewton and the thrust, T , to determine the nominal fuel flow, f nom .This is given by:f nom = ηT,(1)where η, for jet aircraft, is:η = C f l 1 + V T AS C f 2 . (2)In this equation, C f 1 and C f 2 are two thrust-specific fuel consumption coefficients reported in the BADA dataset and V T AS is the true airspeed.The thrust depends on the aircrafts phase of flight.For the majority of the resolutions discussed in this study, the aircraft is in the level cruise portion of flight.In this phase, the thrust is equal to the drag and can be represented by the following:T = ρC D S(V T AS ) 2 2 , (3)where ρ is the air density, C D is the drag coefficient reported by BADA, and S is the wing reference area.For idle thrust descent conditions the fuel flow, f min , is measured in kilogram/minute as:f min = C f 3 1 - h C f 4 , (4)where h is the altitude above sea level in feet, C f 3 is the first descent fuel flow coefficient and C f 4 is the second descent fuel flow coefficient.For climb portions of flight, the fuel flow is still given by equation 1, but the thrust is computed based on the type of climb performed by the aircraft.These equations show that, among other things, the fuel burn is a function of thrust, airspeed, and altitude.Even though these equations are only an approximation of the actual fuel burn of an aircraft, they will be used as the true fuel burn for the results which follow.
+ResultsThe six simulation runs presented in Table 1 were performed, and the results were analyzed.Three aspects of the performance of the conflict resolution algorithm were compared between the delay cases and the fuel burn cases: system safety, system efficiency, and resolution selection.Although results were compiled and analyzed on an individual Center basis, no significant differences were observed between Centers.Accordingly, the results are presented in the aggregate.
+SafetyThe main focus of this study is on how the resolution selection criterion affects system efficiency and resolution selection.However, it is also important to determine whether this selection criterion impacts safety.As a first-order look at safety, two metrics were analyzed: the total number of conflicts and the percentage of conflicts successfully resolved.A significant increase in the number of conflicts as a result of selecting resolutions based on fuel burn might suggest increased risk.The total number of conflicts for the delay cases and the fuel burn cases is presented in Table 2. Selection based on fuel burn leads to approximately 5% more conflicts than selection based on delay.This increase may be a by-product of the resolution selection process, but it is not considered large enough to have an impact on system safety.To illustrate this point, only one conflict remained unresolved in all of the delay cases, and only one conflict remained unresolved in all of the fuel burn cases.So, over 99.98% of all conflicts were resolved when either selection criterion was used.
+Delay
+EfficiencyThe operational efficiency of the resolution trajectories produced by the algorithm is important in understanding the advantages and disadvantages of resolution selection based on fuel burn or delay.Successful resolution trajectories that require less fuel or reduce delay are preferable to those that cause an increase in either quantity.
+Cumulative DelayDelay is the time associated with executing resolution maneuvers.Figure 2 shows the cumulative delay for the system when selecting resolution trajectories based on delay or fuel burn.As expected, the results show that the cumulative delay when selecting resolutions based on delay is 25% less than the cumulative delay when selecting resolutions based on fuel burn.This reduced delay can result in economic and system efficiency advantages to selecting resolutions based on minimum delay.The histograms in Figure 3 provide insight into how the delay imposed by the algorithm is distributed for the two resolution selection criteria.These histograms are divided into 30-second time bins with negative times corresponding to resolutions which generate time savings relative to the selected original trajectory.Negative delay results when Direct-To resolutions (Figure 1(a)) are included within the successful resolutions.Direct-To resolutions resolve conflicts by redirecting the aircraft to a downstream waypoint.This directly bypasses a dogleg in the flight plan.Figure 3(a) shows the delay for the cases where resolutions are selected based on minimum delay, and Figure 3(b) shows the delay for the cases where resolutions are selected based on minimum fuel burn.The mean delay for resolutions in Figure 3(a) is 18 seconds.Over 22% of the resolutions in these cases result in a time savings.These values can be contrasted with the results for fuel-burn selection cases shown in Figure 3(b).The mean delay for these resolutions is 40 seconds, and only 11% of the resolutions result in a time savings.It can be seen that for the fuel-burn cases the histogram is more heavily weighted to the right.
+Cumulative Fuel BurnFigure 4 shows the cumulative fuel burn required for conflict resolution for the system when selecting resolutions based on delay or fuel burn.As expected, when the algorithm is selecting based on fuel burn the cumulative fuel burn of the system is 27% less than the fuel burn for delay selection.This fuel-burn reduction could lead to environmental and economic reasons for selecting resolutions maneuvers based on minimum fuel burn.There are tradeoffs evident when comparing Figure 2 for delay selection and Figure 4 for fuel burn selection, and these tradeoffs will be discussed further in the Environmental and Economic Impact Section.The mean fuel burn for the delay selection case is 22 pounds.It can be seen from Figure 5(a) that, for the delay case, the fuel burn distribution is more heavily weighted to the right side of zero, with only a small percentage of resolutions resulting in a fuel savings.In contrast, when selecting resolutions based on fuel burn (Figure 5(b)), the distribution is more evenly weighted with nearly half of the resolutions producing a fuel savings.For this case the mean fuel burn is 12 pounds.
+Resolution SelectionIn the previous sections, the system-wide trade-offs of selecting conflict resolutions based on minimum fuel burn or minimum delay were presented.Since these two different cases produce different results for total delay and total fuel burn, it is interesting to try to comprehend the mechanism for this difference.As a first attempt to understand the underlying differences, the impact of this selection criterion on the types of resolutions that are likely to be selected will now be analyzed.For this analysis, the many different types of resolutions attempted by the Autoresolver will be categorized in three groups according to the dominant method of conflict resolution: horizontal maneuvers, vertical maneuvers, and speed maneuvers.Figure 6 shows the percentage of each type of maneuver selected for the two cases.When selecting based on fuel burn, the percentage of vertical maneuvers is about equal to the percentage when selecting based on delay.The percentage of speed maneuvers based on fuel burn is higher by 3.5% and the percentage of horizontal maneuvers is lower by 3%.
+Figure 6. Selected maneuver typesTo understand the causes of this difference in maneuver selection, the relationship between delay and fuel burn are plotted for a single aircraft type for the Atlanta Center case. Figure 7 shows this relationship for all resolutions (not just selected ones) for Airbus 319 aircraft in the simulation.A single aircraft type was selected to reduce the fuel flow variance.Generally, it might be thought that reducing the delay will reduce fuel burn.Figure 7(a) shows that this is indeed the relationship for horizontal maneuvers.There is a linear variation where increasing delay leads to increasing fuel burn.The multiple trend lines evident in the figure are from resolutions at different altitudes and at different cruise speeds.The relationship is a bit more complex for vertical resolutions (Figure 7(b)).Many of the resolutions plotted in this figure show a linear positive correlation, but there are some cases where resolutions with negative delay led to positive fuel burn.These are probably from altitude-hold resolutions where the aircraft maintains a lower altitude than the cruise altitude for a certain amount of time to avoid a conflict.For speed resolutions (Figure 7(c)) the relationship between delay and fuel burn does not show clearly identifiable trends therefore a linear regression was included to aid in identification.There are many resolutions where increases in delay lead to decreases in fuel burn.These are speed resolutions that command a reduction of cruise speed to avoid the conflict.This speed reduction results in less fuel burn, but greater delay.The relationships between delay and fuel burn for speed resolutions illustrate the differences between the resolution selections shown in Figure 6 as well as the differences in cumulative delay and fuel burn shown in Figures 2 and4.
+Environmental and Economic ImpactThe development of algorithms in support of automated separation assurance should not only be concerned with safe and efficient operations but also be environmentally and economically responsible.In recent years public concern has grown regarding the potential impact of the byproducts of aviation, particularly noise and emissions.It is estimated that aircraft world-wide contribute about 3.5% of the total radiative forcing (a measure of change in climate) off all human activities, and this percentage is projected to grow [8].A contributor to this projected growth is the impending expansion in the level of air traffic demand in the NAS.Ensuring safe and environmentally responsible systems is of utmost importance if the aviation industry is to meet projected levels of growth and demand.
+Fuel BurnThe potential reduction in fuel burn presented in the results section amounted to 10 pounds per resolution when selecting resolutions based on fuel burn.Expanding on this result using a jet fuel (Jet A) price of 220.1 cts/gal and 5,233 the number of resolved conflicts in the three centers from Table 2, the total savings in US dollars over the course of the 36-hour period is $16,963 which would amount to approximately $4 million per year [9].This fuel cost translates to a savings in carbon dioxide emissions.From the above it can be seen that the selection based on minimum fuel consumption would save approximately 52,330 pounds of fuel.Using the weight of the jet fuel, the amount of carbon dioxide released can be determined.The Energy Information Administration estimates that burning a gallon of jet fuel emits 21 pounds of carbon dioxide [10].One gallon of jet fuel weighs on the order of 6.79 pounds per gallon, which would bring the amount of carbon dioxide saved to 161,845 pounds.The projected savings would come at no cost to airlines, as they do not require any modification of existing aircraft or controller practices.The reduction in emissions and cost stems from a change in the way resolutions are selected within the automated conflict resolver.
+DelayThe effects of delay play an important role in airspace management and decision making.For shorter delays the system-wide impact can be relatively small and result in longer flight times that influence the direct operating cost of the airline.However, longer delays can propagate through the system as the day progresses.These delays can prove to be disruptive to activities such as crew scheduling, gate scheduling and even delay later flights.Although selecting resolutions based on minimum fuel burn results in fuel savings, delay is increased.The mean delay of resolutions when selecting based on fuel burn is 40 seconds.Using a delay cost of $20 per minute for the number of resolved conflicts in the three centers from Table 2, and assuming the number of seats in the aircraft is between 65-150, the cost of the delay is $69,773 [11].This is significantly more than the cost of the fuel savings associated with the same number of resolved conflicts when selecting conflict resolutions based on minimum fuel burn.However, cost is only one of several factors that must be taken in to account when evaluating resolution selection criterion.
+ConclusionThe AAC algorithm was modified to select the preferred resolution based on minimum fuel burn by com-paring the aircraft weight at a point along the original trajectory with a common point downstream.In fast-time simulation of three airspace regions, the resolution trajectories were found to incur an average of 40 seconds more delay when selecting conflict resolutions based on minimum fuel burn.This represents a 25% increase over resolutions selected based on minimum delay.Similarly, the trajectories required an average of 10 pounds more fuel when selecting based on delay when compared to selection based on fuel burn.A preference for speed maneuvers was established when selecting resolutions based on minimum fuel burn.Horizontal and vertical maneuvers were found to be less fuel efficient than speed maneuvers when selecting based on fuel burn.When executing horizontal and altitude maneuvers, optimization based on delay was found to be more efficient.Changing the selection criteria from delay to fuel burn was found to have no impact on the ability of the algorithm to successfully detect and resolve conflicts.Despite the modifications, the algorithm was able to successfully detect and resolve 99.98% of all conflicts regardless of the resolution selection criterion.
+Future WorkIn this study, speed maneuvers were found to be the most fuel efficient when selecting resolutions based on minimum fuel burn.However, the number of speed resolutions executed in comparison to horizontal and vertical resolutions is significantly less.Further modification of the algorithm to generate a greater number of speed resolutions would yield higher fuel savings.Similarly, an additional reduction in fuel consumption can be achieved by combining Direct-To resolutions with speed changes.This would serve to reduce the speed by an amount that would cancel the negative delay of the resulting Direct-To resolution.Figure 1 .1Figure 1.Resolution trajectories of type (a) horizontal, (b) vertical, and (c) speed [5].
+Figure 2 .2Figure 2. Cumulative delay.
+Figure 3 .3Figure 3. Delay histograms for (a) minimum delay and (b) minimum fuel burn.
+Figure 4 .4Figure 4. Cumulative fuel burn.
+Figure 5 .5Figure 5. Fuel burn histograms for (a) minimum delay and (b) minimum fuel burn.
+Figure 7 .7Figure 7. Fuel burn versus delay for Airbus 319 aircraft for (a) horizontal maneuvers, (b) vertical maneuvers, and (c) speed maneuvers.
+Table 1 .1Experiment test matrix and simulated flights.CaseFlights SimulatedZIDDelay5413ZIDFuel Burn5413ZMPDelay8577ZMPFuel Burn8577ZTLDelay10049ZTLFuel Burn10049
+Table 2 .2Conflict resolution results.Fuel Burn
+		
+		
+			
+
+				
+
+
+	
+		FAA Aviation Forecasts: Fiscal Years 1981-1992. Federal Aviation Administration, U.S. Department of Transportation, Washington, D.C. 20591. 1980. 69p
+		10.1177/004728758102000159
+		FAA HQ-08371
+	
+	
+		Journal of Travel Research
+		Journal of Travel Research
+		0047-2875
+		1552-6763
+		
+			20
+			1
+			
+			2008
+			SAGE Publications
+		
+	
+	Federal Aviation Administration, 2008, "Terminal Area Forecast Summary, Fiscal Years 2007-2025", FAA HQ-08371.
+
+
+
+	
+		A review of conflict detection and resolution modeling methods
+		
+			JamesKKuchar
+		
+		
+			LCYang
+		
+		10.1109/6979.898217
+	
+	
+		IEEE Transactions on Intelligent Transportation Systems
+		IEEE Trans. Intell. Transport. Syst.
+		1524-9050
+		
+			1
+			4
+			
+			2000
+			Institute of Electrical and Electronics Engineers (IEEE)
+		
+	
+	Kuchar, James K., L C. Yang, 2000, " A Review of Conflict Detection and Resolution Modeling Meth- ods", IEEE Transactions on Intelligent Transportation Systems, Vol. 1, No. 4, pp. 179-189.
+
+
+
+	
+		Automated Conflict Resolution: A Simulation Evaluation Under High Demand Including Merging Arrivals
+		
+			ToddFarley
+		
+		
+			MichaelKupfer
+		
+		
+			HeinzErzberger
+		
+		10.2514/6.2007-7736
+	
+	
+		7th AIAA ATIO Conf, 2nd CEIAT Int'l Conf on Innov and Integr in Aero Sciences,17th LTA Systems Tech Conf; followed by 2nd TEOS Forum
+		Barcelona, Spain
+		
+			American Institute of Aeronautics and Astronautics
+			2007
+		
+	
+	Farley, T C., H. Erzberger, 2007, "Fast-Time Simula- tion Evaluation of a Conflict Resolution Algorithm Un- der High Air Traffic Demand", 7th USA/Europe ATM 2007 R&D Seminar, Barcelona, Spain.
+
+
+
+	
+		Automated Conflict Resolution for Air Traffic Control
+		
+			HErzberger
+		
+	
+	
+		25th International Congress of the Aeronautical Sciences (ICAS)
+		Hamburg, Germany
+		
+			2006
+		
+	
+	Erzberger, H., 2006, "Automated Conflict Resolution for Air Traffic Control", 25th International Congress of the Aeronautical Sciences (ICAS), Hamburg, Germany.
+
+
+
+	
+		Automated conflict resolution, arrival management, and weather avoidance for air traffic management
+		
+			HErzberger
+		
+		
+			TALauderdale
+		
+		
+			Y-CChu
+		
+		10.1177/0954410011417347
+	
+	
+		Proceedings of the Institution of Mechanical Engineers, Part G: Journal of Aerospace Engineering
+		Proceedings of the Institution of Mechanical Engineers, Part G: Journal of Aerospace Engineering
+		0954-4100
+		2041-3025
+		
+			226
+			8
+			
+			2010
+			SAGE Publications
+			Nice, France
+		
+	
+	Erzberger, H., T Lauderdale, Y. C Chu, 2010, "Automated Conflict Resolution, Arrival Management and Weather Avoidance For ATM", 27th International Congress of the Aeronautical Sciences (ICAS), Nice, France.
+
+
+
+	
+		Fast-Time NAS Simulation System for Analysis of Advanced ATM Concepts
+		
+			DouglasSweet
+		
+		
+			VikramManikonda
+		
+		
+			JesseAronson
+		
+		
+			KarlinRoth
+		
+		
+			MatthewBlake
+		
+		10.2514/6.2002-4593
+		AIAA- 2002-4593
+	
+	
+		AIAA Modeling and Simulation Technologies Conference and Exhibit
+		Monterey, CA
+		
+			American Institute of Aeronautics and Astronautics
+			2002. 2004
+		
+	
+	User Manual For The Base of Aircraft Data (BADA). Revision 3.6
+	Sweet, D., V. Manikonda, J. Aronson, K. Rot, M. Blake, 2002, "Fast-Time Simulation System for Analy- sis of Advanced Air Transportation Concepts", AIAA- 2002-4593, AIAA Modeling and Simulation Technolo- gies Conference and Exhibit, Monterey, CA [7] European Organisation For the Safety of Air Naviga- tion, 2004, "User Manual For The Base of Aircraft Data (BADA)", Revision 3.6
+
+
+
+	
+		Intergovernmental Panel on Climate Change
+		10.4135/9781452218564.n376
+	
+	
+		Aviation and the Global Atmosphere
+		
+			SAGE Publications, Inc.
+			
+		
+	
+	Intergovernmental Panel on Climate Change, 1999, "Aviation and the Global Atmosphere"
+
+
+
+	
+		International Air Transport Association (IATA)
+		
+			BarryTurner
+		
+		10.1007/978-1-349-58635-6_31
+		
+	
+	
+		The Statesman’s Yearbook
+		
+			Palgrave Macmillan UK
+			2010
+			
+		
+	
+	Air Transport Association, 2010 "Jet Fuel Price Mon- itor", www.iata.org
+
+
+
+	
+		Voluntary reporting of greenhouse gases 1997
+		10.2172/348897
+		
+		
+			
+			Office of Scientific and Technical Information (OSTI)
+		
+	
+	Energy Information Administration
+	Energy Information Administration, " Vol- untary Reporting of Greenhouse Gases Program", www.eia.doe.gov
+
+
+
+	
+		
+			A QKara
+		
+		Estimating Domestic U.S Airline Cost of Delay based on European Model" ICRAT 2010 29th Digital Avionics Systems Conference
+		
+			2010. October 3-7, 2010
+		
+	
+	Kara. A . Q , 2010, "Estimating Domestic U.S Air- line Cost of Delay based on European Model" ICRAT 2010 29th Digital Avionics Systems Conference October 3-7, 2010
+
+
+				
+			
+		
+	

file102.txt

@@ -0,0 +1,330 @@
+
+	
+	
+		
+IntroductionAir traffic demand is projected to double in the next 20 years (ref.1).The human workload associated with conflict detection and resolution is expected to limit this increase and thereby limit the economic growth that aviation facilitates.Automated separation-assurance systems are proposed as a way to safely and efficiently separate aircraft in highly dense traffic situations up to two to three times current levels, thereby fostering increased economic growth for the nation.Numerous algorithms have been proposed to provide separation assurance in the future air traffic system (ref.2).Maintaining safe separation is the first-order objective of all such algorithms; the second-order objectives vary, but most of the proposed algorithms optimize the selection of conflict resolution maneuvers to minimize airborne delay in order to mitigate the effect on schedule.An alternative objective is to optimize based on fuel burn (refs. 2 and 3).The Advanced Airspace Concepts (AAC) Autoresolver is strategic conflict resolution algorithm capable of deconflicting aircraft.AAC is a concept for automating separation assurance in the future that includes multiple layers of separation assurance for increased reliability.One component of AAC is the Autoresolver, a strategic problem-solving tool that is responsible for strategic separation assurance as well as weather avoidance and arrival metering, although for this study the focus is only on the separation-assurance function (refs. 4 and 5).In reference 6, the system performance of a conflict resolution algorithm that selected resolutions based on minimum delay was compared to the system performance of the same algorithm when selecting resolutions based on minimum fuel burn.The most effective resolution maneuver when optimizing for airborne delay was a Direct-to maneuver, which identifies wind-favorable shortcuts along the planned route of an aircraft that reduce its flying time while resolving the predicted conflict (ref.7).The most effective resolution maneuver when optimizing for fuel burn was a speed reduction maneuver, which employs a temporary speed reduction to resolve the predicted conflict.However, speed reductions were selected less frequently by the algorithm than other, lessfuel-efficient maneuvers.Additionally, when utilized, these maneuvers significantly increase the cumulative delay.It is hypothesized that the availability of a compound maneuver combining a Direct-To maneuver with the fuel efficiency of a speed reduction would improve the performance of the separation-assurance algorithm.This study compares the system performance of a conflict resolution algorithm in realistic traffic scenarios with and without the availability of a compound Direct-to/speed-reduction maneuver, hereafter referred to as a Variable Speed Direct-To maneuver.The objective is to quantify the operational benefit of adding the proposed new maneuver to the set of maneuvers already available to the automated separation-assurance algorithm.The next section describes the conflict resolution algorithm under test and the new compound maneuver.Then the experimental approach, procedure, and assumptions are discussed.The results are then categorized according to safety and efficiency.Lastly, a summary of the study findings is given, along with suggestions for future research.
+Test ArticleThe conflict resolution algorithm evaluated in this study is the Advanced Airspace Concept (AAC) Autoresolver (refs. 4 and 5) .It is a groundbased algorithm that resolves conflicts in pairwise fashion and can be configured to select resolutions based on minimum delay or minimum fuel burn.The Autoresolver selects a maneuver from one of the following categories: horizontal, vertical, altitude, Direct-To, or compound.For this study, only conflicts with en-route flight maneuvers are analyzed; arrivals are not included because they adhere to additional constraints such as metering.In the following study only one compound maneuver is enabled: the Variable Speed Direct-To maneuver.
+AAC AutoresolverThe AAC Autoresolver is a strategic conflict resolution algorithm designed to deconflict aircraft that are predicted to lose separation more than 2 minutes in the future.For aircraft en route, the look-ahead time is 8 minutes and up to 20 minutes for arriving aircraft.In this study, for every minute of simulation time the future trajectories of all aircraft are computed and processed to determine if there are any predicted losses of separation where two aircraft come within 5 nautical miles horizontally and 1,000 feet vertically of one another.The Autoresolver receives a list of aircraft conflict pairs ordered by predicted time to first loss of separation.For each conflict in the con-flict list, the Autoresolver follows an iterative approach for resolution.Accounting for characteristics such as aircraft type, speed, and airspace boundaries, the resolver calculates a future route composed of waypoints, speeds, and altitudes that may possibly resolve the conflict.Figure 1 shows the types of future routes attempted by the Autoresolver, grouped by whether they are horizontal, vertical, or speed resolutions.This future route is then sent to a trajectory engine that computes a four-dimensional (4-D) trial resolution trajectory based on this route.In order for the resolution to be viable, it must resolve the primary conflict and be free of predicted losses of separation with the primary aircraft in the conflictas well as any other aircraft in the airspacefor a specified period of time.If these conditions are met, the Autoresolver has successfully generated a candidate resolution trajectory and stores it.If the resolution is not free of primary or secondary conflicts, the Autoresolver computes a new resolution route and checks to determine if it is successful.For each resolution type this iteration is continued until a successful resolution is found or all possibilities of that type have been exhausted.For each successful resolution, both the associated delay and the fuel burn are calculated.A common spatial point on the original trajectory and the resolution trajectory is found.To calculate the delay, the time on the original trajectory at the common point is subtracted from the time on the resolution trajectory at the common point.Similarly for the fuel burn, the weight of the aircraft at the common point on the resolution trajectory is subtracted from the aircraft weight at that point, after the aircraft has flown the original trajectory.Figure 2 shows an example trajectory with a resolution maneuver represented by segments 3a and 3b.The algorithm evaluates the cost of segments 1, 2, and 3 versus the cost of segments 1, 2, 3a, and 3b.A discussion of how the aircraft weight is calculated and converted to fuel burn is given in a subsequent section.The resolver will generate up to 18 successful resolutions per aircraft in conflict for a maximum of 36 candidate resolution maneuvers between the two aircraft.In this study, the algorithm selects a resolution from among the set of successful resolutions using either the minimum delay or the minimum fuel-burn criterion, depending on how the algorithm is configured.The selected resolution is then implemented via fast-time, closed-loop simulation as discussed in the following sections.Further discussion regarding the design of the algorithm and the types of resolutions that are generated is presented in references 4 and 5 .
+Variable Speed Direct-To ManeuverThe Autoresolver was modified to allow for the combination of a Direct-To maneuver with a reduction in speed.The reduced speed is chosen to exactly negate the time savings normally associated with a Direct-To maneuver; this reduction in speed produces a fuel-burn benefit while maintaining the flight-plan schedule.This compound maneuver is referred to as a Variable Speed Direct-To maneuver.A Direct-To maneuver resolves a conflict by taking an aircraft directly to a downstream waypoint, thus bypassing a dogleg in the flight plan.This modification augmented the existing Direct-To maneuver, thus allowing the algorithm to continue to have the option to utilize a Direct-To maneuver when efficient.The equation that describes a Direct-To maneuver is shown in reference 1, where ∆t [[need to fix all these symbol callouts]] represents delay in hours, D 1 is the previous distance along the route in nautical miles, D 2 is the new distance in nautical miles, and S is speed in knots:∆t = D 1 S - D 2 S(1)Augmenting equation ( 1) to produce a maneuver that results in zero delay requires setting d to zero, yielding equation (2), where S new represents the new (slower) speed in order to result in a Variable Speed Direct-To maneuver.The algorithm abides by the original Direct-To constraints where the maneuver will not be considered if:• the aircraft is less than 20 minutes from the arrival fix,• the aircraft cannot return to the route within 50 n.mi. of the final fix,• the path of the aircraft along the Direct-To route is greater than 250 n. mi.(dotted line in fig.1(a)), and• the point where the aircraft rejoins the trajectory is within 50 n.mi. of the current Air Route Traffic Control Center boundary.In addition, it will not attempt to execute the maneuver if S new is within 5 knots of the original speed.S new = ( D 2 D 1 )S(2)For example, a Variable Speed Direct-To maneuver by an aircraft traveling 450 knots that will reduce the distance along the route from 400 to 360 n. mi.would reduce the speed to 405 (by 45) knots in order to produce no delay.When performing a Variable Speed Direct-To maneuver, the intent is for the aircraft to recapture the route at the same time it would have if it had not performed the maneuver.Figure 3 illustrates the Variable Speed Direct-To maneuver where A1 and A2 are aircraft predicted to conflict.To avoid this conflict, A1 is selected to execute a Variable Speed Direct-To maneuver.The new trajectory for A1 (dashed line) removes several waypoints and reduces the speed as shown in the neighboring profile.The Mach number of A1 is decreased for the duration of the maneuver and eventually returns to its original speed after clearing the conflict.
+Experiment DesignThis section describes the simulation approach and the metrics used.
+Simulation EnvironmentThe Advanced Concepts Evaluations System (ACES) is a fast-time, agent-based simulation of the National Airspace System (NAS) that uses four-degree-of-freedom equations based on the Base of Aircraft Data (BADA) to generate aircraft trajectories (ref.8).ACES was developed specifically to provide a general-purpose environment for evaluating future air traffic management and control concepts, including automated resolution algorithms.Essential to the simulation of resolution algorithms is the ability to generate 4-D trajectories.In ACES these trajectories begin at the departure fix and end at the arrival fix.By using aircraft-type-specific performance data together with guidance and navigation models, the ACES trajectory engine can generate representative trajectories for many aircraft.In the ACES simulation environment these aircraft trajectories are entirely deterministic; aircraft conflicts can be predicted with perfect accuracy, and resolution trajectories are guaranteed to be followed precisely by the simulated aircraft.In addition to deterministic aircraft trajectories, simplifications were made in the modeling and execution of the experiment.Negotiations of resolution trajectories between aircraft operators and/or the air navigation service provider were not modeled, nor were data-link transmission delays or pilot-action delays.Once a resolution trajectory was determined by the automation, it was executed immediately and precisely.
+Simulated AirspaceIn this study, the Autoresolver resolved conflicts in three pairs of adjacent Air Route Traffic Control Centers (ARTCCs).Each of the airspaces was simulated independently of each other and was selected based on its operational conflict properties as defined in reference 9.These properties fall into three categories that characterize the conflict, its relationship between two or more conflicts, and the locations of the conflicts within the NAS.In this study statistical clustering analysis was employed to categorize ARTCCs based on normalized conflict properties.As a result, three ARTCC pairs: Oakland-Los Angeles (ZOA-ZLA), Indianapolis-Chicago (ZID-ZAU), and Boston-New York (ZBW-ZNY) were identified that provided a wide representation of conflict properties.In order to create three distinct NAS regions, an adjacent center was chosen for each pair, creating (ZOA-ZLA) as representative of West Coast air traffic flow, (ZAU-ZID) as representative of Midwest air traffic flow, and (ZBW-ZNY) as representative of East Coast air traffic flow.By using the clusters shown in figure 4, we can model behavior seen over the entire NAS, thus allowing a more complete assessment of the performance of the algorithm.Figure 4: The ARTCCs simulated in this study.• CL/CL -Both aircraft are climbing.
+Demand Set• CL/CR-One aircraft is climbing while the other is cruising.• CL/DE -One aircraft is climbing while the other is descending.• CR/CR-Both aircraft are cruising.• CR/DE -One aircraft is cruising while the other is descending.• DE/DE-Both aircraft are descending.
+Independent VariablesTo evaluate the difference between the current state-of-the-art conflict resolution algorithm and the addition of a Variable Speed Direct-To maneuver, a test plan was developed that examines the behavior of the algorithm with and without this maneuver enabled in three pairs of ARTCCs under two conflict resolution optimization schemes.Table 1 shows the independent variables and settings.Each of the possible permutations is representative of a simulation run.
+Dependent VariablesThe dependent variables for the experiment were the number of conflicts and the airborne delay and fuel burn incurred by flying the conflict resolution trajectories.In the development of a robust, efficient algorithm for implementation in the Next-Generation Air Transportation System (NextGen), safety is of the utmost concern.The number of conflicts is the metric used here to reflect the safety of the system.Efficiency in terms of delay and fuel burn is important once safety is assured.The fuel consumed per resolution is computed by ACES using aircraft-specific coefficients selected from the BADA (ref.8).The BADA comprises the performance and operating procedure coefficients of 295 aircraft types.These coefficients encompass those that are used to calculate thrust, drag, and fuel flow along with those used to specify nominal cruise, climb, and descent speeds.Further discussion of the specific equations used to calculate the fuel burn is included in references 6 and 8. Evaluating the number of conflicts per simulation provides insight into the impact of the modifications made to the algorithm.A significant increase in the number of conflicts as a result of the availability of the Variable Speed Direct-To maneuver suggests increased risk.The safetyand efficiency-related results are presented in the section Results.
+ResultsThis experiment seeks to evaluate the benefit of augmenting the AAC Autoresolver to consider a Variable Speed Direct-To maneuver when resolving a given conflict.The subsequent results address the safety and efficiency of potential implementation.
+SafetyThe primary safety metric for the experiment is the number of conflicts.A conflict occurs when aircraft are predicted to come within 5 n.mi.horizontally and 1,000 feet vertically from each other in en-route airspace.As expected, the addition of the Variable Speed Direct-To maneuver did not adversely affect the safety of the system, as measured by the total number of predicted conflicts.Figure 6 shows that in none of the test airspaces did the number of conflicts significantly increase when the Variable Speed Direct-To maneuver was enabled.On average, the percent difference between the baseline number of conflicts and the Variable Speed Direct-Toenabled scenario is less than 1%, suggesting that the inclusion of this maneuver does not adversely affect the ability of the algorithm to resolve conflicts, and there are no major gaps in its implementation.
+Efficiency
+Fuel BurnWhen a Variable Speed Direct-To maneuver is executed, the maneuvered aircraft is slowed by an amount such that it will traverse its now, shorter Direct-To route in the same amount of time that it planned to traverse its original route.Figure 7 shows the distribution of speed-reduction magnitudes for ZID-ZAU.Seventy-five percent of all speed reductions observed in the experiment were less than 30 knots.A typical Boeing 737 aircraft at 35,000 feet will cruise between Mach 0.72 (415 knots) and Mach 0.76 (438 knots), approximately a 30-knot variation, indicating that most of the speed-reduction values required to obtain the desired fuel benefit are reasonable.Within our simulation, the range observed adhered to aircraft performance limitations.The speed-reduction ranges vs. the number of Variable Speed Direct-To maneuvers for the selected airspaces are shown in appendix A. To evaluate the fuel burn associated with a resolution maneuver, the weight of the aircraft at the termination point on the resolution trajectory (where the aircraft rejoins the original trajectory) is subtracted from the aircraft weight at the same point, after the aircraft has flown the original trajectory.Fuel-burn savings were higher by 92% in ZID-ZAU, 55% in ZBW-ZNY, and 47% in ZLA-ZOA when resolving conflicts with the Variable Speed Direct-To maneuver enabled.Figure 8 shows the average fuel burn per resolution for the selected airspaces.The negative fuel burn seen in ZBW-ZNY and ZID-ZAU is an indication that the modification made to the algorithm causes it to outperform the nominal case when selecting resolutions based on minimum fuel burn.The average fuel burn per resolution in ZID-ZAU is 4.01 pounds less than when selecting resolutions based on minimum fuel burn with Variable Speed Direct-To maneuvers enabled.Similarly, in ZBW-ZNY the average fuel burn per resolution is 7.04 when optimizing for fuel burn with the maneuver enabled, a 2.43pound-per-resolution decrease.In ZLA-ZOA the average fuel burn per resolution is 2.73 pounds, 2.41 pounds less than when Variable Speed Direct-To is disabled.Though these numbers are small, they are not insignificant when extrapolated to potential savings per year.In this study there were 3,276 conflicts in ZID-ZAU over the course of the day.Each of these conflicts requires one of the two aircraft to be maneuvered.Considering the average fuel savings of 4 pounds per resolution in ZID-ZAU, this savings amounts to roughly 4.8 million pounds of fuel per yearenough fuel to fill the tank of a Boeing 737-700 approximately 100 times.Furthermore, 20 ARTCCs within the continental United States could benefit from these savings.Variation in traffic density and route length accounts for most of the difference in the magnitude of savings between the centers.ZID-ZAU center executed nearly twice as many resolution maneuvers as ZLA-ZOA and ZBW-ZNY, suggesting that the fuel efficiency of the resolutions the algorithm selects increases with the air traffic demand.However, the improvement seen in the delay cases is not as significant.When selecting resolutions based on delay, the algorithm finds Direct-To maneuvers to be more efficient.This increase in efficiency can be attributed to the fact that the selection of a Direct-To maneuver can result in negative delay and thus a time savings, whereas the most time-efficient zero-delay solution is zero and will not yield a time savings.Figure 9 shows the resolutions selected by the algorithm for ZID-ZAU for fuel-burn optimization with Variable Speed Direct-To maneuvers enabled and disabled.Overall, the number of resolutions other than Direct-To or Variable Speed Direct-To remains consistent between scenarios.When Variable Speed Direct-To maneuvers were disabled, 306 Direct-To maneuvers were executed.When enabled, 181 Direct-To and 147 Variable Speed Direct-To maneuvers were executed, representing a 41% decrease in the number of Direct-To maneuvers.When optimizing for minimum fuel burn, the algorithm frequently selected Variable Speed Direct-To maneuvers over traditional Direct-To maneuvers.However, in a small number of cases, a Direct-To maneuver was selected despite the fact that a Variable Speed Direct-To maneuver was available.In these instances, the additional fuel savings did not outweigh a decrease in flight time.The maneuver types for delay and fuel-burn optimization for each of the ARTCCs are shown in appendix B.
+DelayAirborne delay is defined as the difference in time between the arrival time of an aircraft as given in the flight schedule and its actual arrival time.Although there are many sources of delay (e.g., air traffic control, weather, maintenance, crew availability), in the following analysis the source of delay is attributed to time difference between the modified trajectory and original trajectory of an aircraft at a common point in the en-route airspace.Positive delay occurs when the modified trajectory incurs additional flight time to avoid a loss of separation, akin to a detour.Negative delay is a reduction in flight time that can occur when a dogleg in the flight plan is eliminated or more favorable winds are encountered.As expected, the inclusion of a Variable Speed Direct-To maneuver has almost no impact on delay when selecting resolutions based on delay.Figure 10 shows the average delay per resolution.When selecting resolu-tions based on minimum fuel burn, the average delay per resolution with the Variable Speed Direct-To maneuver enabled for ZID-ZAU is 10.86 seconds.For ZLA-ZOA under the same conditions, the average delay is 16.84 seconds, and it is 11.08 seconds for ZBW-ZNY.This delay translates to a 20% increase in cumulative delay in ZID-ZAU, but the absolute difference is only 4.7 minutes.Likewise, in ZLA-ZOA and ZBW-ZNY the difference is less than 1% when selecting resolutions based on delay.This finding supports the initial assertion that cumulative delay would only marginally increase with the availability of the Variable Speed Direct-To maneuver when resolution trajectories are optimized for airborne delay.Selecting resolutions based on minimum fuel burn increases the cumulative delay in each center.Figure 11 shows the cumulative delay per center for each optimization.This effect can be attributed to the selection of Variable Speed Direct-To maneuvers to resolve the associated conflict.As opposed to Direct-To maneuvers that can potentially yield a time savings, these maneuvers result in zero delay benefit.Additionally, when optimizing for fuel burn, the algorithm prefers speed-reduction maneuvers that tend to increase delay within the system.Each implementation of Variable Speed Direct-To changes the way the primary and, consequently, secondary conflicts are solved.Because only one aircraft of the pair will be maneuvered to avoid a conflict, the average delay per resolution can be thought of as per aircraft.Generally, increasing the delay is considered to be undesirable.However, there are strategic instances in which this increase could be of value, such as an aircraft that needs to be slowed in order to meet the requested time of arrival.The magnitude of additional delay per resolution is small when compared to the 15-minute FAA definition of a reportable delay (ref. 11).The largest amount of delay per resolution observed when optimizing for fuel burn utilizing the Variable Speed Direct-To maneuver was 4 minutes.Even if marginal, the system-wide effects of an increase in delay are difficult to determine.
+ConclusionsTwelve conditions were simulated to evaluate the benefit of modifying the AAC Autoresolver to consider a Variable Speed Direct-To maneuver when resolving a given conflict.Two methods of resolution selection were used: minimum delay and minimum fuel burn.The experiment was conducted in a fast-time environment using data representing a reasonable traffic day in the NAS.The results showed that augmenting the existing algorithm to include the compound maneuver did not significantly influence the ability of the algorithm to resolve conflicts, nor did it affect the number of conflicts observed.The inclusion of Variable Speed Direct-To increased the cumulative fuel-burn savings by 92% in ZID-ZAU, 55% in ZBW-ZNY, and 47% in ZLA-ZOA when selecting resolutions based on minimum fuel burn.In these results, the average penalty in delay per aircraft was on the order of a few seconds.Further analysis is required to determine the effect of increasing the delay as well as the balance between delay and fuel-burn benefit.The cumulative fuel-burn savings observed in this study suggests that the Variable Speed Direct-To maneuver could provide significant fuel savings with no significant effect on safety or schedule.Figure 1 :1Figure 1: Resolution trajectories of type horizontal (a), vertical (b), and speed (c).
+Figure 2 :2Figure 2: Delay and fuel burn estimation.
+Figure 3 :3Figure 3: Variable Speed Direct-To maneuver.
+Flightoperations over a 24-hour period were simulated based on Aircraft Situation Display to Industry (ASDI) data recorded March 8, 2007.ASDI data come from the Federal Aviation Administrations (FAA's) Enhanced Traffic Management System (ETMS) and contain information about flights controlled by air traffic control.The dataset included 62,970 flights, their associated routes, and their departure times.This dataset had mixed aircraft types representing the current fleet mix.The data used in this study represent reasonable daily traffic in the NAS.The Rapid Update Cycle wind data were used to model winds in the selected ARTCCs (ref.10). Figure 5 shows the conflict types represented within the demand set by ARTCC. Figure 5 illustrates a diversity of traffic flow types, with the East Coast containing primarily transitioning traffic, the Midwest predominately cruising traffic, and the West Coast a mix of all traffic types.The conflicts are coded as follows:
+Figure 5 :5Figure 5: Conflict types per center.
+Figure 6 :6Figure 6: Number of conflicts.
+Figure 7 :7Figure 7: Variable Speed Direct-Toenabled speed reduction for ZID-ZAU; fuel burn optimal.
+Figure 8 :8Figure 8: Average fuel burn.
+Figure 9 :9Figure 9: Resolution types in ZID-ZAU when optimizing for fuel burn.
+Figure 10 :10Figure 10: Average delay
+Figure 11 :11Figure 11: Cumulative delay
+FigureFigure A2: -Variable Speed Direct-Toenabled speed reduction, ZBW-ZNY.
+FigureFigure A3: -Variable Speed Direct-Toenabled speed reduction, ZLA-ZOA
+FigureFigure B2: -Maneuver types for all centers, fuel-burn optimization.
+
+
+TABLE 1 .1-INDEPENDENT VARIABLES
+Table 1 .1INDEPENDENT VARIABLES.Independent VariablesSettingsVariable Speed Direct-To ManeuverEnabled, DisabledOptimizationDelay, Fuel BurnAirspaceZID-ZAU, ZBW-ZNY, ZOA-ZLA
+		
+		
+
+			
+AcknowledgmentsThe authors wish to acknowledge Dr. Todd Lauderdale, whose contributions to this work were invaluable.The authors also thank Todd Farley and Drs.Antony Evans and Banavar Sridhar for their insightful suggestions and thoughtful review.
+			
+
+			
+Appendix AA.0 A.0 This appendix includes supplemental plots for speed distribution, maneuver types, and distance from final fix.When a Variable Speed Direct-To maneuver is executed, the maneuvered aircraft is slowed by an amount such that it will traverse its now, shorter Direct-To route in the same amount of time that it planned to traverse its original route.Figures A-1 through A-3 show the distribution of speed-reduction magnitudes for the simulated airspaces.The majority of all speed reductions observed in the experiment were less than 30 knots.
+REPORT DOCUMENTATION PAGE
+Form Approved OMB No. 0704-0188The public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information.Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden, to Department of Defense, Washington Headquarters Services, Directorate for Information Operations and Reports (0704-0188), 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202-4302.Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to any penalty for failing to comply with a collection of information if it does not display a currently valid OMB control number.
+PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS.Standard Form 298 (Rev.8/98)Prescribed by ANSI Std.Z39.18
+REPORT DATE (DD-MM-YYYY)01-11-2012
+REPORT TYPE
+Technical Memorandum
+DATES COVERED (From -To)
+TITLE AND SUBTITLEA Fuel-Efficient Conflict Resolution Maneuver for Separation Assurance  An electronic version can be found at http://ntrs.nasa.gov.
+ABSTRACTThis experiment seeks to evaluate the benefit of augmenting a conflict detection and resolution algorithm to consider a fuel-efficient, Variable Speed Direct-To maneuver when resolving a given conflict based on either minimum fuel burn or minimum delay.Twelve conditions were tested in fast-time simulation conducted in three airspace regions with mixed aircraft types and nominal traffic.Inclusion of this maneuver had no appreciable effect on the ability of the algorithm to safely detect and resolve conflicts.Cumulative fuel-burn savings were significantly higher when selecting resolutions based on minimum fuel burn; average delay per resolution was only marginally higher.			
+			
+
+				
+
+
+	
+		Terminal Area Forecast 1977-1987. Aviation Forecast Branch, Office of Aviation Policy, Federal Aviation Administration, Department of Transportation, Washington, D.C. 20591. February 1976. Various paging
+		10.1177/004728757701500317
+	
+	
+		Journal of Travel Research
+		Journal of Travel Research
+		0047-2875
+		1552-6763
+		
+			15
+			3
+			
+			2011
+			SAGE Publications
+		
+	
+	Tech. Rep. HQ121529
+	Terminal Area Forecast Summary Fiscal Years 2011-2040. Tech. Rep. HQ121529, Federal Aviation Administration, 2011.
+
+
+
+	
+		A review of conflict detection and resolution modeling methods
+		
+			JKKuchar
+		
+		
+			LCYang
+		
+		10.1109/6979.898217
+	
+	
+		IEEE Transactions on Intelligent Transportation Systems
+		IEEE Trans. Intell. Transport. Syst.
+		1524-9050
+		
+			1
+			4
+			
+			2000
+			Institute of Electrical and Electronics Engineers (IEEE)
+		
+	
+	Kuchar, J.K.; and Yang, L.C.: A Review of Conflict Detec- tion and Resolution Modeling Methods. IEEE Transactions on Intelligent Transportation Systems, vol. 1, no. 4, 2000, pp. 179189.
+
+
+
+	
+		Foundations of mediation training: A literature review of adult education and training design
+		
+			TimothyHedeen
+		
+		
+			SusanSRaines
+		
+		
+			AnsleyBBarton
+		
+		10.1002/crq.20018
+	
+	
+		Conflict Resolution Quarterly
+		Conflict Resolution Quarterly
+		1536-5581
+		
+			28
+			2
+			
+			2010. 2010
+			Wiley
+		
+	
+	Tech. Rep
+	Literature Review of Conflict Resolution Research. Tech. Rep. 2010, Federal Aviation Administration, 2010.
+
+
+
+	
+		Automated Conflict Resolution for Air Traffic Control
+		
+			HErzberger
+		
+		
+			2006
+		
+	
+	25th International Congress of the Aeronautical Sciences
+	Erzberger, H.: Automated Conflict Resolution for Air Traf- fic Control. 25th International Congress of the Aeronautical Sciences, 2006.
+
+
+
+	
+		Automated conflict resolution, arrival management, and weather avoidance for air traffic management
+		
+			HErzberger
+		
+		
+			TALauderdale
+		
+		
+			Y-CChu
+		
+		10.1177/0954410011417347
+	
+	
+		Proceedings of the Institution of Mechanical Engineers, Part G: Journal of Aerospace Engineering
+		Proceedings of the Institution of Mechanical Engineers, Part G: Journal of Aerospace Engineering
+		0954-4100
+		2041-3025
+		
+			226
+			8
+			
+			2010
+			SAGE Publications
+			Nice, France
+		
+	
+	Erzberger, H.; Lauderdale, T.A.; and Cheng, Y.: Automated Conflict Resolution, Arrival Management and Weather Avoid- ance for ATM. 27th Intl. Congress Aeron. Sci., Nice, France, 2010.
+
+
+
+	
+		Selecting conflict resolution maneuvers based on minimum fuel burn
+		
+			AishaBowe
+		
+		
+			ToddLauderdale
+		
+		10.1109/dasc.2010.5655529
+	
+	
+		29th Digital Avionics Systems Conference
+		
+			IEEE
+			2010
+		
+	
+	Bowe, A.; and Lauderdale, T.: Selecting conflict resolution maneuvers based on minimum fuel burn. Digital Avionics Sys- tems Conf., 2010.
+
+
+
+	
+		Direct-To Tool For En Route Controllers
+		
+			HeinzErzberger
+		
+		
+			DavidMcnally
+		
+		
+			MichelleFoster
+		
+		
+			DannyChiu
+		
+		
+			PhilippeStassart
+		
+		10.1007/978-3-662-04632-6_11
+	
+	
+		New Concepts and Methods in Air Traffic Management
+		Capri, Italy
+		
+			Springer Berlin Heidelberg
+			1999
+			
+		
+	
+	Erzberger, H.; McNally, B.D.; Forester, M.; Chiu, D.; and Stassart, P.: Direct-To Tool for En Route Controllers. ATM '99: IEEE Workshop on Advanced Technologies and their Im- pact on Air Traffic Management in the 21st Century, Capri, Italy, 1999.
+
+
+
+	
+		User Manual for the Base of Aircraft Data (BADA) Revision 3.8
+		
+			ANuic
+		
+		
+			April 2010
+			EUROCONTROL Experimental Centre
+		
+	
+	Tech. Rep. 2010-003
+	Nuic, A.: User Manual for the Base of Aircraft Data (BADA) Revision 3.8. Tech. Rep. 2010-003, EUROCONTROL Exper- imental Centre, April 2010.
+
+
+
+	
+		Analysis of the Aircraft to Aircraft Conflict Properties in the National Airspace System
+		
+			MikePaglione
+		
+		
+			ConfesorSantiago
+		
+		
+			RobertOaks
+		
+		
+			AndrewCrowell
+		
+		10.2514/6.2008-7143
+	
+	
+		AIAA Guidance, Navigation and Control Conference and Exhibit
+		
+			American Institute of Aeronautics and Astronautics
+			2008
+		
+	
+	Paglione, M.M.; Santiago, C.; Crowell, A.; and Oaks, R.D.: Analysis of the Aircraft to Aircraft Conflict Properties in the National Airspace System. AIAA Guidance, Navigation, and Control Conf., 2008.
+
+
+
+	
+		National Oceanic and Atmospheric Administration (NOAA)
+		10.4135/9781412994064.n179
+		
+	
+	
+		The Rapid Update Cycle (RUC)
+		
+			SAGE Publications, Inc.
+			April 2012
+		
+	
+	The Rapid Update Cycle (RUC). Tech. rep., National Oceanic and Atmospheric Administration, April 2012, http://ruc.noaa.gov/Welcome.cgi.
+
+
+
+	
+		Table 3: The percentages of the top 20% NMI features from each omics data.
+		10.7717/peerj.9440/table-3
+		N: 100 Max: 4164 nmi 75%: 1318.75 nmi Median: 394 25%: 236.5 nmi Min: 132 nmi
+	
+	
+		Federal Aviation Administration
+		
+			PeerJ
+			2011
+		
+	
+	Tech. rep
+	Operational Data Reporting Requirements (OPSNET). Tech. rep., Federal Aviation Administration, 2011. N: 100 Max: 4164 nmi 75%: 1318.75 nmi Median: 394 25%: 236.5 nmi Min: 132 nmi
+
+
+
+	
+		Figure 3: A C major scale, starting from C4 and finishing at C5, and going back to C4.
+		10.7717/peerjcs.229/fig-3
+	
+	
+		Distance from fix
+		
+			PeerJ
+			null
+			4
+		
+	
+	Figure C4: -Distance from fix, ZBW-ZNY.
+
+
+				
+			
+		
+	

file103.txt

@@ -0,0 +1,288 @@
+
+	
+	
+		
+I. IntroductionA ir traffic demand is projected to increase significantly in the upcoming years. 1 The human workload associated with conflict detection and resolution is expected to limit this increase and thereby limit the economic growth that aviation facilitates.Automated separation assurance systems are proposed as a way to safely and efficiently separate aircraft in highly dense traffic conditions up to two to three times current levels.Numerous algorithms have been proposed to provide separation assurance in the future air traffic system. 2,3 aintaining safe separation is the first-order objective of all such algorithms; however, the secondorder objectives can vary and are the focus of much research in the field of air traffic management.With any automated resolution tool, the resolution selected is based on some criterion function.The majority of the proposed algorithms optimize the selection of conflict resolution maneuvers to minimize airborne delay in order to mitigate the effect on schedule.However, the true cost of operations is more complex with considerations beyond delay.With the tremendous rise in fuel price over the past few years, examination of the implications of fuel price has increased in relevancy.A additional objective, then, is to optimize based on fuel burn.Prior research, as in Ref. 4, showed that the minimum-delay solution was rarely the same as the minimum-fuel-burn solution.In Ref. 4, the system performance of a conflict resolution algorithm that selected resolutions based on minimum-delay was compared to the system performance of the same algorithm when selecting resolutions based on minimum-fuel-burn.The most effective maneuver when minimizing for fuel burn was a speed reduction maneuver, which employs a temporary speed reduction to resolve the predicted conflict.However, speed reductions were selected less frequently than other, less fuel-efficient maneuvers.Additionally, when utilized, speed reduction maneuvers significantly increased the cumulative delay.When selecting resolutions based on minimum-fuel-burn, a 40% reduction in fuel burn was realized as compared to the the conventional minimum-delay approach.However, the delay incurred with those more fuel-efficient resolution maneuvers was nearly twice that observed with the minimum-delay approach.The stark contrast suggests there could be value to an approach that considers the cost of both delay and fuel burn.The desire to balance the costs of delay and fuel burn is evident in today's Flight Management Systems (FMS).Most airlines use a ratio of the two costs to determine the economy speed for a given flight on a given day.This ratio is called the Cost Index, and it determines the "economy" speed profile for a flight by minimizing the total cost of operation.The Cost Index is the ratio of the time-related operating costs of the aircraft vs. the cost of fuel.This process can be applied when determining how best to resolve a conflict.Where previous studies have explored resolution selection based on minimizing either delay or fuel burn, the algorithm in this study was modified to minimize cost given a parameterized expression of their relative importance.This study examines the system performance of a conflict resolution algorithm capable of selecting maneuvers based on minimum cost in realistic traffic scenarios.The paper is organized as follows.Section II presents the cost function that considers the price of fuel and the"price" of airborne flight delay.The conflict resolution algorithm into which this cost function was embedded is described in Section III.Section IV sets forth the experiment design, and Section V presents the results.A summary of the conclusions and discussion of future work conclude the paper.
+II. Balancing Delay and Fuel BurnThis study evaluates the performance of modifying the resolution selection method to evaluate fuel and delay together.An approach to accomplish this is to design a cost function that normalizes delay and fuel burn.Since delay and fuel burn produce a cost in a true sense of the term, the monetary amount of resolutions is modeled within our system and used as the optimization criteria in our resolution selection scheme.Furthermore, a mechanism to vary the weight of delay and fuel burn in a cost function creates a separation assurance feature similar to the Cost Index of a FMS.This section describes an approach for quantifying the cost of a resolution as a function of delay and fuel burn, and selecting the least-cost resolution for a given conflict.Often, the cost of delaying a flight differs from flight to flight.This cost can be most accurately estimated by the airlines; unfortunately these models are not regularly available.In Ref. 5, the cost of different types of delay (airborne, ground, etc.) were estimated for an array of aircraft sizes.The average airborne delay price of $20.00 per minute, for passenger aircraft of 100 seats or more, was used as the nominal delay price for this study.Since most conflict resolutions produce delays of less than a minute, delay price translated to $0.33 per second.Present-day fuel price of approximately $0.43 per pound was used for resolution selection.The structure of the employed conflict resolution algorithm allows for the tabulation of both delay and fuel burn for each resolution considered.Using these metrics, a cost function descriptive of the relationship between the price of fuel and the amount of fuel and the price of delay and the amount of delay was derived.Equation (1) describes the operational cost:C O = (F B × P F B ) + (D × P D ) (1)where FB is the fuel burn in pounds, P F B is the fuel price in dollars per pound, D is the delay in seconds and P D is the price of delay.For the purpose of this study a resolution cost function, C R , was developed.The relative importance of delay to fuel burn within C R is represented by the inclusion of a user specified weight parameter: alpha.In this scenario, the user represents the Federal Aviation Administration (FAA).The availability of the alpha parameter could allow the FAA to balance system wide preferences for delay and fuel burn thus allowing national optimization of the air traffic control system.For example, the user could shift alpha to favor fuel burn savings when flights are not constrained by time (i.e. a flight is early and would otherwise be delayed because of traffic flow management or the gate is not ready) and shift to delay for aircraft that need to be scheduled more closely.The range of alpha is shown in Eq.( 2):0 ≤ α ≤ 1 (2)Using alpha to represent the weight of a given parameter to another, the resolution cost [Eq.(1)] can be expressed as Eq.(3):C R = [α(D × P D ) + ((1 -α) × (F B × P F B ))](3)where α=0 represents minimum fuel burn optimization and α=1 is minimum delay optimization.The calculation of fuel burn and delay is discussed in Section IV.E.For this study, resolution cost, C R , is used as the criteria in which the optimal resolution is selected to resolve a conflict.The results of this process is controlled by the user-specified alpha value based on the importance of fuel burn and delay.Since resolution cost is a theoretical term, in later sections, operational cost, C O , is analyzed to represent the actual price to the airspace users.
+III. Implementation
+A. Advanced Airspace Concept AutoresolverThe Advanced Airspace Concept Autoresolver (AAC Autoresolver) is a strategic conflict resolution algorithm designed to deconflict aircraft that are predicted to lose separation more than two minutes in the future.The Autoresolver resolves aircraft conflict pairs ordered by predicted time to first loss of separation.For each conflict in the conflict list, the Autoresolver follows an iterative approach for resolution.These trajectories take into account characteristics such as aircraft type, speed and airspace boundaries.The Autoresolver calculates future trajectories composed of waypoints, speeds and altitudes which may possibly resolve the conflict.Figure 1 shows the types of trajectory changes attempted by the Autoresolver grouped in terms of horizontal, vertical, or speed maneuvers.The dashed lines in Figure 1 indicate the suggested trajectory changes to avoid the predicted conflict.This trajectory change is then sent to a trajectory engine that computes a corresponding trial resolution trajectory.A resolution trajectory is considered viable, successful (and stored), if it resolves the primary conflict, and is free of predicted losses of separation with all aircraft for a specified period of time.If the trial resolution is not conflict free, the Autoresolver computes a new trial resolution and checks if it is successful.For each resolution type this iteration is continued until a successful resolution is found or all possibilities of that type have been tried.For each successful resolution, both the associated delay and the fuel burn are calculated.The Autoresolver will generate up to 18 successful resolutions per aircraft in conflict for a total of up to 36 between the two aircraft.In this study, the algorithm selected a resolution from among the set of successful resolutions by calculating the cost per resolution and selecting the resolution with the lowest cost.The selected resolution was then implemented via fast-time, closed-loop experiment as discussed in the following sections.Using the equation formulated in the previous section, the result of the AAC computations is a list of resolutions and their associated costs.Further discussion regarding the design of the algorithm and the types of resolutions that are generated is presented in Refs.6, 7.
+IV. Experiment DesignThis section describes the fast-time simulation environment, test parameters, and the metrics used in the study.
+A. Simulation EnvironmentThe Airspace Concept Evaluation System (ACES) is a fast-time, agent-based simulation of the National Airspace System (NAS) that uses four-degree-of-freedom (4 D.O.F) equations of motion based on the Base of Aircraft Data (BADA) to generate aircraft trajectories. 8ACES was developed specifically to provide a general purpose environment for evaluating future air traffic management and control concepts, including automated resolution algorithms.Each flight's trajectory is simulated from the departure fix associated with its original airport and ends at the arrival fix associated with its destination airport.By using aircrafttype-specific performance data together with guidance and navigation models, the ACES trajectory engine can generate representative trajectories for many aircraft.For the purposes of this study, the aircraft trajectories were entirely deterministic with no trajectory uncertainty.Aircraft conflicts were predicted with perfect accuracy, and resolution trajectories were guaranteed to be followed precisely by the simulated aircraft.In addition to deterministic aircraft trajectories, certain simplifications were made in the modeling and execution of the experiment: negotiation of resolution trajectories between aircraft operators and/or the air navigation service provider were not modeled, and neither data link transmission delays nor pilotaction delays were modeled.Once a resolution trajectory was selected by the automation it was executed immediately and precisely.
+B. Airspace and Traffic
+C. Test MatrixTable 1 shows the test matrix used in this study to investigate the benefits of selecting conflict resolution maneuvers based on minimum cost.The matrix includes two independent variables: alpha and price index.Nine test points were chosen for alpha evenly distributed between 0 and 1 at 1/8 increments.Three test points were chosen for price index: Nominal, Double the fuel price, and Double the cost of delay."Nominal"represents a fuel price and delay costs at current-day values.Double the Fuel Price and Double the Delay Price describe test points for which P F B or P D is doubled, respectively.
+D. Dependent VariablesThree metrics were selected for comparison: the number of conflicts per flight hour, delay and fuel burn.A conflict is said to occur when two aircraft are predicted to come within 5 nautical miles horizontally and 1,000 feet vertically from each other some time in the future (i.e.20 minutes).The flight hour metric is calculated by summing the total flying time within ZAU and ZID of every flight in the simulation.In the study, the number of conflicts per flight hour is used as a proxy for complexity.The delay metric is defined as the additional delay incurred per resolution, in seconds, as compared to the original (i.e., conflicted) trajectory.The fuel-burn metric is defined as the additional fuel burned per resolution, in pounds, as compared to the original trajectory.Fuel burn is modeled as a function of thrust, true airspeed, and altitude using BADA.
+E. Delay and Fuel Price ParametersThe cost of airborne delay used in this study was approximated from the values for airborne delay costs presented in Table 4
+V. ResultsThis study evaluates the effects of a cost-based resolution selection criterion on system efficiency.Metrics for complexity and cost are examined to quantify the impact of modifying the AAC Autoresolver.The cost metrics and resolution-type cost results are presented parametrically in terms of the delay cost and fuel burn cost of selected resolutions.The cost associated with various resolution types is investigated, and their implications are discussed.
+A. ComplexityIn order to assess how the resolution selection criterion (alpha) affects the complexity of the conflict resolution problem generally, the number of conflicts per flight hour was examined.The number of conflicts per flight hour provides insight into the algorithm's response to the inclusion of cost-based resolution selection.A significant increase in the number of conflicts per flight hour, as a result of including the minimum-cost resolution selection approach, might suggest an increase in problem complexity.Figure 3 shows that using cost-based resolution selection does not significantly increase the observed number of conflicts per flight hour.The small increase in the number of conflicts per flight hour observed between α=0 and α=1 may be a by-product of the resolution selection process.In Bowe et al. (Ref.4) minimum-delay resolution selection was found to favor timesaving maneuvers such as route shortcuts.When selecting resolutions based on minimum fuel burn the algorithm displayed a preference for speed reduction maneuvers which were shown to increase the cumulative delay.
+B. Costs Associated with Balancing Delay and Fuel BurnMost flight management systems in operation today have configurable cost index settings to select the most efficient speed profile according to the users' needs.This allows the user to weigh the importance of saving time or saving fuel per flight.For example, if a given flight is ahead of schedule, and connecting flights are in question, a user may change the cost index in the FMS to favor fuel savings.Likewise, if a flight is behind schedule the user could change the index to increase the importance of delay.Similar to this cost index paradigm for favoring delay or fuel burn in certain cases, the results of this experiment can be modeled in a way that allows one metric to be weighed more heavily than the other.Figure 4(a) shows the cost as a function of alpha.As expected, when α=0 the fuel burn cost is minimized, and when α=1 delay cost is minimized.When α=1 the contribution of the fuel cost to the operational cost equation is zero but each resolution still has an associated amount of fuel burn.Conversely, in the minimum fuel burn case, the total cost is dominated by the delay cost.The lowest total operational cost roughly occurs when α=0.5.The figure reveals the minimum delay case to be the most expensive overall, as reflected by the Total Operational Cost curve, which is primarily dominated by the fuel cost.This result suggests that optimizing conflict resolution maneuvers for minimum delay may be the least cost effective approach.Figure 4(b) shows the total operational cost as a function of alpha for the three price indices.The observed trends suggest that, when the cost of delay is doubled, the overall operational cost is higher than when the fuel price is doubled, with the exception of when alpha is between 0.875 and 1.As expected, the nominal price index produces the lowest total operating cost and the least dramatic fluctuation in cost over the range of alpha, until alpha is greater than 0.875.Of interest in Figure 4(b) is the increase in total operational cost when the price of delay is doubled and α=0.5.Further investigation revealed a large disparity in the total operational cost for en route and arrival conflicts for this simulation setting.An arrival conflict is defined when a maneuvered aircraft is predicted to conflict with another within 20 minutes of its arrival fix.All other instances of conflict are considered en route.Figure 5 shows the total operational cost when the price of delay is doubled for en route and arrival conflicts.The rise in operational cost is directly due to the spike in operational cost for arrival conflicts when α=0.5.Results show that when delay and fuel costs were evenly balanced in the Double the Delay Price simulation runs, a large amount of speed reductions were selected as the optimal cost-based resolution, thus creating arrival sequencing congestion where additional resolutions were required to separate the flow.More analyses are needed to evaluate the impact of optimizing cost when resolving arrival conflicts.The curve for total operational cost across alpha considering only en route conflicts is much smoother and almost symmetrical at α=0.5.This indicates that as the importance of delay and fuel costs becomes more unbalanced, operational cost increases, and this trend is generally the same regardless which parameter (delay or fuel burn) is favored.
+C. Resolution Type CostsFurther analysis illustrates the influence that the selection of resolution type has on aircraft operating costs.When resolving conflicts there are several categories of maneuvers that can be utilized to prevent a predicted conflict.The AAC Autoresolver captures most of the different resolution types used in the field, and these types are illustrated in Figure 1.For most resolution types their impact on the performance of delay and fuel burn can be generally hypothesized through intuition into the physics of the maneuvers.For instance, maneuvers such as a Direct-to which identify wind favorable shortcuts along the planned route are known to save time and fuel, step altitude climbs generally reduce fuel burn, step altitude descents generally increase fuel burn, path stretches are known to increase delay and fuel burn, and speed reductions save fuel, but increase delay.The performance of delay and fuel burn does not have a direct correlation based on resolution type, however the operational cost of resolutions creates a normalization of the two metrics and comparisons can be made.In this section, the operational cost among 13 different resolution types are investigated in an attempt to uncover which resolutions theoretically cost more than others.Figure 6(a) provides the average operational cost per resolution executed for each resolution type.The data was computed using simulation results of α=0.5, in order to evenly balance the cost of delay and fuel burn, and the nominal price index.As expected, the two best resolutions with respect to operational cost are Direct-to and Variable Speed Direct-to ($36.70 and $55.70 savings, respectively).Both maneuvers result in a shorter horizontal path thus saving fuel and delay, and Variable-Speed Direct-to (D2Speed ) initiates a speed reduction simultaneously with a Direct-to for added benefit.The most expensive resolution types are the horizontal maneuvers: path stretch, offset, and horizontal vector turn (HVT ).Path stretches produce the overall highest price with $48.40 per resolution.Interestingly, seven of the 13 resolution types produce a mean price savings (negative cost), however it should be noted this does not translate to a total price savings as the magnitude of horizontal maneuvers and speed reductions overtake any savings for a incurred cost.In the speed domain, increases tend to save $8.30 on average and reductions incur a price of $3.20 on average.Both step altitude descents and climbs produce a mean price savings with descents saving approximately four times more than climbs.Furthermore, temporary altitude descents incur a mean price of $3.20, by contract temporary altitude climbs save $7.40.A maneuver not illustrated in Figure 6 called an extended temporary altitude (ExtTempAlt), is a maneuver where in order to resolve a conflict an aircraft, already performing a temporary altitude, remains at the current temporary altitude for a specified period of time (i.e., 12 minutes).These maneuvers produce a mean price savings of $3.40.The selection rate of each resolution type is important when making comparisons, especially when a broad distribution exists as shown in Figure 6(b).In our simulations, path stretches were the single most utilized maneuver when resolving conflicts.These maneuvers are the most frequently selected (approximately 29% of the time), because they are extremely successful at creating the separation minima required to clear a conflict.Consequently, they are also the most expensive resolution, and the dominant factor in the overall price of resolving conflicts.By contrast, the Direct-to maneuver significantly reduces operational cost when selected, however is only utilized 3.1% of the time.Regardless of the optimization problem or advancements to conflict resolution algorithms, the performance of resolutions will never be more important than the criticality of avoiding actual losses of separation.It is likely there will always be a net price to resolving conflicts.In order to uncover which cost, delay or fuel burn, has a greater effect on the results in Figure 6(a), the percentage that derives from the price of fuel is shown in Figure 6(c).For example, the value of 57% for the path stretches means 57% of the $48.40 per resolution comes from the price of fuel, thus 43% of the price derives from additional delay.Eight of the 13 resolution types produce mean costs mostly made up by the price of fuel burn, i.e. fuel burn price percentages greater than 50% Most of the price savings for Variable-Speed Direct-to maneuvers come from fuel burn (98%).This is expected because the maneuver employs a speed reduction to produce zero delay when performing a Direct-to, thus no delay change to benefit from.The price of path stretches and Direct-to maneuvers are nearly impacted evenly between delay and fuel burn.Moreover, temporary altitude descents and speed increases were primarily impacted by their respective delay price (savings).
+VI. ConclusionsFast-time simulations of current-day traffic levels in two regional airspaces under nominal weather conditions were simulated to evaluate the benefit of modifying a conflict resolution algorithm to select resolution maneuvers based on minimum cost.The study employed the use of a parameter, alpha, to represent the relative importance of fuel price to the cost of delay, similar to FMS cost index.In terms of operational cost, the most efficient choice of alpha is roughly 0.5, i.e., when the cost of fuel and delay are weighted evenly.The operational cost is highest when the cost of fuel is ignored by the algorithm (i.e., α=1), which is the case for most conflict resolution algorithms in the literature, including the baseline prototype (AAC Autoresolver) that was modified for this study.The most cost-effective resolution maneuvers were the Direct-to and the Variable-Speed Direct-to, owing to the fact that both maneuvers result in a shorter horizontal path, thus saving time and fuel.Conversely, the most costly maneuver type was the Path Stretch.The cost of fuel burn is the predominant factor in the total operational cost of most (eight) of the 13 resolution maneuver types.
+VII. Future WorkAn approach for evaluating cost based conflict resolution was presented.The context for the evaluation was an approximation of delay and fuel cost.However, the price of delay varies based on the number of passengers per aircraft.In the future, an investigation of how different aircraft fleet mixes (low, medium, and high-occupancy aircraft) affect the results of selecting resolutions based on minimum cost could be performed.Furthermore, other ARTCCs could be simulated to test the results of different conflict geometries (i.e different resolution type distribution).Each simulation in this study used the same value of alpha for every, conflict therefore, each conflict had the same defined importance of delay vs. fuel burn.Further research into calculating an optimal alpha value for each conflict based on characteristics of the flight pair could improve the operational practicality of approach.For example, a flight approaching its arrival fix may tend to prefer to minimize delay (higher alpha value), especially if it is already behind schedule and there are no apparent arrival sequencing issues.By contrast, a flight that is early may tend to prefer a fuel-efficient resolution (lower alpha value), because any further timesaving may be nullified by additional delays in the terminal.Figure 1 :1Figure 1: Resolution trajectories of type horizontal (a), vertical (b) and speed (c & d)
+In this study, the Autoresolver resolved roughly 1,885 conflicts in two Air Route Traffic Control Centers (ARTCCs): Indianapolis (ZID) and Chicago (ZAU).The ARTCCs selected contain primarily cruising traffic which is of interest as this study focused on the resolution of en route conflicts.Flight operations over a 6-hour period were simulated based on Aircraft Situation Display to Industry (ASDI) data recorded March 8, 2007 which represented a"low weather," high volume day in the NAS.The data set included 23,000 flights of varying types, their associated routes, and their departure times.The Rapid Update Cycle wind data was used to model winds in the selected ARTCCs.Figure2shows a subset of the ARTCCs in the central region of the United States with ZID and ZAU shaded.
+Figure 2 :2Figure 2: The ARTCCs studied in this experiment.
+Figure 3 :3Figure 3: Conflicts per Flight Hour vs alpha.
+Figure 4 :4Figure 4: (a) Total Cost vs. alpha, (b) Operational Cost vs. alpha
+Figure 5 :5Figure 5: Total Operational Cost vs. alpha for En Route and Arrival Conflicts for Double Delay Price Index.
+Figure 6 :6Figure 6: Comparison of results by distinct resolution types for α=0.5 and nominal price.(a) mean operational cost per resolution type, (b) percentage of all conflicts each resolution type was selected, (c) fuel burn price ratio.
+Table 1 .1Experiment Factors and Levels.Experiment FactorsLevelsOptimizationResolution Costα0, 0.125, 0.25, 0.375, 0.5, 0.625, 0.75, 0.875, 1Price IndexNominal, Double the Fuel Price, Double the Delay Price
+Table 2 .2Delay and fuel burn prices for the various Price Index levels.Price IndexDelay Price ($/seconds) Fuel Price ($/pound)Nominal0.330.43Double the Fuel Price0.330.86Double the Delay Price0.660.43of Ref.5for passenger aircraft greater than 100 persons.The operational data used to generate the cost figures in the referenced study were collected from European Airlines.A summary of the cost of delay to airlines during various trip segments is presented in Ref. 9. The fuel price used in this study was taken from the International Air Transport Association (IATA) Jet Fuel Price Monitor.10Thisstudyused a price from August 2012 to represent the Nominal Cost Index.The delay and fuel prices for the different Price Index levels used in this study is presented in Table2.
+			of 10 American Institute of Aeronautics and Astronautics Downloaded by NASA AMES RESEARCH CENTRE on April 17, 2013 | http://arc.aiaa.org| DOI: 10.2514/6.2012-5416
+			of 10 American Institute of Aeronautics and Astronautics Downloaded by NASA AMES RESEARCH CENTRE on April 17, 2013 | http://arc.aiaa.org| DOI: 10.2514/6.2012-5416
+			of 10 American Institute of Aeronautics and Astronautics Downloaded by NASA AMES RESEARCH CENTRE on April 17, 2013 | http://arc.aiaa.org| DOI: 10.2514/6.2012-5416
+			of 10 American Institute of Aeronautics and Astronautics Downloaded by NASA AMES RESEARCH CENTRE on April 17, 2013 | http://arc.aiaa.org| DOI: 10.2514/6.2012-5416
+			of 10 American Institute of Aeronautics and Astronautics Downloaded by NASA AMES RESEARCH CENTRE on April 17, 2013 | http://arc.aiaa.org| DOI: 10.2514/6.2012-5416
+			of 10 American Institute of Aeronautics and Astronautics Downloaded by NASA AMES RESEARCH CENTRE on April 17, 2013 | http://arc.aiaa.org| DOI: 10.2514/6.2012-5416
+			of 10 American Institute of Aeronautics and Astronautics Downloaded by NASA AMES RESEARCH CENTRE on April 17, 2013 | http://arc.aiaa.org| DOI: 10.2514/6.2012-5416
+		
+		
+			
+			
+			
+
+				
+
+
+	
+		Terminal Area Forecast 1977-1987. Aviation Forecast Branch, Office of Aviation Policy, Federal Aviation Administration, Department of Transportation, Washington, D.C. 20591. February 1976. Various paging
+		10.1177/004728757701500317
+	
+	
+		Journal of Travel Research
+		Journal of Travel Research
+		0047-2875
+		1552-6763
+		
+			15
+			3
+			
+			2010
+			SAGE Publications
+		
+	
+	Tech. Rep. HQ111308
+	"Terminal Area Forecast Summary Fiscal Years 2010-2030," Tech. Rep. HQ111308, Federal Aviation Administration, 2010.
+
+
+
+	
+		A review of conflict detection and resolution modeling methods
+		
+			JKKuchar
+		
+		
+			LCYang
+		
+		10.1109/6979.898217
+	
+	
+		IEEE Transactions on Intelligent Transportation Systems
+		IEEE Trans. Intell. Transport. Syst.
+		1524-9050
+		
+			1
+			4
+			
+			2000
+			Institute of Electrical and Electronics Engineers (IEEE)
+		
+	
+	Kuchar, J. K. and Yang, L. C., "A Review of Conflict Detection and Resolution Modeling Methods," IEEE Transactions on Intelligent Transportation Systems, Vol. 1, No. 4, 2000, pp. 179-189.
+
+
+
+	
+		Foundations of mediation training: A literature review of adult education and training design
+		
+			TimothyHedeen
+		
+		
+			SusanSRaines
+		
+		
+			AnsleyBBarton
+		
+		10.1002/crq.20018
+	
+	
+		Conflict Resolution Quarterly
+		Conflict Resolution Quarterly
+		1536-5581
+		
+			28
+			2
+			
+			2010
+			Wiley
+		
+	
+	Tech. Rep. 2010
+	"Literature Review of Conflict Resolution Research," Tech. Rep. 2010, Federal Aviation Administration, 2010.
+
+
+
+	
+		Selecting conflict resolution maneuvers based on minimum fuel burn
+		
+			AishaBowe
+		
+		
+			ToddLauderdale
+		
+		10.1109/dasc.2010.5655529
+	
+	
+		29th Digital Avionics Systems Conference
+		
+			IEEE
+			2010
+		
+	
+	Bowe, A. and Lauderdale, T., "Selecting conflict resolution maneuvers based on minimum fuel burn," Digital Avionics Systems Conference, 2010.
+
+
+
+	
+		
+			AKara
+		
+		
+			JFerguson
+		
+		Estimating Domestic U.S Airline Cost of Delay based on European Model," 4th International Conference on Research in Air Transportation
+		
+			2010
+		
+	
+	Kara, A., Ferguson, J., and et. al,"Estimating Domestic U.S Airline Cost of Delay based on European Model," 4th International Conference on Research in Air Transportation, 2010.
+
+
+
+	
+		Automated conflict resolution, arrival management, and weather avoidance for air traffic management
+		
+			HErzberger
+		
+		
+			TALauderdale
+		
+		
+			Y-CChu
+		
+		10.1177/0954410011417347
+	
+	
+		Proceedings of the Institution of Mechanical Engineers, Part G: Journal of Aerospace Engineering
+		Proceedings of the Institution of Mechanical Engineers, Part G: Journal of Aerospace Engineering
+		0954-4100
+		2041-3025
+		
+			226
+			8
+			
+			2010
+			SAGE Publications
+			Nice, France
+		
+	
+	Erzberger, H., Lauderdale, T. A., and Cheng, Y., "Automated Conflict Resolution, Arrival Management and Weather Avoidance for ATM," 27th Iternational Congress of the Aeronautical Sciences, Nice, France, 2010. 9 of 10
+
+
+
+	
+		An Approach for Balancing Delay and Fuel Burn in Separation Assurance Automation
+		
+			AishaBowe
+		
+		
+			ConfesorSantiago
+		
+		10.2514/6.2012-5416
+		
+	
+	
+		12th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference and 14th AIAA/ISSMO Multidisciplinary Analysis and Optimization Conference
+		
+			American Institute of Aeronautics and Astronautics
+			2013
+		
+	
+	American Institute of Aeronautics and Astronautics Downloaded by NASA AMES RESEARCH CENTRE on April 17, 2013 | http://arc.aiaa.org | DOI: 10.2514/6.2012-5416
+
+
+
+	
+		Automated Conflict Resolution for Air Traffic Control
+		
+			HErzberger
+		
+		
+			2006
+		
+	
+	25th Iternational Congress of the Aeronautical Sciences
+	Erzberger, H., "Automated Conflict Resolution for Air Traffic Control," 25th Iternational Congress of the Aeronautical Sciences, 2006.
+
+
+
+	
+		Build 8 of the Airspace Concept Evaluation System
+		
+			SapaGeorge
+		
+		
+			GoutamSatapathy
+		
+		
+			VikramManikonda
+		
+		
+			KeePalopo
+		
+		
+			LarryMeyn
+		
+		
+			ToddLauderdale
+		
+		
+			MichaelDowns
+		
+		
+			MohamadRefai
+		
+		
+			RichardDupee
+		
+		10.2514/6.2011-6373
+	
+	
+		AIAA Modeling and Simulation Technologies Conference
+		Portland, OR
+		
+			American Institute of Aeronautics and Astronautics
+			2011
+		
+	
+	George, S., Satapathy, G., Manikonda, V., Palopo, K., Meyn, L., Lauderdale, T. A., Downs, M., Refai, M., and Dupee, R., "Build 8 of the Airspace Concept Evaluation System," AIAA Modeling and Simulation Technologies Conference, Portland, OR, 2011.
+
+
+
+	
+		Evaluating the True Cost to Airlines of One Minute of Airborne or Ground Delay
+		
+			PRUnit
+		
+	
+	
+		Tech. rep., University of Westminster Final Report
+		
+			May 2004. August 2012
+		
+	
+	Unit, P. R., "Evaluating the True Cost to Airlines of One Minute of Airborne or Ground Delay," Tech. rep., University of Westminster Final Report, May 2004. 10 International Air Transport Association Jet Fuel Price Monitor, August 2012.
+
+
+				
+			
+		
+	

file104.txt

@@ -0,0 +1,310 @@
+
+	
+	
+		
+IntroductionAir traffic demand is projected to increase significantly in the upcoming years [1].The human workload associated with conflict detection and resolution is expected to limit this increase and thereby limit the economic growth that aviation facilitates.Automated separation assurance systems are proposed as a way to safely and efficiently separate aircraft in highly dense traffic situations up to two to three times current levels, thereby fostering increased economic growth for the nation.Numerous algorithms have been proposed to provide separation assurance in the future air traffic system [2,3].Maintaining safe separation is the first-order objective of all such algorithms; the second-order objectives vary, however.The majority of the proposed algorithms optimize the selection of conflict resolution maneuvers to minimize airborne delay in order to mitigate the effect on schedule.An alternative objective is to optimize based on fuel burn.In [4] the system performance of a conflict resolution algorithm that selected resolutions based on minimum delay was compared to the system performance of the same algorithm when selecting resolutions based on minimum fuel burn.The most effective resolution maneuver when optimizing for airborne delay was a "direct-to" maneuver, which identifies wind-favorable shortcuts along an aircraft's planned route that reduce its flying time while resolving the predicted conflict [5].The most effective resolution maneuver when optimizing for fuel burn was a "speed reduction" maneuver, which employs a temporary speed reduction to resolve the predicted conflict.However, speed reductions were selected less frequently than other, less fuel-efficient maneuvers.Additionally, when utilized, these maneuvers significantly increase the cumulative delay.It is hypothesized that the availability of a compound maneuver combining a Direct-To maneuver with the fuel efficiency of a speed reduction would improve the performance of the separation assurance algorithm.This study compares the system performance of a conflict resolution algorithm in realistic traffic scenarios with and without the availability of a compound Direct-to/speedreduction maneuver, hereafter referred to as a Zero-Delay Direct-To maneuver.The objective is to quantify the operational benefit of adding the proposed maneuver to the set of maneuvers already available to the automated separation assurance algorithm.This paper is organized as follows.Section 2 describes the conflict resolution algorithm under test and the new compound maneuver.Section 3 presents the experimental approach, procedure, and assumptions.The results are then categorized according to safety and efficiency.Lastly, a summary of the study's findings is given, along with suggestions for future research.
+Test ArticleThe conflict resolution algorithm evaluated in this study is the Advanced Airspace Concept (AAC) Autoresolver [6,7].It is a ground-based algorithm that resolves conflicts in pairwise fashion and can be optimized for airborne delay or for fuel burn.The Autoresolver selects a maneuver from one of the following categories: horizontal, vertical, altitude, Direct-To or compound.For this study, only conflicts where en route flights were maneuvered are analyzed; arrivals were not included because they adhere to additional constraints such as metering.In the following study only one compound maneuver is enabled: the Zero-Delay Direct-To maneuver.
+Zero-Delay Direct-To ManeuverThe Autoresolver was modified to allow for the combination of a Direct-To maneuver with a reduction in speed.This compound maneuver is referred to as a Zero-Delay Direct-To maneuver.This modification augmented the existing Direct-To maneuver, thus allowing the algorithm to continue to have the option to utilize a Direct-To maneuver when efficient.The equation that describes a Direct-To maneuver is shown in (1) where d represents delay in hours, D 1 is the previous distance along the route in nautical miles, D 2 is the new distance in nautical miles and S is speed in knots:! d = D 1 S " D 2 S (1)Augmenting the above equation to produce a maneuver that results in zero delay requires setting d to zero.This yields equation 2 where S new represents the new (slower) speed in order to result in a Zero-Delay Direct-To maneuver.The algorithm abides by the original Direct-To constraints where the maneuver will not be considered if the aircraft ( 1) is less than 20 minutes from the arrival fix, (2) cannot return to the route within 50 nmi of the final fix, (3) path along the Direct-To route is greater than 250 nmi (dotted line in Figure 1), and (4) if the point where the aircraft rejoins the trajectory is within 50 nmi of the current Air Route Traffic Control Center boundary.In addition, it will not attempt to execute the maneuver if S new is within 5 knots of the original speed.!S new = D 2 D 1 " # $ % & ' S (2)For example, a Direct-To maneuver by an aircraft traveling 450 knots that will reduce the distance along the route from 400 to 360 nmi would reduce the speed to 405 (by 45) knots in order to produce no delay.When performing a Zero-Delay Direct-To maneuver, the aircraft would recapture the route at the same time it would had it not performed the maneuver.Figure 1 illustrates the Zero-Delay Direct-To maneuver where A 1 and A 2 are aircraft predicted to conflict.To avoid this, A 1 is selected to execute a Zero-Delay Direct-To maneuver.The new trajectory for A 1 (dashed line) removes several waypoints and reduces the speed as shown in the neighboring profile.The Mach number of A 1 decreases for the duration of the maneuver and eventually returns to its original speed after clearing the conflict.
+Resolution Selection CriteriaThe resolver will generate up to 18 successful resolutions per aircraft in conflict for a total of 36 available between the two aircraft.More specifically, the algorithm selects a successful resolution in each of the following categories for each aircraft:• Vector Left • Vector Right • Climb • Descend • Speed Increase • Speed Decrease • Direct-To • Zero-Delay Direct-To • Left Horizontal Vector Turn • Right Horizontal Vector TurnIn this study, the algorithm selects a resolution from among the set of successful resolutions using either the minimum delay or the minimum fuel burn criterion, depending on how the algorithm is configured.The selected resolution is then implemented via fast-time, closed-loop experiment as discussed next.Further discussion regarding the design of the algorithm and the types of resolutions that are generated is presented in [7].
+Experiment
+Test BedThe Airspace Concept Evaluation System (ACES) was used to simulate the National Airspace System (NAS) in a fast-time simulation [8].ACES was also used to compute and archive the dependent variables: the number of losses of separation and the airborne delay and fuel burn incurred flying the conflict resolution trajectories.
+ProcedureTo evaluate the difference between the current state-ofthe-art conflict resolution algorithm and the addition of a Zero-Delay Direct-To maneuver, a test plan was developed that examines the behavior of the algorithm with and without this maneuver enabled in two pairs of Air Route Traffic Control Centers (ARTCCs) under two conflict resolution optimization schemes.In [9], statistical clustering analysis was employed to categorize ARTCCs based on normalized conflict properties.The two ARTCC pairs selected for this experiment-Indianapolis (ZID) -Chicago (ZAU) and Los Angeles (ZLA) -Oakland (ZOA) -were chosen because they provide a wide representation of conflict properties.Table 1 shows the independent variables and settings.
+Independent Variables SettingsZero-Delay Direct-To maneuver Available, Unavailable Optimization Delay, Fuel Burn Center ZID-ZAU, ZLA-ZOA Table 1.Independent Variables
+Demand SetFlight operations over a 24-hour period were simulated based on Aircraft Situation Display to Industry (ASDI) data recorded March 8, 2007.ASDI data comes from the FAA's Enhanced Traffic Management System (ETMS) and contains information about flights controlled by air traffic control.The data set included 62,970 flights, their associated routes, and their departure times.This dataset had mixed aircraft types representing the current fleet mix.The data used in this study represents reasonable daily traffic in the NAS with a relatively small amount of weather induced delay.The Rapid Update Cycle wind data was used to model winds in the selected ARTCCs.
+Dependent VariablesThe dependent variables for the experiment were the number of losses of separation and the airborne delay and fuel burn incurred by flying the conflict resolution trajectories.In the development of a robust, efficient algorithm for implementation in the Next Generation Air Transportation System (NextGen), safety is of the utmost concern.The number of losses of separation is the metric used here to reflect the safety of the system.Those results are presented in Section 4.1.Efficiency in terms of delay and fuel burn are important once safety is assured.To calculate the delay imposed by executing a resolution maneuver, the time on the original trajectory at a common point is subtracted from the time on the resolution trajectory at the common point.Similarly, to estimate the fuel burn associated with a resolution maneuver, the weight of the aircraft at the common point for the resolution trajectory is subtracted from the aircraft weight for the original trajectory.The fuel consumed per resolution is computed by ACES using aircraft-specific coefficients selected from the Base of Aircraft Data (BADA) [10].The BADA is comprised of the performance and operating procedure coefficients of 295 aircraft types.These coefficients encompass those that are used to calculate thrust, drag, and fuel flow along with those used to specify nominal cruise, climb and descent speeds.Further discussion of the specific equations used to calculate the fuel burn is included in [4].The efficiencyrelated results are presented in Section 4.2.
+ResultsThis experiment seeks to evaluate the benefit of augmenting the Autoresolver to consider a Zero-Delay Direct-To maneuver when resolving a given conflict.The subsequent results address the safety and efficiency of potential implementation.
+SafetyThe primary safety metric for the experiment is the number of losses of separation.A loss of separation occurs when aircraft are less than 5 nmi horizontally and 1,000 feet vertically from each other in en route airspace.As expected, the addition of the Zero-Delay Direct-To maneuver did not adversely affect the safety of the system, as measured by losses of separation.Evaluating the number of conflicts per simulation provides insight into the impact of the modifications made to the algorithm.A significant increase in the number of conflicts as a result of the availability of the Zero-Delay Direct-To maneuver might suggest increased risk.Figure 2 shows that enabling the compound maneuver does not significantly increase the number of conflicts in either ARTCC.On average, the percent difference between the baseline number of conflicts and the Zero-Delay Direct-To enabled scenario is less than 1%.This suggests that the inclusion of this maneuver does not adversely affect the ability of the algorithm to resolve conflicts, and there are no major gaps in its implementation.
+EfficiencyThe following section uses fuel burn and delay as a metric to evaluate the efficiency of the addition of the Zero-Delay Direct-To maneuver to the base algorithm.This is different than discussing delay and fuel burn as an optimizing factor because for each resolution implemented these metrics are computed.Although the algorithm is selecting resolutions based on fuel burn or delay, both the delay and fuel burn values per maneuver were tabulated, thus allowing for comparison.
+En Route Delay MetricFlight arrival delay is defined as the difference in time between the arrival time of an aircraft as given in the flight schedule and its real arrival time.Although there are many sources of delay (e.g., air traffic control, weather, maintenance, crew availability), in the following analysis the source of delay is attributed to time difference between the aircraft's modified trajectory and original trajectory at a common point in the en route airspace.Positive delay occurs when the modified trajectory incurs additional flight time to avoid a loss of separation, much akin to a detour.Negative delay is a reduction in flight time that can occur when a dogleg in the flight is eliminated or more favorable winds are encountered.The inclusion of Zero-Delay Direct-To maneuver has almost no impact on delay when selecting resolutions based on delay.Figure 3 shows the average delay per resolution.When selecting resolutions based on minimum fuel burn the average delay per resolution with the Zero-Delay Direct-To maneuver enabled for ZID-ZAU is 10.86 seconds.For ZLA-ZOA under the same conditions the average delay is 16.84 seconds.This translates to a 20.29% increase in cumulative delay in ZID-ZAU, but the absolute difference is only 4.7 minutes.Likewise, in ZLA-ZOA the difference is less than 1% when selecting resolutions based on delay.This supports the initial assertion that cumulative delay would only marginally increase with the availability of the Zero-Delay Direct-To maneuver when resolution trajectories are optimized for airborne delay.Selecting resolutions based on minimum fuel burn with the Zero-Delay Direct-To maneuver available increased the cumulative delay by 45.12% in ZID-ZAU and 10.49% in ZLA-ZOA.The increase in cumulative delay is caused by the selection of fewer Direct-To's due to the fact that Zero-Delay Direct-To maneuvers are more optimal than Direct-To's when optimizing for minimum fuel burn.There is a greater negative effect in ZID-ZAU than ZLA-ZOA because the traffic in ZID-ZAU had a greater number of Direct-To's that were no longer implemented.This finding will be further discussed in the next section.Selecting resolutions based on minimum fuel burn appears to result in an increase in cumulative delay in each center.This effect can be attributed to secondary conflicts.Each implementation of Zero-Delay Direct-To changes the way the primary and consequently, secondary conflicts are solved.Because only one aircraft of the pair will be maneuvered to avoid a conflict, the average delay per resolution can be thought of as per aircraft.Generally, increasing the delay is considered to be undesirable.However, the magnitude of additional delay per resolution is small when compared to the 15-minute Federal Aviation Administration (FAA) definition of a "reportable" delay [11].Even if marginal, the system-wide effects of an increase in delay are difficult to determine.
+Fuel Burn MetricWhen a Zero-Delay Direct-To maneuver is executed the maneuvered aircraft is slowed by a specified amount.Figure 4 shows the distribution of speed reduction magnitudes for ZID-ZAU.75% of all speed reductions observed in the experiment were less than 30 knots.A typical Boeing 747-400 aircraft at 35,000 feet will cruise between Mach 0.8 (533 knots) and Mach .85(566 knots) approximately a 30-knot variation.This indicates that the majority of the speed reduction values required to obtain the desired fuel benefit are reasonable.Within our simulation the range observed adhered to aircraft performance limitations.To evaluate the fuel burn associated with a resolution maneuver, the weight of the aircraft at a common point on the resolution trajectory is subtracted from the aircraft weight for the original trajectory.The utilization of Zero-Delay Direct-To maneuvers increases the fuel burn savings by 91.85% in ZID-ZAU and by 47.48% in ZLA-ZOA when resolving conflicts.Figure 5 shows the average fuel burn per resolution for ZID-ZAU and ZLA-ZOA.The negative fuel burn seen in ZID-ZAU is an indication that the modification made to the algorithm causes it to outperform nominal case when selecting resolutions based on minimum fuel burn.The average fuel burn per resolution in ZID-ZAU is 4.01 pounds less than when selecting resolutions based on minimum fuel burn with Zero-Delay Direct-To maneuvers enabled.In ZLA-ZOA the average fuel burn per resolution is 2.73 pounds, this is 2.41 pounds less than when Zero-Delay Direct-To is disabled.Though these numbers may seem insignificant the potential fuel benefit is great when considering the savings per year.In this study there were 3,276 conflicts in ZID-ZAU over the course of the day.Each of these requires one of the two aircraft to be maneuvered.Considering the average fuel savings of 4 pounds per resolution in ZID-ZAU, this amounts to roughly 4.8 million pounds of fuel per year.This is enough fuel to fill the tank of a Boeing 737-700 approximately 100 times.Furthermore, there are 20 ARTCCs within continental United States that could benefit from these savings.Variation in traffic density and route length accounts for most of the difference in the magnitude of savings between the two centers.ZID-ZAU center executed nearly twice as many resolution maneuvers as ZLA-ZOA, suggesting that the fuel efficiency of the algorithm increases with the air traffic demand.However, the improvement seen in the delay cases is not as significant.When selecting resolutions based on delay the algorithm finds Direct-To maneuvers to be more efficient.This can be attributed to the fact that the selection of a Direct-To can result in negative delay and thus a time saving whereas the most time-efficient Zero Delay solution is zero and will not yield a time savings.Figure 6 shows the resolutions selected by the algorithm for ZID-ZAU for fuel burn optimization with Zero-Delay Direct-To maneuvers enabled and disabled.Overall, the number of resolutions other than Direct-To or Zero-Delay Direct-To remains consistent between scenarios.When Zero-Delay Direct-To maneuvers were disabled there were 306 Direct-To's executed.When enabled there were 181 Direct-To's and 147 Zero-Delay Direct-To's.This represents a 41% decrease in the number of Direct-To maneuvers.When optimizing for minimum fuel burn, the algorithm frequently selects Zero-Delay Direct-To maneuvers over traditional Direct-To maneuver.However, in a small number of cases, a Direct-To maneuver is selected despite the fact that a Zero-Delay Direct-To maneuver is available.In these instances, the additional fuel savings does not outweigh a decrease in flight time.
+ConclusionEight conditions were simulated to evaluate the benefit of modifying the AAC Autoresolver to consider a Zero-Delay Direct-To maneuver when resolving a given conflict.Two methods of resolution selection were used: minimum delay and minimum fuel burn.The experiment was conducted in a fast-time environment using data representing a reasonable traffic day in the NAS.The results showed that augmenting the existing algorithm to include the compound maneuver did not significantly influence the algorithm's ability to resolve conflicts or effect the number of losses of separation observed.The inclusion of Zero-Delay Direct-To increased the cumulative fuel burn savings by 91.85% in ZID-ZAU and 47.48% in ZLA-ZOA when selecting resolutions based on minimum fuel burn.In this scenario, the average penalty in delay per aircraft is on the order of seconds.Further analysis is required to determine the effect of increasing the delay as well as the balance between delay and fuel burn benefit.The cumulative fuel burn savings observed within this study suggests that the Zero-Delay Direct-To maneuver could provide significant fuel savings in future systems while maintaining safety and schedule.
+Future WorkIn en route airspace, aircraft operate within desired performance envelopes and operational speed limitations.To address these factors a survey concerned with evaluating the effects of distinct performance envelopes on the feasibility of the Zero-Delay Direct-To maneuver is planned.In addition, the operational soundness of the speed reduction distribution requires validation by subject matter experts.Furthermore, the prior work introduced the addition of a Zero-Delay Direct-To maneuver within the Advanced Airspace Concept Autoresolver.The experiment then looked at the performance of the Zero-Delay Direct-To maneuver when selecting resolutions for either minimum fuel burn or minimum delay.This leaves a gap in coverage for a follow-on simulation to explore a hybrid selection scheme where resolution selection is based on a tradeoff between the two cost functions.Figure 1 .1Figure 1.Zero-Delay Direct-To
+Figure 2 .2Figure 2. Number of Conflicts
+Figure 3 .3Figure 3. Average Delay
+Figure 4 .4Figure 4. Zero-Delay Direct-To Enabled Speed Reduction ZID-ZAU Fuel Burn Optimal
+Figure 5 .Figure 6 .56Figure 5. Average Fuel Burn
+		
+		
+
+			
+AcknowledgmentThe authors wish to acknowledge Dr. Todd Lauderdale whose contributions to this work were invaluable.The authors also thank Todd Farley, and Drs.Antony Evans, and Banavar Sridhar for their insightful suggestions and thoughtful review.
+			
+
+			
+
+				
+
+
+	
+		FAA Aviation Forecasts: Fiscal Years 1981-1992. Federal Aviation Administration, U.S. Department of Transportation, Washington, D.C. 20591. 1980. 69p
+		10.1177/004728758102000159
+	
+	
+		Journal of Travel Research
+		Journal of Travel Research
+		0047-2875
+		1552-6763
+		
+			20
+			1
+			
+			2010
+			SAGE Publications
+		
+	
+	Federal Aviation Administration
+	Federal Aviation Administration, "Terminal Area Forecast Summary, Fiscal Years 2010-2030", FAA HQ111308, 2010.
+
+
+
+	
+		A review of conflict detection and resolution modeling methods
+		
+			JKKuchar
+		
+		
+			LCYang
+		
+		10.1109/6979.898217
+	
+	
+		IEEE Transactions on Intelligent Transportation Systems
+		IEEE Trans. Intell. Transport. Syst.
+		1524-9050
+		
+			1
+			4
+			
+			2000
+			Institute of Electrical and Electronics Engineers (IEEE)
+		
+	
+	Kuchar, J.K., Yang, L.C., " A Review of Conflict Detection and Resolution Modeling Methods", IEEE Transactions on Intelligent Transportation Systems, Vol.1, No. 4, pg. 179-189, 2000.
+
+
+
+	
+		TRB Special Report 301: Traffic Controller Staffing in the En Route Domain
+		10.17226/13022
+	
+	
+		Literature Review of Conflict Resolution Research
+		
+			Transportation Research Board
+			2010
+		
+	
+	Federal Aviation Administration Task Order white Paper, "Literature Review of Conflict Resolution Research" 2010.
+
+
+
+	
+		Selecting conflict resolution maneuvers based on minimum fuel burn
+		
+			AishaBowe
+		
+		
+			ToddLauderdale
+		
+		10.1109/dasc.2010.5655529
+	
+	
+		29th Digital Avionics Systems Conference
+		
+			IEEE
+			Oct. 2010
+			4
+			
+		
+	
+	Bowe, A., Lauderdale, T., "Selecting conflict resolution maneuvers based on minimum fuel burn," Digital Avionics Systems Conference (DASC), 2010 IEEE/AIAA 29th , vol., no., pp.1.A.4-1-1.A.4-9, 3-7 Oct. 2010.
+
+
+
+	
+		Direct-To Tool For En Route Controllers
+		
+			HeinzErzberger
+		
+		
+			DavidMcnally
+		
+		
+			MichelleFoster
+		
+		
+			DannyChiu
+		
+		
+			PhilippeStassart
+		
+		10.1007/978-3-662-04632-6_11
+	
+	
+		New Concepts and Methods in Air Traffic Management
+		Capri, Italy
+		
+			Springer Berlin Heidelberg
+			Sep. 1999
+			
+		
+	
+	Erzberger, H., McNally, B. D., Foster, M., Chiu, D., and Stassart, P., "Direct-To Tool for En Route Controllers," ATM '99: IEEE Workshop on Advanced Technologies and their Impact on Air Traffic Management in the 21st Century, Capri, Italy, 26-30 Sep. 1999.
+
+
+
+	
+		Automated Conflict Resolution for Air Traffic Control
+		
+			HErzberger
+		
+	
+	
+		25th International Congress of the Aeronautical Sciences (ICAS)
+		Hamburg, Germany
+		
+			2006
+		
+	
+	Erzberger, H., "Automated Conflict Resolution for Air Traffic Control", 25th International Congress of the Aeronautical Sciences (ICAS), Hamburg, Germany, 2006.
+
+
+
+	
+		Automated conflict resolution, arrival management, and weather avoidance for air traffic management
+		
+			HErzberger
+		
+		
+			TALauderdale
+		
+		
+			Y-CChu
+		
+		10.1177/0954410011417347
+	
+	
+		Proceedings of the Institution of Mechanical Engineers, Part G: Journal of Aerospace Engineering
+		Proceedings of the Institution of Mechanical Engineers, Part G: Journal of Aerospace Engineering
+		0954-4100
+		2041-3025
+		
+			226
+			8
+			
+			2010
+			SAGE Publications
+			Nice, France
+		
+	
+	Erzberger, H., Lauderdale, T., Chu, Y.C., "Automated Conflict Resolution, Arrival Management and Weather Avoidance For ATM", 27th International Congress of the Aeronautical Sciences (ICAS), Nice, France, 2010.
+
+
+
+	
+		Build 4 of the Airspace Concept Evaluation System
+		
+			LarryMeyn
+		
+		
+			RobertWindhorst
+		
+		
+			KarlinRoth
+		
+		
+			DonaldVan Drei
+		
+		
+			GregKubat
+		
+		
+			VikramManikonda
+		
+		
+			SharleneRoney
+		
+		
+			GeorgeHunter
+		
+		
+			AlexHuang
+		
+		
+			GeorgeCouluris
+		
+		10.2514/6.2006-6110
+	
+	
+		AIAA Modeling and Simulation Technologies Conference and Exhibit
+		
+			American Institute of Aeronautics and Astronautics
+			2006
+		
+	
+	Meyn, L., Windhorst, R., Roth, K., Drei D.V., Kubat, G., Manikonda, V., Roney, S., Hunter, G., and Couluris, G., Build 4 of the airspace concepts evaluation system. Proc AIAA Modeling and Simulation Technologies Conference and Exhibit, 2006.
+
+
+
+	
+		Analysis of the Aircraft to Aircraft Conflict Properties in the National Airspace System
+		
+			MikePaglione
+		
+		
+			ConfesorSantiago
+		
+		
+			RobertOaks
+		
+		
+			AndrewCrowell
+		
+		10.2514/6.2008-7143
+		AIAA 2008-7143
+	
+	
+		AIAA Guidance, Navigation and Control Conference and Exhibit
+		Honolulu, Hawaii
+		
+			American Institute of Aeronautics and Astronautics
+			August 18-21, 2008
+		
+	
+	Paglione, M. M., Santiago, C., Crowell, A., Oaks, R.D., "Analysis of the Aircraft to Aircraft Conflict Properties in the National Airspace System", American Institute of Aeronautics and Astronautics Guidance, Navigation, and Control Conference, AIAA 2008-7143, Honolulu, Hawaii, August 18-21, 2008.
+
+
+
+	
+		Sensitivity Analysis of Aviation Environmental Impacts for the Base of Aircraft Data (BADA) Family 4
+		10.2514/6.2021-0457.vid
+		
+			2004
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	User Manual For The Base of Aircraft Data (BADA). Revision 3.6
+	European Organisation For the Safety of Air Navigation, 2004, "User Manual For The Base of Aircraft Data (BADA)", Revision 3.6.
+
+
+
+	
+		Practice for Application of Federal Aviation Administration (FAA) Federal Aviation Regulations Part 21 Requirements to Unmanned Aircraft Systems (UAS)
+		10.1520/f2505
+		
+	
+	
+		Operational Data Reporting Requirements
+		
+			ASTM International
+			November 4, 2011
+		
+	
+	Federal Aviation Administration
+	Federal Aviation Administration, Operational Data Reporting Requirements (OPSNET) url: https://aspm.faa.gov/opsnet/sys/Default.asp , November 4, 2011.
+
+
+				
+			
+		
+	

file105.txt

@@ -0,0 +1,248 @@
+
+	
+	
+		
+I. IntroductionLTHOUGH a number of airport surface movement models exist 1,2,3 and have been successfully used for analysis of airport operations, validation of these models has been a challenge due to a lack of advanced airport surface surveillance.With such data, it is now possible to conduct detailed validation of these models.In this study we have conducted a limited set of analyses to empirically derive operational techniques that are used by controllers in sequencing flights on the airport surface at taxiway intersections.These operational techniques provide greater insight regarding airport surface traffic control and can be used to validate and enhance airport simulation modeling capabilities.We have used the Surface Operations Data Analysis and Adaptation (SODAA) tool 4 to collect and analyze these detailed airport surface operations.
+ATo model airport surface operations with detail and accuracy, it is necessary to consider current techniques and strategies used to determine the taxi route of an aircraft and to establish the sequence to be used whenever two or more aircraft place demand on a taxiway or runway resource simultaneously.Until recently such analysis could only be conducted through visual observation of sequencing decisions, 5,6 whereas now it is possible to analyze such details using airport surface surveillance data through the use of the SODAA tool.The SODAA tool supports NASA's NextGen research 7,8 with a focus on advanced airport surface and terminal operations.SODAA provides the infrastructure and information necessary for NASA researchers and industry analysts to achieve a deep level of knowledge and understanding of airport surface operations.This tool provides data querying and analysis capabilities, as well as advanced data mining features to support analysis of taxi routing, sequencing, and congestion management strategies used by air traffic controllers.The objective of this paper is to describe a novel method of airport surface operations analysis and to provide initial results demonstrating the viability of the technique.The analysis was conducted using airport surface operations data from the Dallas/Fort Worth International (DFW) airport.The Surface Management System (SMS) 9 installation at the North Texas Research Station (NTX) was used to collect surface operations data, which was then analyzed using SODAA to empirically derive sequencing practices.The first section of this paper describes this new methodology used for this work.The second section presents the initial results, and the third section discusses future opportunities and research direction.
+II. MethodologyAirport surface operations at DFW airport provide a useful case study environment for sequencing analysis.For the analyses presented here, we focus on the intersection of taxiways K and EL, as shown in Fig. 1.Taxiway K is a primary north/south taxiway located just to the east of the ramp area for DFW terminals A, C, and E. This taxiway is used for both departures and arrivals as they leave the ramp and taxi to their assigned runway for departure or as the flights taxi toward the ramp after landing.The EL taxiway is one of the primary routes used by flights that have arrived on runway 17C and are crossing runway 17R to reach their parking areas.Thus, the K/EL intersection appears to be an interesting case for a sequencing study.During periods of peak airport demand, both arrival and departure taxi times tend to increase.This is due to departure queuing, communication frequency congestion, and traffic congestion on the airport surface.Any time two flights are in contention for the same intersection at roughly the same time, the Ground Controller must decide which flight passes through first and which must hold.Once this decision is made, the trailing flight must wait.The total elapsed wait time can be broken into the time required for several events to occur: 1) The leading aircraft must reach the intersection.
+K/EL2) The lead aircraft must then pass through the intersection.3) A certain amount of following distance must be established (if following will occur).4) If not previously provided, the trailing aircraft must obtain clearance to continue.The first three steps must happen in sequential order, while the last step may be handled concurrently if the controller gives the direction to proceed after the traffic crosses.This sequencing decision regarding which aircraft leads and the resulting delay experienced by the following aircraft has a significant impact on surface operations.For example, as departing aircraft taxi toward a departure runway, the sequencing decision will ultimately determine the departure order.A difference of one position in the departure order will change the taxi time for a particular flight by a minimum of one or two minutes.For arriving flights, the additional time spent waiting for crossing traffic is the primary consideration when modeling taxi time.A special case of sequencing delay incurred more often by American Institute of Aeronautics and Astronautics Using surface surveillance data, it is possible to determine the location on the taxiway at which flights wait for runway crossings or other sequencing decisions.Figure 2 shows a portion of the taxi path for a single flight that must cross runway 17R and through the intersection between taxiways K and EL on its way to its parking gate.As shown in the figure, the flight stops and waits at both points B and D. However, it is not possible to determine from this information alone what the reasons were for the decision to hold the flight at each of those points.In many cases, a flight is held on the airport surface to implement a sequencing decision.Such sequencing decisions are the focus of this study.We have a two-fold approach for determining how sequencing decisions are made.The first step is to detect situations where flights are in contention for the same intersection and to identify the intersections of interest-those at which sequencing decisions are actually being made versus those where sequencing is merely First-Come-First-Served (FCFS).The second step is to analyze the relevant intersections and corresponding sequencing events to determine the factors that influence the sequence order.American Institute of Aeronautics and Astronautics 3 At any taxiway intersection, SODAA automatically identifies situations in which two aircraft could have crossed the intersection at the same time along crossing or converging paths, as illustrated in Fig. 3.This figure shows a plan view of several DFW terminals, taxiways and runways.The yellow and green lines show the paths of different aircraft that land on different runways, and share a common path during taxi in to the ramp area.In this case, a decision had to be made by a controller to direct one of the aircraft to go first, while the other aircraft would give way to allow the first aircraft to cross.After collecting hundreds or thousands of such sequencing events at each sequencing intersection, SODAA can perform automatic, detailed data mining analysis to find parameters and correlations that provide the strongest indicator of which aircraft would be selected to proceed and which one would be held.For example, at a given intersection, SODAA may find that aircraft on taxiway A are given priority over aircraft on taxiway F over 90% of the time if the aircraft on A can reach the sequencing intersection at or before the time the contending aircraft on taxiway F arrives.Once determined by SODAA, these sequencing parameters can be directly applied to improve the airport surface modeling capabilities of SMS or other fast-time models used to evaluate the benefits of future airport operational procedures.In order to analyze current airport operations, and to also provide the ability to model procedural changes to implement NextGen concepts, it is necessary to develop a modeling system that can both mimic current operational characteristics and implement future procedures.Airport surface sequencing behavior must be a model parameter.For example, future research may identify a novel runway crossing procedure.In order to evaluate the benefits of this procedure, the airport modeling system must be able to conduct both a baseline case model run without the new procedure and a future case model run with the new procedure.SODAA sequencing analysis utilizes recorded target positions on the airport surface to determine locations (e.g., taxiway intersections) that were used by a pair of aircraft.For each situation in which a common intersection was found, SODAA calculates the earliest time that each of the aircraft could have reached the common intersection based on a nominal taxi speed.This earliest crossing time and the actual crossing time for each of the aircraft are used to determine whether the two aircraft could have been at the intersection at the same time.If the following aircraft could have been at the intersection at or before the time that the leading aircraft actually crossed the intersection, then a sequencing event is identified by SODAA.Once a sequencing event has been identified by SODAA, quantitative characteristics of the event are computed and recorded.Example sequencing statistics include the following:-the actual separation time between the two aircraft at the common intersection; -the initial time offset between the two aircraft at the common intersection, which indicates how much earlier one aircraft could have reached the common intersection than the other; and -the amount of delay experienced by each of the aircraft in their taxi from their starting point to the common intersection.To accomplish this analysis, we have extended SODAA to populate two new tables in the SODAA database when flight data is processed.The first table will contain one record for each (flight, node) pair for all nodes through which a flight actually passed.This table stores the earliest estimated time for crossing that node and the actual time that node was crossed.The second table contains one record for each combination of flight, node, and time, where "time" corresponds to a surface surveillance update.For each surveillance update, we calculate the distance to each node remaining in the actual taxi route and estimate the earliest time that flight may reach each of those nodes by dividing distance by a relatively fast nominal taxi speed.After we have the distance and time data populated, we can create "waterfall charts" by plotting, for one node, the distance versus time profile of all flights as they approach that common node.Thus, if a flight stops on the taxiway, its distance to the common node will remain constant as time progresses, and the waterfall diagram will show a flat line.A flight taxiing at a nominal taxi speed will appear in the waterfall diagram as a descending line.This will aid in the exploratory analysis of how the flights behave.Figure 4 shows a sample waterfall diagram.Note that in this sample waterfall diagram, we are only showing the time and distance relative to the intersection at a very limited set of discrete points along the taxi path.In a full waterfall diagram, we would expect to see flat spots in the diagram for flights that are stopped on their taxi path, and we expect to see many instances of crossing lines close to the X axis for intersections where sequencing is not simply based on the order of arrival (i.e., FCFS).If there are many instances of crossing lines far away from the node of interest, we expect that we may have to traverse the network to upstream nodes to determine whether sequencing decisions are made there.If we start the analysis at an intersection where sequencing is known to occur, such as the threshold of a departure runway, we can learn how to graphically identify sequencing events.We may then recursively move through the network to identify sequencing events at upstream intersections.
+American Institute of Aeronautics and Astronautics
+III. Initial ResultsFollowing the methodology described above, SODAA was used to analyze multiple sets of airport surface sequencing operations at DFW.The first analysis that we conducted evaluates sequencing characteristics at all intersections at DFW over a six-hour period.Figure 5 compares the initial predicted arrival time of each aircraft in the sequencing event pair at the common intersection.We computed the difference between the two aircraft arrival times to generate the histogram.Positive differences indicate the aircraft that was originally predicted to be able to reach the common intersection first was actually sequenced first.Negative results indicate that the flights crossed the intersection in a non-FCFS order because the flight that crossed the common intersection first was originally predicted to reach the common intersection after the second flight.All intersections that were found to have a sequencing event are included in this set of results.The figure indicates that an FCFS sequence was used in the majority of cases.However, there were some cases in which the sequencing decision resulted in a non-FCFS sequence (at least according to our definition and computation method).American Institute of Aeronautics and Astronautics Figure 6 shows a histogram of the actual separation times observed in surface surveillance data over a 24-hour period at the intersection of taxiways K and EL.The figure shows the distribution of separation times between aircraft that required sequencing on the airport surface.Note that this data only applies to situations in which the two aircraft could have been at the same intersection at the same time.This separation time data provides valuable information about the throughput capacity of an intersection.If all aircraft were able to move freely through the intersection and to continue taxiing without delay, how much time separation would be required between successive aircraft?Physically, this depends on the taxi speed, length of the aircraft, and required buffer distance.As shown in the figure, this information can be derived empirically.Using the detailed data that has been computed, including the predicted crossing time at this intersection for each flight as a function of time, we have analyzed controller decision-making regarding the sequence of flights through this intersection.To accomplish this analysis, we created a set of geospatial regions in the SODAA tool and used a query to obtain the first entry time of each flight into each of the geospatial regions.The geospatial regions were located on the taxi routes approaching the K/EL intersection, as shown in Fig. 7.As flights taxi through each of the geospatial regions, the SODAA query provides the time of entry.Using a nominal taxi speed and the distance from each of the geospatial regions to the K/EL intersection, we compute the earliest crossing time of the K/EL intersection for each flight at each of the regions.
+Initial Predicted Time Offset at
+Time Separation Between Flights in IntersectionSequencing Events
+Figure 6. Actual separation times between aircraft at common intersection (K and EL).Although these calculations do not give us a full waterfall diagram for the flight, we can analyze the time of intersection crossing predicted at each of the geospatial regions that the flight crosses to evaluate whether or not the flight is sequenced at the intersection in an FCFS order.American Institute of Aeronautics and Astronautics 7 Note that the layout of the geospatial regions has been designed to monitor multiple approaches to the K/EL intersection.Our hypothesis in designing these geospatial regions in this manner is that the direction and route that is used to approach the intersection has a significant impact on the controller decision-making process regarding the flight sequence.On the left side of Fig. 7, three geospatial regions have been created that encompass a group of 'spots'.A 'spot' or apron entrance/exit point marks a location on the airport surface at which flights transition from the Airport Movement Area (AMA) to the ramp area.The spots that are included in the geospatial regions are 42, 43, and 44 in the first group, 45 and 46 in the second region, and 47 and 48 in the last region.Departure flights stop and hold in these regions waiting for taxi clearance from the tower to proceed onto taxiway K. On taxiway K, we have multiple geospatial regions.A geospatial region at the intersection of K and EM is shown in the figure.The largest portion of traffic through K/EL goes through the K/EM intersection.The traffic through this intersection includes arrival flights heading for their parking gates and departure flights that have left spots further south of the EM taxiway.
+K/EL
+Runway 17RTaxiway EL Taxiway K The geospatial regions used for this analysis also include intersections on taxiway EL.This taxiway is used by arrival flights to cross runway 17R and to proceed toward their gates.We have created two geospatial regions-one immediately before the flights cross runway 17R, and one after 17R has been crossed and before the intersection with taxiway L.
+K/EMBy identifying the earliest time at K/EL for each flight at each of these geospatial regions, we determine whether or not flights are handled at K/EL in an FCFS order.If a flight's earliest crossing time at K/EL is earlier than the actual crossing time of the flight ahead of it at the intersection, then we consider the flight to have been handled in a non-FCFS order.As the flight enters each geospatial region, we compute the earliest time of arrival at K/EL.Since flights may progress towards K/EL with varying average velocities, the predicted K/EL sequencing order will change from one geospatial region to the next.Figure 8 shows sequencing analysis results for 414 flights that traversed the K/EL intersection during a 24-hour period.In the figure, a pair of numbers is shown for each of the geospatial regions.The number below the line is the total number of flights that went through the indicated geospatial region on the way to the intersection of taxiways K and EL.The number above the line is the number of flights that were handled at K/EL in non-FCFS order.For example, 21 flights out of 89 that waited east of 17R on EL were not handled in FCFS order at K/EL.Of those 21, three of them took their non-FCFS sequencing delay before getting to the geospatial region east of the L taxiway.The other 18 took their non-FCFS delay between the geospatial region on EL east of L and the K/EL intersection.This result indicates, as would be expected, that flights taxiing on the K taxiway, which is the primary route for departure and arrivals, are more likely to be sequenced ahead of flights that are merging onto the K taxiway.Flights that are leaving the spots seem to have a higher percentage of cases in which they are sequenced out of FCFS order.Although we have not analyzed the taxi route pairs for each sequencing decision, based on the number of flights traveling on the K taxiway, we expect that most of the sequencing decisions for flights coming out of the spots are made with respect to flights taxiing north on the K taxiway.It is reasonable to expect that flights on the K taxiway would receive some preference because they are more likely to be up to speed, whereas the flights leaving the spots are more likely to be at a full stop while they wait for a taxi clearance.It appears from the data that flights taxiing across runway 17R and entering the K taxiway are handled in a non-FCFS order at a higher frequency than those traveling north on the K taxiway.This may be due to the fact that flight crews must be on the Local Controller's frequency while crossing runway 17R, and then (usually) they must switch to the Ground Controller's frequency before receiving the remainder of their taxi clearance.The data indicate that aircraft often stop short of taxiway K after crossing runway 17R, which would be a common result of the need to change frequency and receive further clearance to taxi into the ramp.
+American Institute of Aeronautics and AstronauticsNote that the results shown in Fig. 8 include all flights crossing the K/EL intersection during a 24-hour period.During that period of time, however, there would be many flights that crossed through the K/EL intersection that did not require a sequencing decision to be made at all because there were no other flights that were competing for access to the intersection.By limiting the data to only those pairs of flights that could have arrived at the K/EL intersection within one minute of each other, we can more accurately evaluate the decision-making and sequencing techniques used by the controller when sequencing is necessary.Figure 9 shows the results when we consider only flights that require a sequence decision to be made.Notice that the flights leaving the spots are even more likely to be held for other traffic when there a sequencing decision to be made.This version of the results also illustrates another surface operations phenomenon.Note that there are 49 flights that pass through the geospatial region on EL east of runway 17R that are considered to be part of a sequence event at the K/EL intersection.However, after the flights have held and waited to cross runway 17R, only 43 of the original 49 flights are still involved in a sequencing event at the K/EL intersection.This is an example of the upstream effects on a downstream intersection that must be considered when formulating conclusions about controller decision-making at a given intersection.Finally, in Fig. 10, we present an analysis of the delay allocated to aircraft that are sequenced in a non-FCFS order.Although there are many different reasons that flights may be held on the airport surfaceincluding a parking gate that is not available, a Traffic Flow Management (TFM) restriction, or a mechanical issue-we have designed this analysis to exclusively evaluate the amount of delay assigned to a flight because of the sequencing of that flight behind another.The results shown in the figure indicate that the sequencing delay is generally less than two minutes.
+Histogram of Delay for Non-FCFS Sequences
+IV. ConclusionsIn this study we have conducted a limited set of analyses to empirically derive operational techniques used by controllers in sequencing flights on the airport surface at intersections between taxiways.While controller techniques may vary, initial results suggest that consistent sequencing patterns can be identified.Further, the results indicate that sequencing decisions are dependent on the flight status (in motion vs. stopped) and taxiway location.For example, our results for this particular case study at DFW indicate that almost 90% of flights that are established on a major taxi route (taxiway K) are handled in an FCFS order, while only 50% of flights leaving the spots and merging onto the taxiway are handled in an FCFS order.These initial results, and the analysis techniques that have been developed through this study, provide the means by which airport surface decision support tools and airport surface models can be improved to accurately represent microscopic decisions on the airport surface that can have significant effects on the flow of the overall air transportation system.American Institute of Aeronautics and Astronautics
+V. Future WorkFuture work will include analysis to characterize the decision factors involved in sequencing situations.The present study has identified some indicators of controller technique at specific intersections at DFW.In future studies we intend to construct a logistic regression model to predict the likelihood of a flight being the next one to proceed through an intersection.Logistic regression is a generalized linear model that fits a binomial response variable to a linear combination of independent variables.The basic model is described by Eqs.(1-3) below.In our case, π would represent the probability of being the next flight to use the intersection.) exp( 1) exp( ) ( The independent variables (x 1 …x k ) may be designed to represent a combination of categorical or numeric values.Figure 11 shows the shape of the linking function in Eq. ( 2), which maps the linear model to the non-linear probability.In a fashion somewhat similar to multiple linear regression, we can test the fit of the model with and without each factor to identify those that provide a statistically significant improvement in the fit of the model.We plan to build the model data set by sampling at random times from a relatively large time interval of data, where the data for one intersection at one time will consist of all flights that meet all of the following conditions at the sample time:1 1 0 1 1 0 k k k k x x x x x β β β β β β π Κ Κ + + + + + = (1) ⎥ ⎦ ⎤ ⎢ ⎣ ⎡ - = ) ( 1 ) ( ln ) ( x x x g π π (2) k k x x x g ) ( β + β = 1 1 0 + Κ β1) The actual taxi path includes the intersection, 2) The flight is active (meaning we have taxi surveillance data), and 3) The flight has not yet reached the intersection.The set of state variables to be used is as yet undetermined, but will likely include the following:1) Distance to the intersection; 2) Current taxiway link; 3) Speed, possibly categorized into {stopped, slow, fast, etc..}; 4) Timeliness of flight, possibly categorized into {early, on-time, late, etc...}; 5) Aircraft type; 6) Airline; 7) Controlled departure time, if any (departures only); and 8) Departure fix/procedure.To these state variables, we can add prior knowledge of whether or not the flight actually was next to pass through the intersection, which is the binomial independent variable we are trying to fit.Once we compute probabilities for being the next in sequence, we must develop a model that applies them to automatic sequencing decisions.For example, we may choose to simply pick the flight with the highest probability or combine the probabilities of potential candidates into an odds ratio and use that to decide.The development of that model will depend heavily on the outcome of the logistic regression and our ability to create a model that will accurately estimate the likelihood of being the next in the sequence.Other aspects of sequencing decisions will be evaluated as well, such as decisions regarding the sequencing of arrivals and departures on a runway, as well as the sequencing of aircraft crossing runways with arrival and departure traffic.Figure 1 .1Figure 1.DFW airport east-side taxiway layout.
+Figure 2 .2Figure 2. SODAA display showing taxi track and speed on the surface.arrivals is the crossing of active runways.In this case, the arriving aircraft must wait for departures, and possibly arrivals, to use the runway prior to obtaining clearance from the Local Controller to cross.Using surface surveillance data, it is possible to determine the location on the taxiway at which flights wait for runway crossings or other sequencing decisions.Figure2shows a portion of the taxi path for a single flight that must cross runway 17R and through the intersection between taxiways K and EL on its way to its parking gate.As shown in the figure, the flight stops and waits at both points B and D. However, it is not possible to determine from this information alone what the reasons were for the decision to hold the flight at each of those points.In many cases, a flight is held on the airport surface to implement a sequencing decision.Such sequencing decisions are the focus of this study.We have a two-fold approach for determining how sequencing decisions are made.The first step is to detect situations where flights are in contention for the same intersection and to identify the intersections of interest-those at which sequencing decisions are actually being made versus those where sequencing is merely First-Come-First-Served (FCFS).The second step is to analyze the relevant intersections and corresponding sequencing events to determine the factors that influence the sequence order.
+Figure 3 .3Figure 3. Intersecting aircraft surface tracks at DFW.
+Figure 4 .4Figure 4. Sample waterfall diagram for intersection K and EL (point E in figure).
+Figure 5 .5Figure 5.Initial predicted arrival time difference at common intersections.
+Figure 7 .7Figure 7. Geospatial regions surround the intersection of taxiway K and taxiway EL.
+Figure 8 .8Figure 8. Flights sequenced in non-FCFS order compared to total flights.
+Figure 9 .9Figure 9. Flights sequenced in non-FCFS order when sequencing is necessary.
+Figure 10 .10Figure 10.Sequencing delay for flights that are handled in non-FCFS order.
+of fitted model coefficients 0...k x are the independent variables included in the model 1...k π (x) is the binomial response variable being fit g (x) is the linking function
+Figure 11 .11Figure 11.Shape of non-linear linking function.
+		
+		
+			
+
+				
+
+
+	
+		Airport Simulation for Rapid Decision-Making: TAAM for DFW
+		
+			JCrites
+		
+		
+			EMeyer
+		
+	
+	
+		Airport-Airspace Simulations-A New Outlook, TRB Annual Meeting
+		
+			13 Jan. 2001
+		
+	
+	Crites, J., Meyer, E., "Airport Simulation for Rapid Decision-Making: TAAM for DFW", Airport-Airspace Simulations-A New Outlook, TRB Annual Meeting, 13 Jan. 2001.
+
+
+
+	
+		Airport and Airspace Simulation Model, Software Package
+		
+			Simmod
+		
+		
+			2007
+			Sunnyvale, CA
+		
+	
+	Ver 7.3, ATAC
+	SIMMOD, Airport and Airspace Simulation Model, Software Package, Ver 7.3, ATAC, Sunnyvale, CA, 2007.
+
+
+
+	
+		
+			ATrani
+		
+		
+			HBaik
+		
+		
+			JMartinez
+		
+		
+			VKamut
+		
+		A New Paradigm to Model Aircraft Operations at Airports:The Virginia Tech Airport Simulation Model (VTASIM)
+		
+			13 Nov. 2000
+		
+	
+	Nextor Research Symposium
+	Trani, A., Baik, H., Martinez, J., Kamut, V., "A New Paradigm to Model Aircraft Operations at Airports:The Virginia Tech Airport Simulation Model (VTASIM)", Nextor Research Symposium, 13 Nov. 2000.
+
+
+
+	
+		Surface Operations Data Analysis and Adaptation tool, Software Package, Ver. 1.8, Mosaic ATM
+		
+			Sodaa
+		
+		
+			2008
+			Leesburg, VA
+		
+	
+	SODAA, Surface Operations Data Analysis and Adaptation tool, Software Package, Ver. 1.8, Mosaic ATM, Leesburg, VA, 2008.
+
+
+
+	
+		Macrocognition in Systems Engineering: Supporting Changes in the Air Traffic Control Tower
+		
+			CBonaceto
+		
+		
+			SEstes
+		
+		
+			PMoertl
+		
+		
+			KBurns
+		
+		10.1201/9781315597584-15
+	
+	
+		Naturalistic Decision Making and Macrocognition
+		Amsterdam, The Netherlands
+		
+			CRC Press
+			Jun. 2005
+			
+		
+	
+	Bonaceto, C., Estes, S., Moertl, P., and Burns, K., "Naturalistic Decision Making in the Air Traffic Control Tower: Combining Approaches to Support Changes in Procedures," Proceedings of the Seventh International NDM Conference, Amsterdam, The Netherlands, Jun. 2005.
+
+
+
+	
+		Observations of Departure Processes at Logan Airport to Support the Development of Departure Planning Tools
+		
+			HusniRIdris
+		
+		
+			IoannisAnagnostakis
+		
+		
+			BertrandDelcaire
+		
+		
+			RJohnHansman
+		
+		
+			John-PaulClarke
+		
+		
+			EricFeron
+		
+		
+			AmedeoROdoni
+		
+		10.2514/atcq.7.4.229
+	
+	
+		Air Traffic Control Quarterly
+		Air Traffic Control Quarterly
+		1064-3818
+		2472-5757
+		
+			7
+			4
+			
+			1998
+			American Institute of Aeronautics and Astronautics (AIAA)
+			Orlando, FL
+		
+	
+	Idris, H., Delcaire, B., Anagnostakis, I., Hall, W., Clarke, J., Hansman, R., Feron, E. and Odoni, A., "Observations of Departure Processes at Logan Airport to Support the Development of Departure Planning Tools," 2nd USA/Europe ATM R&D Seminar, Orlando, FL, 1998.
+
+
+
+	
+		Next Generation Air Transportation System (NGATS) Air Traffic Management (ATM) -Airportal Project
+		
+			DHinton
+		
+		
+			JKoelling
+		
+		
+			MMadson
+		
+		
+	
+	
+		NASA External Release Version
+		
+			23 May 2007
+		
+	
+	Hinton, D., Koelling, J., and Madson, M., "Next Generation Air Transportation System (NGATS) Air Traffic Management (ATM) -Airportal Project," NASA External Release Version: http://www.aeronautics.nasa.gov/nra_pdf/airportal_project_c1.pdf, 23 May 2007.
+
+
+
+	
+		Next Generation Air Transportation System (NGATS) Air Traffic Management (ATM) -Airspace Project
+		
+			HSwenson
+		
+		
+			RBarhydt
+		
+		
+			MLandis
+		
+		
+	
+	
+		NASA External Release Version
+		
+			1 Jun. 2006
+		
+	
+	Swenson, H., Barhydt, R., and Landis, M., "Next Generation Air Transportation System (NGATS) Air Traffic Management (ATM) -Airspace Project," NASA External Release Version: http://www.aeronautics.nasa.gov/nra_pdf/airspace_project_c1.pdf, 1 Jun. 2006.
+
+
+
+	
+		Concept Description and Development Plan for the Surface Management System
+		
+			SAtkins
+		
+		
+			CBrinton
+		
+	
+	
+		Journal of Air Traffic Control
+		
+			2002
+		
+	
+	Atkins, S., and Brinton, C., "Concept Description and Development Plan for the Surface Management System," Journal of Air Traffic Control, 2002.
+
+
+				
+			
+		
+	

file106.txt

@@ -0,0 +1,915 @@
+
+	
+	
+		
+I. IntroductionW ith the expected introduction of unmanned aircraft to move goods [1] and people [2-4] and conduct other novel operations like structural monitoring, surveying, etc, future airspace could be filled with traffic orders of magnitude higher than it can bear today.How many such aircraft operations can be accommodated in low-altitude airspace given a set of technological capabilities, operational requirements and protocols, while maintaining safety, stability, performance efficiency and an optimal flow of traffic?In this paper, we address that question by proposing a throughput-based airspace capacity metric.Historically, airspace capacity has been constrained by manual air traffic controller workload [5][6][7][8].The system has evolved with very stringent requirements on safety, as any loss of flight is catastrophic.This may change for unmanned operations for two main reasons.First, the constraint of a manual controller is relaxed.Automated traffic management should accommodate higher traffic densities.It has been shown to do so to some extent even for manned aviation [9,10].Second, not all crashes will be catastrophic.Most may instead result in property damage and not injury or death.Hence, this opens up the opportunity to explore new approaches to estimate capacity for operations in low-altitude airspace.Our throughput idea is inspired by the concept of the fundamental diagram [11], a component of kinematic wave theory.An extensive application of the concept in road transportation [12] relates the freeway traffic flow to the traffic density (Fig. 1).This has been researched for over seven decades and is well understood and utilized in road transportation [13].Further, the expected future demand of over 100,000 flights per day [14] (just for package delivery in a single metropolitan region) is closer to volumes traditionally handled in road transportation.Hence, it provides a reasonable starting point for further research into estimating airspace capacity for novel air traffic operations.Intuitively, as inflow into an airspace volume increases from zero, the throughput (i.e.number of aircraft traversing the airspace per unit time) increases as well.This induces a corresponding rise in accumulation (density).However as aircraft begin to excessively impede each other to avoid losses of separation, the traffic becomes congested and throughput decreases.The aircraft must slow down or deviate significantly from their intended path.The throughput should eventually drop to a minimum steady-state value at a maximum aircraft density that preserves safety.The presence of a peak throughput value suggests that operating beyond that regime will be inefficient, even if it is still safe.Therefore, this constrains the capacity of the airspace.In our current work, we study whether such traffic behavior is actually exhibited by aircraft traversing an airspace.The study is restricted to small Unmanned Aircraft Systems (sUAS) traffic in this paper.Furthermore, capacity is a function of technology.Technology dictates the conflict detection and resolution (CD&R) [15] capability and the allowable minimum separations between the aircraft.Hence, we evaluate the throughput behavior for different CD&R algorithms and separation minima and use a simulation paradigm to produce the results.The rest of this paper is structured as follows.We first present a review of related work that motivates this effort under section II.Section III lays out our approach in detail.Metrics, CD&R methods and the simulation platforms are discussed under section IV.In V we present the results showing the peak throughput behavior with different CD&R methods and separation minima.Section VI concludes this paper with a discussion of proposed extensions of the work.
+II. Literature ReviewApproaches in manned aviation literature frequently estimate capacity as a function of controller and pilot workload [5][6][7][8].Capacity is derived from air traffic complexity measures such as Monitor Alert Parameter (MAP) [16], the maximum number of aircraft an Air Traffic Control (ATC) controller can handle at any given time and Dynamic Density (DD) [17,18], a weighted summation of factors that affect the air traffic complexity.There is an inherent assumption of a structured airspace and Air Traffic Management (ATM) that includes monitors, sectors and airways [19][20][21].Capacity is then estimated using fast-time and real-time simulation methods [22] in a highly subjective manner biased by the judgment of human air traffic controllers during the experiments, who are also assumed to be the bottleneck of the system.A second approach called the Eurocontrol Care-Integra models the ATM system as a combination of several information processing agents, each with an associated information processing load (IPL) [23].The system reaches capacity when one of the agents overloads.This is deterministic for machine agents but again needs subjective judgment for human agents.However, it finds the bottleneck in the system instead of assuming it.Road transportation practice uses a third approach that pins down the bottleneck by measuring the change in lane throughput as a function of the freeway traffic density (see Fig. 1).
+Fig. 1 Fundamental Diagram of Traffic FlowOur throughput-based capacity metric is inspired by the latter two approaches.Further, the fundamental diagram approach was recently studied for highly structured onedimensional sUAS traffic flow in sky lanes in urban areas and shown to exhibit a threshold behavior [24].We want to explore if this also holds true for unstructured free-flow traffic in an area (2D) and eventually in an airspace volume.Future operations both for sUAS and Urban Air Mobility (UAM) [4] may be free flight in nature; i.e. individual flights could be responsible for determining their own courses, independent of a global plan or system.Unmanned Aircraft Systems (UAS) Traffic Management (UTM) should therefore support userpreferred flight trajectories to the extent possible.Any chosen metrics should account for this.ATM architectures that transfer some of the separation responsibility to the cockpit for manned free flight were researched by Bilimoria et al. as part of their Distributed Air/Ground Traffic Management (DAG-TM) concept [9,10,25,26].Based on their work, the following types of metrics can potentially be used to evaluate any UTM architecture for free flight (the sample measures used for manned ATM [9] are listed in parenthesis): Safety (number of actual conflicts and conflict alerts), Performance (Change in direct operating cost), Stability (number of forced conflicts (domino effect)) and DD (aircraft density, average proximity and average point of closest approach).Of these, we focus on the first two as the basis of our chosen comparative metrics.
+Fig. 2 CDR geometry based on choosing the lower cost choice between the frontside and backside maneuver [9]Next comes the choice of a CD&R algorithm.CD&R methods in aviation literature have been primarily developed for large aircraft [15] flying at higher altitudes and lower densities than the expected future sUAS traffic.An example of a simple rule used by Krozel et al. [9] is shown in Figure 2. Smaller unmanned aircraft provide a unique opportunity for simpler conflict resolution algorithms.Proposed future operations [27] might primarily be done by aircraft that have Vertical Take Off and Landing (VTOL) capability and better maneuverability.Under these considerations, we chose three simple algorithms in this work.We discuss these further under section IV-IV.B Finally, we need a simulator that can simulate sUAS traffic densities so that the throughput behavior can be studied.Existing simulation and evaluation tools developed for ATM (like BlueSky, TMX, ACES, AEDT, FACET) may not fit for our purposes -they are designed to handle manned-aircraft with a much lower traffic density than our study.They take into account interactions with a variety of actors (air traffic controllers, etc.) that are not needed in low-altitude UTM questions of traffic behavior and capacity.The fast-time Fe3 simulator developed by NASA Ames [28] provides the capability of statistically analyzing the high-density, high-fidelity, and low-altitude traffic system.It can be used for effectively evaluating policies and concepts, and performing parameter studies in a higher-fidelity environment like the one in which we are interested.Hence, we use it in conjunction with a simple kinematic model-based Matlab simulator developed for simpler algorithms.
+III. ApproachAirspace capacity is the maximum number of aircraft that can traverse an airspace in a given time under a set of requirements.Throughput is a way of quantifying the capacity per unit time.Prior analysis [14] suggested that demand for sUAS package deliveries could be as high as 100,000 flights per day in a metropolitan region like the San Francisco Bay Area.A threshold-based definition was used to study these estimates and establish airspace capacity for such a metropolitan region in terms of "flights per day" considering safety and performance efficiency [29].Such a macroscopic approach, although useful for long-term planning and design of an airspace system, provides no direct method of real-time control.On the other hand, the flow density relation is used as a tool to control road traffic by regulating inflow in real time and improve throughput.Hence, if the peak throughput behavior is exhibited by air traffic, a similar air traffic control method could also be explored for operating the airspace at or close to capacity.There is no empirical data on sUAS traffic in the airspace today on which to base our methods.Hence, we start by considering a representative area, subjecting it to increasing steady state inflow rates of air traffic and measuring the mean outflow rate, which we call throughput of the airspace.Next, to study the feasibility of the throughput metric, we pick other metrics to compare against, use parameters that model the technology, and develop a computational process that quantifies the metrics as a function of the technology parameters.The primary goals of air traffic management are to maximize safety, capacity, and efficiency.In Section IV we discuss the safety and efficiency metrics that are evaluated along with the throughput metric.We make the following operational assumptions about the aircraft and their operations: (a) All aircraft are sUAS with strictly VTOL capability; (b) Their flight plans are straight line paths from entry to exit on the boundary of the study area.These paths change as aircraft fly through the airspace and avoid conflicts with other aircraft using a given CD&R algorithm; and (c) All sUAS have nominal and maximum speeds constrained by the capabilities of typical sUAS in use today.Our setup is two dimensional.Any losses of separation are horizontal (a simplification to evaluate the throughput-based capacity metric).We plan to extend this to a volumetric study in the future.The detailed simulation setup and the chosen metrics and CD&R methods are described next.We first define the notion of a conflict and loss of separation.Any sUAS should stay out of a minimum separation exclusion zone (a disc with radius D) around another sUAS.A loss of separation occurs when two sUAS come within this minimum separation.Given their projected paths in the horizontal plane, if an sUAS is predicted to eventually enter within the minimum separation of another sUAS, the two aircraft are said to be in conflict.Figure 3 illustrates a loss of separation occurring between two sUAS.
+IV. SimulationWe make the following assumptions.We only consider multicopters, which means that the aircraft can hover.The nominal flight speed is assumed to be 15 ms -1 with a maximum value of 20 ms -1 .The maximum acceleration that today's sUAS can achieve is about 2g.If 1g is used to overcome the weight, close to 1g is available for horizontal maneuvers while keeping the aircraft in safe operational limits.To avoid pushing aircraft to their maximum capability all the time, we limited the maximum acceleration to 0.5g.We chose a representative area as a square of 0.5km width.The origins and destinations of aircraft are uniformly distributed along the edges and are spaced such that two aircraft don't enter or exit within loss of separation distance.These are randomly connected to form the flight paths such that no aircraft has an origin and destination on the same edge.This ensures that every aircraft enters the study area.Finally, we estimate the different metrics for two different separation minima -5 meters and 10 meters.
+A. Metrics
+ThroughputOur primary metric -Trip Exits per min captures the average traffic outflow rate through the area (i.e.throughput).Measuring trip exits per second would be too small to capture substantial intended boundary crossings and measurements over an hour would be too long to provide any real-time control over an area.
+SafetySafe operation of the airspace is of utmost importance.Following the proposed requirements by MITRE [30] and its use in our prior macroscopic capacity estimation work, we choose the necessary safety metric as the Total Losses of Separation observed over the simulation interval.
+Performance EfficiencyHigher operating costs (fuel, wear, etc.) lower performance.They are typically caused by longer travel times and distances, which are in turn usually the byproduct of safer operation.We capture this in the current work by measuring the Percentage Extension in Travel Time.This is a direct derivative of the Change in Direct Operating Cost as proposed by Krozel et al. [9].
+B. CD&R methodsApproaches to CD&R may be broadly classified into three categories -force field based, trajectory projection based and offline look-up table-based.We chose three simple CD&R methods that capture different types of control, represent different categories of CD&R approaches and can together encompass most types of aircraft.This makes them flexible for future extensions of this study to different classes of aircraft.First is avoidance by slowing down to "Hover."This captures the effect of pure speed control and encompasses aircraft that can stop in flight.It uses a trajectory projection-based approach.Second is a simplified implementation of "Potential Field" method as used by Mueller [31].This uses a simultaneous speed and direction control.Since the minimum speed can be set greater than 0, it captures all aircraft with a stall speed constraint.It belongs to the broad category of force field-based CD&R approaches.The third is an algorithm derived from the ICAROUS [32] architecture, that is based on DAIDALUS [33], a reference implementation of RTCA-228 Minimum Operational Performance Standards (MOPS)(Appendix G) for UAS DAA (Detect and Avoid) [34].This also uses speed and direction control and is extendable to all aircraft with a stall speed constraint.It acts as a more complex example of trajectory prediction-based approaches under formal consideration.In the rest of this paper, we will refer to this ICAROUS-based algorithm as "ICb" for brevity.We used a kinematic model-based simulator in Matlab to study the throughput behavior for Hover and Potential Field and used the same flight data to study ICb as implemented on the fast-time simulation platform Fe3.Fe3 is highlyparallelized and implemented on Amazon Web Services(AWS) Graphical Processing Unit(GPU) instances.It includes various six-degree-of-freedom vehicle models and CD&R algorithms and also incorporates vehicle communication and sensor models and wind models.Although other components, such as no-fly zones, near-ground static and dynamic obstacles and avoidance, and community effect via noise and pollution, are still under development, Fe3 provides essential functionality necessary for our study.
+V. ResultsIn this section, we present our results that describe the throughput-based capacity metric.Figures 4, 5 and6 show the variation of throughput for hover, potential field and ICb.The figures on the left compare throughput to number of losses of separation during the simulation and the figures on the right compare it to the percentage extension of travel time.In all the figures, solid line and dotted line represent 5 meters and 10 meters minimum separation, respectively.The metrics are evaluated as a function of different steady-state inflow rates.The average area outflow rate measured as Trip Exits per min is plotted in blue.The losses of separation and the mean percentage extension of travel time measured over the entire simulation are plotted in orange on the left and right figures respectively.We observe the following general trends.In all figures, a peak throughput behavior is exhibited at an intermediate steady-state traffic inflow (between 60 to 80 flights per min for Hover and ICb and between 40 to 60 flights per min for Potential Field).Any losses of separation and noticeable extensions of travel times occur at or beyond the peak throughput.In other words, peak throughput is achieved even before safety of the system is compromised.Therefore, in this airspace, the optimal inflow to be maintained is decided by throughput rather than safety.When the tolerance is higher (smaller minimum separation), the throughput is also higher as the aircraft can be safely packed closer together.Under the same steady-state inflow conditions, for example between 60 and 80 flights per min, the highest throughput is shown by both Hover and ICb (about 50 trip exits per min) but Hover exhibits it at a slightly higher inflow rate.Potential Field shows lower peak throughput than the other two.It also peaks at an inflow rate of 40 trips per min, almost half that of the other two.However, this loss comes at a much higher level of safety.This is shown by the loss of separation numbers at and beyond peak throughput.Both Hover and ICb start deteriorating in terms of safety beyond their respective throughput peaks, while Potential Field maintains its low losses of separation (below 4) well beyond.Further, the number of losses of separation rise rapidly for the Hover and ICb cases but they stay low for potential field.Practically, this gives flexibility to the system to operate at peak, while for the other two, from a risk standpoint, it is preferable to operate to the left of the peak.Next we compare the performance efficiency.ICb fares better than Hover in terms of travel time extension.For example, close to the peak throughput, hovering extends mean travel time by 1% for the 5m separation case, while the value for ICb is around 0.8%.The percentage extension is almost thrice at peak (about 3%) for Potential Field.The difference is more pronounced beyond the peaks.In the entire simulation, Hover exhibits a maximum mean travel time extension of 3%.The same metric for ICb is 1.2%, while it is slightly more than thrice for Potential Field at 10%.But as stated earlier, what Potential Field loses in efficiency, it compensates for in safety.These behaviors are explained as follows.Our hover approach slows down aircraft to a stop without deviating them from their trajectory.Hence beyond peak throughput, first there is an excessive slowdown that reduces the throughput.Next, when all aircraft begin to stop while the inflow is still maintained, several aircraft don't have enough distance/time to stop safely.Hence, they start entering each others' minimum separation distances, especially closer to the boundaries.The number of losses of separation rises fast and exit rates continue to fall.However, the aircraft within separation minima are either stopped or moving very slowly.This is comparable to a jam on a freeway where cars are bumper to ICb picks the resolution maneuver from the recovery bands provided by DAIDALUS that has the least secondary conflicts and simultaneously minimizes the deviation from the nominal trajectory.This results in low travel time extensions and higher throughput.As more aircraft accumulate in the airspace, the recovery bands become narrower and hence lead to higher number of losses of separation.In the potential field approach, the aircraft at or close to minimum separation have large repulsive forces, which ensure that the aircraft are kept away from each other and hence very safe.The small number of losses of separation happen when excessive aircraft have accumulated in the system and either the repulsive forces start overwhelming the aircraft operation limits or an aircraft trying to reach its destination ends very close to an originating aircraft.Both of these scenarios can be minimized by implementing appropriate entry and exit rules or providing buffer zones at high inflow rates.Safety is achieved by spreading the aircraft beyond the primary study region boundaries.To understand this better, let us assume that the study region was an urban area.This approach pushes out aircraft at the edges of the area into suburban airspace.Hence, a higher amount of contingency airspace is required.Based on the above insights we find that the throughput metric is not only useful to understand airspace capacity as a function of technology, but our comparative approach also provides a basis for evaluating CD&R methods in terms of the capacity, safety and efficiency they can achieve at high density system level operations.
+VI. Conclusions and Future WorkWe have developed a throughput-based airspace capacity metric for unmanned air traffic in low-altitude airspace.Throughput, safety and performance metrics were evaluated for uniformly distributed air traffic inflow over a square area of 0.5 km width.We used three CD&R methods -Hover, Potential Field and ICb, and two separation minima -5 meters and 10 meters in our simulations.Our results show the throughput behavior as a function of the steady-state air traffic inflow in a representative area.The system stays safe (i.e.no losses of separation) without excessive impact on performance (less than 5.5% mean extension of travel time) until after the accumulation of air traffic has lead to a reduction in throughput.This suggests that the throughput peak may quantify the airspace capacity.The CD&R algorithms themselves exhibited different throughput peaks.Further, smaller separation requirements allowed better throughput no matter which algorithm was used.We also observed that measuring throughput in comparison to safety and efficiency metrics could be used as a tool to evaluate the adequacy of a CD&R algorithm for large-scale operations.A next step in the evaluation of this approach is to use other other robust CD&R methods such as Airborne Collision Avoidance System X (ACAS X) developed for multi-copters [31].This would capture an offline look-up table-based CD&R method, which we didn't explore in this paper.We also need to measure the effect of sensor and navigational uncertainties (such as deviations from trajectory, delays in aircraft detection, wind, etc), static and dynamic obstacles (e.g., buildings and temporary flight restrictions) and specific traffic flow patterns.Fig. 33Fig. 3 Conflict and Loss of Separation.Ao -Own sUAS, Ai 1 & Ai 2 -Intruder sUAS.The aircraft are shown in relative frame of reference
+Fig. 4 Fig. 545Fig. 4 Results for Hover to Avoid
+		
+		
+
+			
+AcknowledgmentWe express our sincere gratitude to Dr. Alex A. Kurzhanskiy from UC, Berkeley, Dr. Parimal H. Kopardekar from the NASA Ames Research Center for valuable insights on the basic ideas.We thank Joseph Silva and the entire Fe3 team at NASA Ames for their ongoing contributions to the fast-time simulator used in the current work.This work is funded in part by the USRA NAMS Student R&D Program.
+			
+
+			
+
+				
+
+
+	
+		Flight Demonstration of Unmanned Aircraft System (UAS) Traffic Management (UTM) at Technical Capa...
+		
+			PHKopardekar
+		
+		10.2514/6.2020-2851.vid
+	
+	
+		NASA Ames Research Center
+		
+			Apr. 1, 2014
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	Tech. rep.
+	Kopardekar, P. H., "Unmanned Aerial System (UAS) Traffic Management (UTM): Enabling Low-Altitude Airspace and UAS Operations," Tech. rep., NASA Ames Research Center, Apr. 1, 2014.
+
+
+
+	
+		Enabling Airspace Integration for High-Density On-Demand Mobility Operations
+		
+			EricRMueller
+		
+		
+			ParmialHKopardekar
+		
+		
+			KennethHGoodrich
+		
+		10.2514/6.2017-3086
+	
+	
+		17th AIAA Aviation Technology, Integration, and Operations Conference
+		
+			American Institute of Aeronautics and Astronautics
+			2017
+			3086
+		
+	
+	Mueller, E., Kopardekar, P., and Goodrich, K. H., "Enabling Airspace Integration for High-Density On-Demand Mobility Operations," 17th AIAA Aviation Technology, Integration, and Operations Conference, 2017, p. 3086.
+
+
+
+	
+		Silicon Valley as an Early Adopter for On-Demand Civil VTOL Operations
+		
+			KevinRAntcliff
+		
+		
+			MarkDMoore
+		
+		
+			KennethHGoodrich
+		
+		10.2514/6.2016-3466
+	
+	
+		16th AIAA Aviation Technology, Integration, and Operations Conference
+		
+			American Institute of Aeronautics and Astronautics
+			2016
+			
+		
+	
+	Antcliff, K. R., Moore, M. D., and Goodrich, K. H., "Silicon Valley as an Early Adopter for On-Demand Civil VTOL Operations," 16th AIAA Aviat. Technol. Integr. Oper. Conf, 2016, pp. 1-17.
+
+
+
+	
+		High Speed Mobility through On-Demand Aviation
+		
+			MarkDMoore
+		
+		
+			KennethHGoodrich
+		
+		10.2514/6.2013-4373
+	
+	
+		2013 Aviation Technology, Integration, and Operations Conference
+		
+			American Institute of Aeronautics and Astronautics
+			2013
+		
+	
+	Moore, M. D., Goodrich, K., Viken, J., Smith, J., Fredericks, B., Trani, T., Barraclough, J., German, B., and Patterson, M., "High Speed Mobility through On-Demand Aviation," 2013.
+
+
+
+	
+		Estimation of European Airspace Capacity from a Model of Controller Workload
+		
+			ArnabMajumdar
+		
+		
+			WashingtonOchieng
+		
+		
+			JohnPolak
+		
+		10.1017/s037346330200190x
+	
+	
+		Journal of Navigation
+		J. Navigation
+		0373-4633
+		1469-7785
+		
+			55
+			3
+			
+			2002
+			Cambridge University Press (CUP)
+		
+	
+	Majumdar, A., Ochieng, W., and Polak, J., "Estimation of European airspace capacity from a model of controller workload," Journal of Navigation, Vol. 55, No. 03, 2002, pp. 381-403.
+
+
+
+	
+		En-route sector capacity estimation methodologies: An international survey
+		
+			ArnabMajumdar
+		
+		
+			WashingtonYottoOchieng
+		
+		
+			JamesBentham
+		
+		
+			MartynRichards
+		
+		10.1016/j.jairtraman.2005.05.002
+	
+	
+		Journal of Air Transport Management
+		Journal of Air Transport Management
+		0969-6997
+		
+			11
+			6
+			
+			2005
+			Elsevier BV
+		
+	
+	Majumdar, A., Ochieng, W. Y., Bentham, J., and Richards, M., "En-route sector capacity estimation methodologies: An international survey," Journal of Air Transport Management, Vol. 11, No. 6, 2005, pp. 375-387.
+
+
+
+	
+		Airspace capacity estimation using flows and Weather-Impacted Traffic Index
+		
+			AlexanderKlein
+		
+		
+			LaraCook
+		
+		
+			BryanWood
+		
+		
+			DavidSimenauer
+		
+		10.1109/icnsurv.2008.4559188
+	
+	
+		2008 Integrated Communications, Navigation and Surveillance Conference
+		
+			IEEE
+			2008
+			
+		
+	
+	Klein, A., Cook, L., Wood, B., and Simenauer, D., "Airspace capacity estimation using flows and weather-impacted traffic index," 2008 Integrated Communications, Navigation and Surveillance Conference, IEEE, 2008, pp. 1-12.
+
+
+
+	
+		Capacity Estimation for Airspaces with Convective Weather Constraints
+		
+			JimmyKrozel
+		
+		
+			JosephMitchell
+		
+		
+			ValentinPolishchuk
+		
+		
+			JosephPrete
+		
+		10.2514/6.2007-6451
+	
+	
+		AIAA Guidance, Navigation and Control Conference and Exhibit
+		
+			American Institute of Aeronautics and Astronautics
+			2007
+		
+	
+	Krozel, J., Mitchell, J., Polishchuk, V., and Prete, J., "Airspace capacity estimation with convective weather constraints," AIAA Guidance, Navigation, and Control Conference, 2007.
+
+
+
+	
+		A decentralized control strategy for distributed air/ground traffic separation
+		
+			JimmyKrozel
+		
+		
+			MarkPeters
+		
+		
+			KarlBilimoria
+		
+		10.2514/6.2000-4062
+	
+	
+		AIAA Guidance, Navigation, and Control Conference and Exhibit
+		
+			American Institute of Aeronautics and Astronautics
+			2000
+		
+	
+	Krozel, J., Peters, M., and Bilimoria, K., "A decentralized control strategy for distributed air/ground traffic separation," 2000.
+
+
+
+	
+		System Performance Characteristics of Centralized and Decentralized Air Traffic Separation Strategies
+		
+			JimmyKrozel
+		
+		
+			MarkPeters
+		
+		
+			KarlDBilimoria
+		
+		
+			ChangkilLee
+		
+		
+			JosephS BMitchell
+		
+		10.2514/atcq.9.4.311
+	
+	
+		Air Traffic Control Quarterly
+		Air Traffic Control Quarterly
+		1064-3818
+		2472-5757
+		
+			9
+			4
+			
+			2001
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	Krozel, J., Peters, M., Bilimoria, K. D., Lee, C., and Mitchell, J., "System performance characteristics of centralized and decentralized air traffic separation strategies," Fourth USA/Europe air traffic management research and development seminar, 2001.
+
+
+
+	
+		Linking Community Visioning and Highway Capacity Planning
+		10.17226/14580
+	
+	
+		Highway research board proceedings
+		
+			National Academies Press
+			1935
+			1935
+		
+		
+			National Research Council (USA), Highway Research Board
+		
+	
+	Greenshields, B., Channing, W., Miller, H., et al., "A study of traffic capacity," Highway research board proceedings, Vol. 1935, National Research Council (USA), Highway Research Board, 1935.
+
+
+
+	
+		Existence of urban-scale macroscopic fundamental diagrams: Some experimental findings
+		
+			NikolasGeroliminis
+		
+		
+			CarlosFDaganzo
+		
+		10.1016/j.trb.2008.02.002
+	
+	
+		Transportation Research Part B: Methodological
+		Transportation Research Part B: Methodological
+		0191-2615
+		
+			42
+			9
+			
+			2008
+			Elsevier BV
+		
+	
+	Geroliminis, N., and Daganzo, C. F., "Existence of urban-scale macroscopic fundamental diagrams: Some experimental findings," Transportation Research Part B: Methodological, Vol. 42, No. 9, 2008, pp. 759-770.
+
+
+
+	
+		E-C149 of Transportation Research Circular, Traffic Flow Theory and Characteristics Committee
+		
+			RKühne
+		
+		
+			NGartner
+		
+	
+	
+		Years of the Fundamental Diagram for Traffic Flow Theory: Greenshields Symposium
+		
+			46
+			75
+		
+	
+	Kühne, R., and Gartner, N., "Years of the Fundamental Diagram for Traffic Flow Theory: Greenshields Symposium, vol. E-C149 of Transportation Research Circular, Traffic Flow Theory and Characteristics Committee," Transportation Research Board of the National Academies, Vol. 46, 75.
+
+
+
+	
+		A ground-delay-based approach to reduce impedance-based airspace complexity
+		
+			VishwanathBulusu
+		
+		
+			RSengupta
+		
+		
+			ZLiu
+		
+		10.2514/6.2021-2340
+	
+	
+		AIAA AVIATION 2021 FORUM
+		
+			American Institute of Aeronautics and Astronautics
+			2016
+		
+	
+	Bulusu, V., Sengupta, R., and Liu, Z., "Unmanned Aviation: To Be Free or Not To Be Free? A Complexity Based Approach," 7th International Conference on Research in Air Transportation, 2016.
+
+
+
+	
+		A review of conflict detection and resolution modeling methods
+		
+			JKKuchar
+		
+		
+			LCYang
+		
+		10.1109/6979.898217
+	
+	
+		IEEE Transactions on Intelligent Transportation Systems
+		IEEE Trans. Intell. Transport. Syst.
+		1524-9050
+		
+			1
+			4
+			
+			2000
+			Institute of Electrical and Electronics Engineers (IEEE)
+		
+	
+	Kuchar, J. K., and Yang, L. C., "A review of conflict detection and resolution modeling methods," IEEE Transactions on intelligent transportation systems, Vol. 1, No. 4, 2000, pp. 179-189.
+
+
+
+	
+		Applications of a Macroscopic Model for En Route Sector Capacity
+		
+			JerryWelch
+		
+		
+			JohnAndrews
+		
+		
+			BrianMartin
+		
+		
+			EricShank
+		
+		10.2514/6.2008-7221
+	
+	
+		AIAA Guidance, Navigation and Control Conference and Exhibit
+		Barcelona, Spain
+		
+			American Institute of Aeronautics and Astronautics
+			2007
+			138
+		
+	
+	Welch, J. D., Andrews, J. W., Martin, B. D., and Sridhar, B., "Macroscopic workload model for estimating en route sector capacity," Proc. of 7th USA/Europe ATM Research and Development Seminar, Barcelona, Spain, 2007, p. 138.
+
+
+
+	
+		AIR TRAFFIC MANAGEMENT
+		
+			IVLaudeman
+		
+		
+			SShelden
+		
+		
+			RBranstrom
+		
+		
+			CBrasil
+		
+		10.1201/9781482267952-27
+	
+	
+		Contemporary Ergonomics 1998
+		
+			CRC Press
+			1998
+			
+		
+	
+	Laudeman, I. V., Shelden, S., Branstrom, R., and Brasil, C., "Dynamic density: An air traffic management metric," 1998.
+
+
+
+	
+		Airspace complexity and its application in air traffic management
+		
+			BSridhar
+		
+		
+			KSSheth
+		
+		
+			SGrabbe
+		
+	
+	
+		Europe Air Traffic Management R&D Seminar
+		
+			
+			1998
+		
+	
+	2nd USA/
+	Sridhar, B., Sheth, K. S., and Grabbe, S., "Airspace complexity and its application in air traffic management," 2nd USA/Europe Air Traffic Management R&D Seminar, 1998, pp. 1-6.
+
+
+
+	
+		The Complexity Construct in Air Traffic Control: A Review and Synthesis of the Literature
+		
+			RHMogford
+		
+		
+			JGuttman
+		
+		
+			SMorrow
+		
+		
+			PKopardekar
+		
+		
+			1995
+		
+	
+	Tech. rep., DTIC Document
+	Mogford, R. H., Guttman, J., Morrow, S., and Kopardekar, P., "The Complexity Construct in Air Traffic Control: A Review and Synthesis of the Literature." Tech. rep., DTIC Document, 1995.
+
+
+
+	
+		Practice for Application of Federal Aviation Administration (FAA) Federal Aviation Regulations Part 21 Requirements to Unmanned Aircraft Systems (UAS)
+		
+			PKopardekar
+		
+		10.1520/f2505-07
+	
+	
+		Federal Aviation Administration
+		
+			ASTM International
+			2000
+		
+	
+	Dynamic density: A review of proposed variables
+	Kopardekar, P., "Dynamic density: A review of proposed variables," FAA internal document. overall conclusions and recommendations, Federal Aviation Administration, 2000.
+
+
+
+	
+		Airspace complexity measurement: An air traffic control simulation analysis
+		
+			PHKopardekar
+		
+		
+			ASchwartz
+		
+		
+			SMagyarits
+		
+		
+			JRhodes
+		
+	
+	
+		International Journal of Industrial Engineering: Theory, Applications and Practice
+		
+			16
+			1
+			
+			2009
+		
+	
+	Kopardekar, P. H., Schwartz, A., Magyarits, S., and Rhodes, J., "Airspace complexity measurement: An air traffic control simulation analysis," International Journal of Industrial Engineering: Theory, Applications and Practice, Vol. 16, No. 1, 2009, pp. 61-70.
+
+
+
+	
+		Air traffic predictability framework – Development, performance evaluation and application
+		
+			GonzaloTobaruela
+		
+		
+			PeterFransen
+		
+		
+			WolfgangSchuster
+		
+		
+			WashingtonYOchieng
+		
+		
+			ArnabMajumdar
+		
+		10.1016/j.jairtraman.2014.04.001
+	
+	
+		Journal of Air Transport Management
+		Journal of Air Transport Management
+		0969-6997
+		
+			39
+			
+			2012
+			Elsevier BV
+		
+	
+	Tobaruela, G., Majumdar, A., and Ochieng, W. Y., "Identifying Airspace Capacity Factors in the Air Traffic Management System," Proceedings of the 2nd International Conference on Application and Theory of Automation in Command and Control Systems, 2012, pp. 219-224.
+
+
+
+	
+		Assessing the capacity of novel ATM systems
+		
+			AJHudgell
+		
+		
+			RGingell
+		
+	
+	
+		4th USA/Europe Air Traffic Management R&D Seminar
+		
+			2001
+		
+	
+	Hudgell, A. J., and Gingell, R., "Assessing the capacity of novel ATM systems," 4th USA/Europe Air Traffic Management R&D Seminar, 2001.
+
+
+
+	
+		Concepts of Airspace Structures and System Analysis for UAS Traffic flows for Urban Areas
+		
+			Dae-SungJang
+		
+		
+			CoreyAIppolito
+		
+		
+			ShankarSankararaman
+		
+		
+			VahramStepanyan
+		
+		10.2514/6.2017-0449
+	
+	
+		AIAA Information Systems-AIAA Infotech @ Aerospace
+		
+			American Institute of Aeronautics and Astronautics
+			2017
+			449
+		
+	
+	Jang, D.-S., Ippolito, C., Sankararaman, S., and Stepanyan, V., "Concepts of Airspace Structures and System Analysis for UAS Traffic flows for Urban Areas," AIAA Information Systems-AIAA Infotech@ Aerospace, 2017, p. 0449.
+
+
+
+	
+		Distributed Air/Ground Traffic Management for En Route Flight Operations
+		
+			StevenMGreen
+		
+		
+			KarlDBilimoria
+		
+		
+			MarkGBallin
+		
+		10.2514/atcq.9.4.259
+	
+	
+		Air Traffic Control Quarterly
+		Air Traffic Control Quarterly
+		1064-3818
+		2472-5757
+		
+			9
+			4
+			
+			2001
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	Green, S. M., Bilimoria, K., and Ballin, M. G., "Distributed air/ground traffic management for en route flight operations," Air Traffic Control Quarterly, Vol. 9, No. 4, 2001.
+
+
+
+	
+		Performance Evaluation of Airborne Separation Assurance for Free Flight
+		
+			KarlDBilimoria
+		
+		
+			ShonRGrabbe
+		
+		
+			KapilSSheth
+		
+		
+			HildaQLee
+		
+		10.2514/atcq.11.2.85
+	
+	
+		Air Traffic Control Quarterly
+		Air Traffic Control Quarterly
+		1064-3818
+		2472-5757
+		
+			11
+			2
+			
+			2003
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	Bilimoria, K. D., Sheth, K. S., Lee, H. Q., and Grabbe, S. R., "Performance evaluation of airborne separation assurance for free flight," Air Traffic Control Quarterly, Vol. 11, No. 2, 2003, pp. 85-102.
+
+
+
+	
+		Trajectory Tracking Control of a Drone-Guided Hose System for Fluid Delivery
+		
+			ABarr
+		
+		
+			GBensinger
+		
+		10.2514/6.2021-1003.vid
+	
+	
+		Wall Street Journal
+		
+			Aug. 29, 2014
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	Barr, A., and Bensinger, G., "Google Is Testing Delivery Drone System," Wall Street Journal, Aug. 29, 2014.
+
+
+
+	
+		Initial Study of An Effective Fast-time Simulation Platform for Unmanned Aircraft System Traffic Management
+		
+			MinXue
+		
+		
+			JosephRios
+		
+		10.2514/6.2017-3073
+	
+	
+		17th AIAA Aviation Technology, Integration, and Operations Conference
+		
+			American Institute of Aeronautics and Astronautics
+			2017
+			3073
+		
+	
+	Xue, M., and Rios, J., "Initial Study of An Effective Fast-time Simulation Platform for Unmanned Aircraft System Traffic Management," 17th AIAA Aviation Technology, Integration, and Operations Conference, 2017, p. 3073.
+
+
+
+	
+		Capacity Estimation for Low Altitude Airspace
+		
+			VishwanathBulusu
+		
+		
+			ValentinPolishchuk
+		
+		
+			RajaSengupta
+		
+		
+			LeonidSedov
+		
+		10.2514/6.2017-4266
+	
+	
+		17th AIAA Aviation Technology, Integration, and Operations Conference
+		
+			American Institute of Aeronautics and Astronautics
+			2017
+			4266
+		
+	
+	Bulusu, V., Polishchuk, V., Sengupta, R., and Sedov, L., "Capacity estimation for low altitude airspace," 17th AIAA Aviation Technology, Integration, and Operations Conference, 2017, p. 4266.
+
+
+
+	
+		http://ljournal.ru/wp-content/uploads/2016/08/d-2016-154.pdf
+		
+			Mitre
+		
+		10.18411/d-2016-154
+		
+	
+	
+		SUAS gaps being worked by SARP
+		
+			ljournal
+			2016
+		
+	
+	UTM Convention
+	MITRE, "SUAS gaps being worked by SARP," UTM Convention, 2016. URL http://www.utm2016.com/Uploads/ Presentations/UTMRDPanelcompress.pdf, researc Panel Presentation.
+
+
+
+	
+		Multi-Rotor Aircraft Collision Avoidance using Partially Observable Markov Decision Processes
+		
+			EricRMueller
+		
+		
+			MykelKochenderfer
+		
+		10.2514/6.2016-3673
+	
+	
+		AIAA Modeling and Simulation Technologies Conference
+		
+			American Institute of Aeronautics and Astronautics
+			2016
+		
+		
+			Stanford University
+		
+	
+	Ph.D. thesis
+	Mueller, E. R., and Kochenderfer, M. J., "Multi-rotor aircraft collision avoidance using partially observable Markov decision processes," Ph.D. thesis, Stanford University, 2016.
+
+
+
+	
+		ICAROUS: Integrated configurable algorithms for reliable operations of unmanned systems
+		
+			MariaConsiglio
+		
+		
+			CesarMunoz
+		
+		
+			GeorgeHagen
+		
+		
+			AnthonyNarkawicz
+		
+		
+			SweeBalachandran
+		
+		10.1109/dasc.2016.7778033
+		IEEE/AIAA 35th
+	
+	
+		2016 IEEE/AIAA 35th Digital Avionics Systems Conference (DASC)
+		
+			IEEE
+			2016. 2016
+			
+		
+	
+	Consiglio, M., Muñoz, C., Hagen, G., Narkawicz, A., and Balachandran, S., "ICAROUS: Integrated configurable algorithms for reliable operations of unmanned systems," Digital Avionics Systems Conference (DASC), 2016 IEEE/AIAA 35th, IEEE, 2016, pp. 1-5.
+
+
+
+	
+		DAIDALUS: Detect and Avoid Alerting Logic for Unmanned Systems
+		
+			CesarMunoz
+		
+		
+			AnthonyNarkawicz
+		
+		
+			GeorgeHagen
+		
+		
+			JasonUpchurch
+		
+		
+			AaronDutle
+		
+		
+			MariaConsiglio
+		
+		
+			JamesChamberlain
+		
+		10.1109/dasc.2015.7311421
+	
+	
+		2015 IEEE/AIAA 34th Digital Avionics Systems Conference (DASC)
+		
+			IEEE
+			2015. 2015
+			
+		
+	
+	Muñoz, C., Narkawicz, A., Hagen, G., Upchurch, J., Dutle, A., Consiglio, M., and Chamberlain, J., "DAIDALUS: detect and avoid alerting logic for unmanned systems," Digital Avionics Systems Conference (DASC), 2015 IEEE/AIAA 34th, IEEE, 2015, pp. 5A1-1.
+
+
+
+	
+		Unmanned Aircraft System (UAS) standards development: RTCA SC-228 status
+		
+			StephenVan Trees
+		
+		10.1109/icnsurv.2015.7121363
+		
+	
+	
+		2015 Integrated Communication, Navigation and Surveillance Conference (ICNS)
+		
+			IEEE
+			
+			
+		
+	
+	"RTCA SC-228 Minimum Operational Performance Standards for Unmanned Aircraft Systems," https://www.rtca.org/ content/sc-228, Accessed: 2018-05-10.
+
+
+				
+			
+		
+	

file107.txt

@@ -0,0 +1,208 @@
+
+	
+	
+		
+I. IntroductionNASA, in collaboration with government and industry partners, is developing concepts for Urban Air Mobility (UAM) vehicles and technologies to support UAM vehicle operations across a network of takeoff and landing areas (TOLAs) in metropolitan regions.These UAM operations of the future will need FAA approval.Hence, the simplest approach to begin operations might be to start with already-existing approved routes, especially for flying into major airports.Can UAM vehicles fly these routes with minimal impact on present-day commercial air traffic (hereafter referred to as conventional traffic)?This paper presents a preliminary modeling and analysis of interactions between proposed UAM operations and conventional traffic, if UAM operations were restricted to FAA-approved helicopter routes and altitude ceilings.UAM operations in Dallas/Fort-Worth (DFW) terminal airspace is chosen for the study.The goal of this work is to assess the extent to which proposed initial UAM operations may trigger Traffic alert and Collision Avoidance System (TCAS) resolution advisories (RA) aboard conventional aircraft in the DFW terminal area.A range of operational scenarios at DFW is evaluated with combinations of UAM vehicle route, speed, altitude, and direction along the DFW "spine route."The analysis is done for both South flow and North flow configurations of DFW.First, results are obtained under the assumption that UAM vehicles fly their routes precisely (i.e.no uncertainty).Then, the impact of altitude uncertainty on these results is also evaluated.Different route alternatives are evaluated between DFW and Frisco, Texas in this paper.DFW-Frisco operation was selected as a test case based upon recommendations derived from traffic demand studies from potential UAM operators in the region.The focus of research presented here is only the impact of operations for this test case on triggering TCAS RAs on conventional aircraft.Interaction with Air Traffic Control(ATC) and other route alternatives and procedures are not studied.Associated work by Verma et.al. [1] explored potential routes and procedures in a Human-In-The-Loop (HITL) experiment expanding the above study region to include Dallas Lovefield (DAL) and Addison (ADS) airspace and Dallas downtown.The rest of this paper is divided into five sections.Section II, Traffic alert and Collision Avoidance System (TCAS), provides a description of TCAS and how its logic is used in this study.Section III, DFW Airport Terminal Area Traffic, describes DFW runways and operational configurations.Section IV, Study Approach, provides a breakdown of modeling assumptions, UAM operational route scenarios, selection of study days and description of simulation set up.Section V, Results, first summarizes the results of the geometric analysis of the routes, followed by their verification using simulations without trajectory uncertainties.It also presents the results with altitude uncertainty.Section VI, Conclusions, discusses the overall findings and recommendations for future studies.
+II. Traffic alert and Collision Avoidance System (TCAS)In conventional operations, TCAS is the last layer to prevent mid-air collisions between aircraft other than the pilot.The design of the TCAS logic is a trade-off between providing necessary safe separation between aircraft and preventing unnecessary advisories.Unnecessary advisories reduce trust in the system and may increase crew and controller workload to unacceptable levels.The system monitors the amount of horizontal and vertical separation and uses the rates of change in horizontal and vertical separation to predict the closest point of approach between the ownship and surrounding aircraft.A Traffic Advisory (TA) or a Resolution Advisory (RA) is issued based on thresholds for estimated time for an intruder aircraft to enter a protected volume of airspace around a TCAS equipped aircraft (τ mod ) and time to co-altitude (τ vert ) [2].The boundary of the protected volume is defined by a slant range distance called DMOD.During a TA, on a conventional aircraft, TCAS displays the intruder aircraft and notifies the pilot about a potential conflict through visual and audio alerts.The pilot is expected to respond to a TA by (1) establishing visual contact with the intruder and other aircraft in the vicinity and (2) maintaining current assigned clearance.During an RA, TCAS issues maneuvers to the pilot to increase or maintain current vertical separation.The pilot is expected to immediately respond to the indicated maneuvers unless doing so would unduly jeopardize the safe operation of the flight.When both aircraft are equipped with TCAS II, aircraft coordinate their RA's through Mode S datalink to ensure that complementary RAs are selected.Due to the assumption that TCAS II is available only on conventional aircraft, RA coordination functionality will not be discussed in this work.To balance the tradeoff between necessary protection and unnecessary advisories, TCAS uses an altitude-based Sensitivity Level (SL), which controls the tau (time) thresholds for TA and RA issuance, and therefore the dimensions of the protected airspace around each TCAS-equipped aircraft (Table 1).DMOD and ZTHR are the slant range and vertical separation threshold of the protected airspace as per the TCAS II manual [2].The higher the SL, the larger the amount of protected airspace and the longer the alerting thresholds.While an aircraft is in close proximity to ground, the SL of TCAS alert and avoidance parameters depends on the altitude of the ownship aircraft above ground level (AGL).TCAS does not provide RAs below 1000ft AGL (SL=2).Between 1000-2350 ft AGL (SL=3), TCAS issues RAs, if both τ mod and τ vert are less than 15 seconds threshold, or τ mod is less than 15 seconds and current altitude separation is less than 600 feet (vertical separation threshold).
+III. DFW Airport Terminal Area TrafficThe study described in this paper uses nominal operation assumptions and procedures for DFW airport as described in this section.
+Fig. 1 DFW Runway Map
+A. DFW RunwaysThe DFW airport has seven physical runways shown in figure 1.These runways are operated in the South-flow and North-flow configurations.The runways in South-flow configuration are designated as, 13L, 13R, 17L, 17C, 17R, 18L and 18R.The runways in the North-flow configuration are designated as, 31L, 31R, 35L, 35C, 35R, 36L and 36R.These designations indicate the runway heading with respect to north.The two inner runways 18L/36R and 17R/35L are primarily used for departures.Runways 18R/36L and 17C/35C are primarily used for arrivals [3].Adjacent primary departure and arrival runways are about 0.2 miles (1200 ft) apart.In North flow operations, runway 35R is used for arrivals as the amount of traffic increases but it is less desirable being shorter and far from terminals and its use requires aircraft to cross runways.As a rule of thumb, easterly arrivals use runway 17C/35C and westerly arrivals use runway 18R/36L.Considering the geometry of the runways, arrival and departure procedures at DFW and the expected nominal trajectories of UAM vehicles along helicopter routes (see figure 2), interactions between commercial aircraft and UAM are most likely to happen between aircraft arriving/departing on runways (18R,17C)/(18L,17R) in South flow and arriving/departing from runways (36L,35C)/(36R/35L) in North Flow.
+B. DeparturesNinety five percent of all departures from DFW are jets using RNAV routes [4].Departures typically are provided temporary level-off altitudes of 10,000ft, if their track crosses under arriving aircraft.After crossing they are cleared to climb to their cruise altitude.The average duration of a 10,000 ft level off is about 1.7 minutes, traveling a distance of 9.4 nmi based on an assumed airspeed of 331 knots [4].
+C. ArrivalsThe arrivals use Runways 13R, 18R, 17C and 17L in South flow and 31R, 35C, 35R, and 36L in North flow configurations.All approaches follow a 3 degree glide slope procedure [5].Hence, aircraft arriving with a stable approach are below 1000 ft AGL, 3.16 nm before the touchdown zone.
+IV. Study ApproachIn this study, two routes for UAM flights operating between the city of Frisco, Texas and DFW airport were analyzed.Frisco was selected based on recommendations derived from traffic studies by potential UAM operators in the region.UAM-conventional aircraft interactions were evaluated from a TCAS perspective.UAM-UAM interactions were not studied as they are out of scope for this work.
+A. Assumptions and Test CasesThe following assumptions were made:-All UAM vehicles in the simulation were modeled as the same aircraft type.-All UAM fly along published helicopter routes.-UAM trajectories are completely deterministic in the simulation for the first set of results and identification of sensitive areas.-Only altitude uncertainty is considered for the second set of results for the identified sensitive areas.-The altitude errors are assumed to be normally distributed with zero mean and varying standard deviations.-Conventional aircraft are modeled with TCAS II version 7.1.-UAM vehicles are not modeled with the above TCAS system but do provide the state information required by the TCAS system on board the conventional aircraft.-Conventional aircraft adhere to published area navigation (RNAV) routes.Same aircraft type assumption has low impact on the relevance of this study, as we only need a representative performance model.In future, a wider range of vehicles could be summoned based on trip length and traffic demand.However, manufacturing and maintenance costs alone will likely push towards a preferred UAM vehicle.Existing helicopter routes are a useful starting point as they are already designed for vehicles that operate at low altitudes with Vertical Take-Off and Landing (VTOL) capability.Eventually, other routes and procedures can be explored (for example see [1]).Since vertical separation was identified as a major factor that could potentially trigger RAs, from the first set of results, altitude uncertainty was a good first candidate to study.Four UAM operational route scenarios were evaluated: Nominal days were characterized by moderate meteorological (temperature between 85 o F -95 o F and low precipitation) and traffic flow conditions.The off-nominal days had maximum temperatures close to 100 o F , clear skies, and minimal weather impacted operations, allowing for the highest traffic flow with least impact on airport operations.On these days, owing to higher temperatures, conventional aircraft may not have been able to climb as quickly as they could on nominal days, which was expected to cause closer encounters and potentially more TCAS RAs.Furthermore, August 7 had the highest conventional traffic in North Flow configuration for the entire year and therefore was also used for the altitude uncertainty study.These test cases together account for different routes, flow directions of UAM traffic between Frisco and DFW and flow variations of conventional traffic.Additionally, in the vicinity of the airport, all UAM flying into DFW were modeled to cruise at 1000 ft MSL, and all UAM flying out were modeled to cruise at 900 ft MSL.This separates UAM vehicles flying into DFW by 100 ft from UAM vehicles flying out.The cruise airspeed of all UAM vehicles was modeled at 130 knots.Although it is important to assess whether this UAM-UAM vertical separation is sufficient, it is beyond the scope of this study.However, it ensures that UAM vehicles taking off and flying against arriving conventional traffic, are at a higher vertical separation, compared to UAM vehicles flying in the same direction.
+C. Analysis ApproachThe analysis was accomplished in three stages.First, TCAS sensitivity level in the simulator was higher than a real system.The value of sensitivity parameter DMOD was 0.66 nm as per UAS in NAS project [7], which would be highly conservative for the near surface operations in this study.τ mod measures the time it takes two aircraft to come closer than DMOD distance.Second, scenarios that produce RAs from above were filtered using geometric aircraft configuration data from the simulator, with the thresholds for a SL 3 TCAS operation, even below 1000 ft AGL (which would technically be SL 2 -No RAs).Third, the correct TCAS SL was used, based on actual AGL altitude of both aircraft consistent with what would be used by a real TCAS II system on a conventional aircraft.If there are no RAs in a higher analysis stage, there can't be any in the next stage.The first two stages are useful for identifying the sensitive areas in the system.However, the results presented in the next section are based on the third and least conservative stage, which is how TCAS would behave in a real scenario.As an example, a sample encounter scenario is shown in figure 3 with the associated encounter parameters.Recall that if τ mod is less than the threshold and either τ vert is less than the threshold or the current vertical separation is less than the threshold, then a TCAS RA is issued.TCAS τ mod (Sim) is evaluated using a DMOD of 0.66 nm as used in UAS-NAS project [7]; in this example, that results in a value of 1 sec, which is below all τ mod thresholds.From the first stage of analysis, the simulation will flag an RA because both τ mod and τ vert are less than the threshold of 15 sec.TCAS τ mod (SL) is derived based on SL-specific DMOD.For SL 3, DMOD is 0.2 nm.At the second stage of analysis, an RA is also issued because both times are below the threshold of 15 sec (SL 3).However, since DMOD is smaller, the projected time for slant range to go below it is greater.Finally, in a real scenario, since the commercial aircraft is below 1000 ft AGL, no RA shall be issued.Results derived following this analysis are presented in the next section.
+V. ResultsOutside the DFW surface-to-2500 MSL class B airspace, under the assumption that conventional aircraft adhere to published RNAV routes, they are far above the 1000-ft MSL ceiling/cruise altitude of the UAM aircraft and therefore well separated (> 1000 ft) in altitude.Hence, the analysis here is focused only on the interactions near DFW.The region of interest along with the sensitive areas is shown in figure 4. The thick blue lines indicate the vertical planes between which the arriving conventional traffic is below 1000 ft AGL.They are 3.16 nm from the touchdown zone of their respective runways.This distance was computed from the 3-degree approach glide slope described in section III.C.All analysis for deterministic UAM trajectories is primarily geometric and hence applies to both nominal and off-nominal days.However, those days were simulated as a secondary confirmation to ensure that no corner cases were neglected.The elevation of DFW airport is 607 ft MSL.TCAS will not issue RAs when the conventional aircraft is below 1607 ft MSL (i.e.<1000 ft MSL).It will operate at SL=3 for conventional aircraft between 1607-2957 ft MSL.The UAM vehicles simulated in this study are below 1607 ft MSL during their entire trajectory portion under consideration, i.e. in the vicinity of DFW.Thus, in principle, RAs can technically be issued only at SL=3 in the worst case.If the conventional aircraft is above 2957 MSL (SL=4), a TCAS RA will not be triggered with a UAM.In order for the conventional aircraft at 2957 MSL to become co-altitude with the UAM vehicle at 1607 MSL within the τ vert threshold of 20 seconds, the conventional aircraft would have to descend faster than 4000 ft/min.This is much higher than a typical descent rate of 800 ft/min that such an aircraft would use on a 3-degree final approach glide slope.Even before the aircraft has intercepted the final glide slope, it would descend at a much lower final descent rate [9].Therefore, the modeling assumptions for this study prevent the time and altitude separation thresholds from ever being violated, if the conventional aircraft is above 2957 MSL near DFW.
+A. Departures AnalysisDeparting conventional aircraft primarily use runways 18L and 17R in South Flow (35L and 36R in North Flow) (Figure 1) and are above 1000 ft AGL 20 secs after departure, based on a climb rate of roughly 50 ft/sec [4].Since the departure runways have a minimum separation of 0.44 nm from the UAM flight paths in their take-off zones, departing conventional aircraft are well separated horizontally from the UAM aircraft.After take-off, the conventional aircraft climb at rates between 2000-3000 ft/min, much faster than the 500 ft/min ascent/descent rate assumed for UAM flights.They are above the incoming UAM flight altitudes in less than 10 sec.Hence, they are always diverging and well separated by 1000ft AGL, the altitude where the TCAS system would start producing any RAs.Therefore, under the assumptions in this study, departing conventional aircraft will not produce RAs, even with UAM trajectory uncertainties.
+B. Arrivals AnalysisResults in this section were determined assuming no UAM trajectory uncertainties.In every case where τ mod was violated and either the vertical separation threshold or τ vert was violated, the conventional aircraft was always below 1000 ft AGL, where TCAS RAs are inhibted.Therefore, under the assumptions of this study, arriving conventional aircraft will also not produce any RAs, if the UAM trajectories are deterministic.North of DFW (figure 5), UAM flights departing against the direction of arriving conventional traffic are separated horizontally by a minimum of 0.44 nm (0.64 nm from the outer runway -0.2 nm separation between the adjacent parallel runways) and primarily follow parallel trajectories.For interaction scenarios south of the blue line, the aircraft are therefore well separated horizontally.Furthermore, conventional aircraft are below 1000 ft AGL with TCAS operating at SL 2 as described at the start of this section.Hence, they issue no RAs.North of the blue line in the same figure, the conventional aircraft are above 1000 ft AGL and have TCAS operating at SL 3.Even where τ mod goes below thresholds (e.g.where the UAM turns right and flies below conventional aircraft arrival path), the vertical separation is greater than 600 ft and the time to co-altitude is greater than the SL 3 threshold of 15 seconds to trigger any RAs.Such a sample encounter where τ mod goes below thresholds is shown in figure 6.The evolution of vertical separation (solid blue line) is plotted on the left axis.At SL 3, the vertical separation threshold, ZTHR is 600 ft (dashed blue line).τ mod (solid red line) and τ vert (dashed red line) are plotted on the right axis.The SL 3 time thresholds are violated below Tau = 15 sec, i.e. when τ mod and τ vert are between the dotted red lines.When τ mod is between 0 and 15 sec, neither the vertical separation threshold nor the τ vert threshold is violated.Hence, this encounter will not trigger an RA. Figure 7 illustrates this is true even at a high descent rate.Assume that the conventional aircraft is at AGL altitude A C in feet (>1000 ft) and the UAM is cruising in level flight at AGL altitude A U in feet.Any aircraft above 1000 ft, under normal operation, should not descend at a rate (ft/min) greater than its AGL altitude, i.e. maximum descent rate or maximum vertical closure rate, V ZC = A C in ft/min.Hence time to co-altitude, τ vert = separation/(closure rate per min/60) = separation/(closure rate in sec) = ((A C -A U )/A C )*60 = (1 -A U /A C )*60 seconds.For maximum A U = 400 ft and minimum A C > 1000 ft, minimum τ vert > 36 sec.Hence time to co-altitude is always greater than 36 seconds, which is more than double the time threshold for SL 3. The vertical separation threshold of 600 ft is also not violated because A C -A U > 600 ft.South of DFW (figure 8), the UAM flight route intersects with conventional aircraft runway approaches after the conventional aircraft is already below 1000 ft AGL (north of blue line) with TCAS operating at SL 2. Hence, no RAs will be issued.South of blue line (region not shown in figure 8), the conventional aircraft are separated by at least 0.5 nautical miles horizontally and more than 600 ft vertically.Following the same explanation as the SL 3 situation north of DFW (see figure 7), the vertical violation criteria is never satisfied.Hence, no RAs are issued.These results cover all combinations of UAM vehicle routes, altitudes and directions studied.They also account for variations in UAM vehicle cruise speeds.When conventional aircraft are below 1000 ft AGL, there will be no RAs irrespective of the closure rates (slant and vertical).When they are above 1000 ft AGL, the lack of RAs is due to vertical closure rates and adequate vertical separation.The entire analysis is therefore agnostic to UAM cruise speeds.
+C. Altitude Uncertainty AnalysisAlthough τ vert was never violated when the conventional aircraft were above 1000 ft AGL, it is noteworthy that the vertical separation threshold (600 ft) was very close to violation at the intersection of the UAM route and the conventional aircraft arrival path south of DFW, when the UAM vehicles fly into DFW (at 1000 ft MSL).A sample encounter from this sensitive area is shown in figure 9.Although there is a time when τ mod is less than 15 sec and the vertical separation is less than 600 feet, the conventional aircraft is already below 1000 ft AGL at that time.Hence, TCAS will ignore the UAM vehicle and not trigger an RA.However, if the UAM vehicle had a vertical position error of even 10 ft, it would have triggered an RA in this scenario.Hence, the natural next step is to explore the impact of trajectory uncertainties.In this paper, only altitude uncertainty was explored.Vertical GPS errors with available technologies today are less than 10 m (≈30 ft) even in the worst case.This was simulated by introducing errors in the altitude of UAM vehicles.The errors were assumed to be normally distributed with zero mean and standard deviation (StD) varying from 5 ft to 30 ft.The maximum error allowed was thrice of the chosen StD.Hence, even though the maximum errors in reality shouldn't exceed 30 ft, errors were simulated up to 90 ft.It is noteworthy that UAM vehicles departing DFW were modeled to cruise at 900 ft MSL (293 ft AGL) and hence, even with a maximum error of 90ft, they wouldn't trigger RAs at the sensitive area identified above.Hence, analysis Fig. 9 A sample encounter South of DFW (north of blue line in figure 4).Note: Actual onboard TCAS II RA logic will ignore diverging aircraft, hence τ mod and τ vert will only be computed when they are non-negative.was only done for Scenario 4, where UAM vehicles fly from Frisco to DFW and enter DFW from the South, passing below the conventional aircraft arrival paths at 1000 ft MSL (393 ft AGL).Furthermore, encounters were simulated with the conventional aircraft track data for August 7, 2017.That day had the highest North Flow traffic of the year and hence, maximum potential encounters.Figure 11 top row shows the probability of triggering an RA for a UAM vehicle departing Frisco at the times shown on the horizontal axis, as the altitude error StD varies from 5 ft to 30 ft.The probability variation followed the conventional traffic demand change through the day.During lean times (before 9a and after 7:30p), even a 30 ft error StD did not produce more than forty percent risk of triggering an RA.During the peak conventional traffic rush between
+Fig. 10 Risk of triggering RAs on a typical day of operation -Noon-3p9a and 7:30p, it is observed that even a 5-ft error StD produced over twenty percent risk of triggering RAs at certain times.This risk increased to over forty percent, when the error StD was increased to 10 ft or more.A potential solution to this problem could be that the UAM fly into DFW at a slightly lower altitude below the arrival paths.To justify this recommendation, the above analysis was repeated by lowering the UAM vehicle mean altitude by 5 ft and 10 ft, respectively.It was observed that, in general, lowering the mean altitude by twice the allowable error StD substantially reduced the risk of triggering RAs.As an example, in figure 11, third row, lowering the UAM vehicle mean altitude by 10 ft, reduced the chance of triggering an RA to under ten percent throughout the day, even with a 10-ft error StD (maximum error 30 ft).For clarity, figure 10 shows a zoomed version of the variation from noon to 3p.
+VI. ConclusionsThe study found that using a basic model of UAM performance, UAM vehicles simulated to operate from Frisco to DFW utilizing existing helicopter routes as shown in figure 2, triggered no RAs on conventional aircraft.These results were obtained under the assumption of deterministic UAM vehicle trajectories, i.e. zero error in observed and true position of the UAM vehicles.Meteorological conditions, such as wind, which could affect this accuracy, were ignored.These results suggest a very high navigational performance requirement on UAM, if operations were to be enabled with high UAM trajectory determinism.From the altitude uncertainty study, it was also observed that the above performance requirements could be slightly relaxed by operating the UAM vehicles at or below 990 ft MSL (383 ft AGL), if they can adhere to a maximum altitude error of 15 ft from their trajectory.Furthermore, the primary reason for the lack of RAs is that conventional aircraft are either already, by procedure, well separated horizontally and vertically; or are below 1000 ft AGL otherwise (which suppresses RAs).Even though this is true for DFW based on its particular runway configuration, a similar analysis can be performed at any other airport to determine the sensitive regions for TCAS RA alerts.It should be noted that TCAS II ignores intruders below 360 ft AGL.This means any UAM flights below 967 ft MSL (around DFW) are automatically ignored by current TCAS operation criteria.This can be interpreted in two ways.In the short term, UAM flights around any airport in the country can be kept below 360 ft AGL to enable early operations, if necessary, without triggering RAs.In the long term, this could create potential issues with high density of near-ground traffic and therefore, might necessitate an update to the TCAS logic to account for the same.Therefore, this is also an important area for further investigation.1 )1Frisco, Texas to DFW entering DFW from the North (Figure 2, Scenario 1); 2) DFW to Frisco exiting DFW towards the South (Figure 2, Scenario 2); 3) DFW to Frisco exiting DFW towards the North (Figure 2, Scenario 3); and 4) Frisco to DFW entering DFW from the South (Figure 2, Scenario 4).Each of these scenarios was studied for every DFW runway operation condition listed below for a total of 16 simulation test cases (4 UAM operation scenarios X 4 runway operation conditions).Conventional traffic data for each runway operation condition was derived from the dates mentioned in parenthesis.-Nominal Day in South Flow (June 03, 2017).-Nominal Day in North Flow (November 11, 2017).-Off-Nominal Day in South Flow (July 20, 2017).-Off-Nominal Day in North Flow (August 7, 2017).
+Fig. 22Fig. 2 Flight routes between Frisco, Texas and DFW airports overlaid on FAA Sectional Charts [6] (Clockwise from top left: Scenario 1, Scenario 2, Scenario 4 and Scenario 3)B.Simulation Platform and Software ComponentsStudy simulations used the SaaControl fast-time simulation software developed by NASA as a testing tool for modeling Detect-And-Avoid (DAA) capability of Unmanned Aircraft Systems (UAS-NAS project[7]).Conflict avoidance algorithms (TCAS as one), surveillance and atmospheric models, and pilot response models have been integrated into its core module.SaaControl is capable of running faster-than-real-time NAS-wide simulations.In this study, it detects potential conflicts from raw input traffic data: flight plans for UAM and track data files for conventional aircraft.For the TCAS logic, FAA-supplied TCAS II version 7.1 software was used with a software wrapper developed by NASA, to integrate it into NASA simulation platforms.The wrapper packages TCAS II into a JAVA library callable by clients.It takes aircraft states from a client, calls TCAS II, and returns TCAS responses to the client.
+Fig. 3 A3Fig. 3 A Sample Scenario (CPA is Closest Point of Approach)
+Fig. 44Fig. 4 Region of Interest [8].UAM flight routes in green (North approach/departure -fluorescent green, South approach/departure -dark green), conventional aircraft paths in orange and red and the sensitive areas for analysis, North (figure 5) and South (figure 8) of UAM DFW vertiport, marked with purple boxes.Between the North and South blue lines, conventional aircraft are below 1000 ft AGL.
+Fig. 5 Fig. 6 A56Fig. 5 Sensitive Area North of DFW.UAM flight route in green, conventional aircraft landing and take-off paths in orange and dark red.Bright red arrows denote horizontal separations.South of blue line, conventional aircraft are below 1000 ft AGL.
+Fig. 77Fig. 7 Worst case scenario, when conventional aircraft is above 1000 ft AGL
+Fig. 1111Fig. 11 Risk of triggering RAs on a typical day of operation
+
+
+
+Table 1 TCAS II Version 7.1 Sensitivity Levels and Thresholds for Resolution Advisories [2] Ownship Altitude SL Tau(sec) DMOD(nmi) ZTHR(ft) (ft1)<1000 (AGL)2N/AN/AN/A1000-2350 (AGL)3150.206002350-50004200.356005000-100005250.5560010000-200006300.8060020000-420007351.10700>420007351.10800
+		
+		
+			
+
+				
+
+
+	
+		Exploration of Near term Potential Routes and Procedures for Urban Air Mobility
+		
+			SavitaVerma
+		
+		
+			JillianKeeler
+		
+		
+			TamsynEEdwards
+		
+		
+			VictoriaDulchinos
+		
+		10.2514/6.2019-3624
+	
+	
+		AIAA Aviation 2019 Forum
+		
+			American Institute of Aeronautics and Astronautics
+			2019
+		
+	
+	Verma, S. A., Keeler, J. N., Edwards, T. E., and ulchinos, V. L., "Exploration of Near Term Potential Routes and Procedures for urban Air Mobility," AIAA Aviation Technology, Integration, and Operations Conference, 2019.
+
+
+
+	
+		Federal Aviation Administration
+		10.4135/9781544377230.n127
+	
+	
+		Federal Regulatory Guide
+		
+			CQ Press
+			2011. Feb 28
+			
+		
+	
+	Federal Aviation Administration
+	Federal Aviation Administration, "Introduction to TCAS II Version 7.1 booklet," , 2011. Feb 28.
+
+
+
+	
+		Aircraft Hardstand Ramp Expansion at DFW International Airport
+		
+			KMBymers
+		
+		
+			MOBejarano
+		
+		10.1061/9780784482476.036
+		
+	
+	
+		Airfield and Highway Pavements 2019
+		
+			American Society of Civil Engineers
+			2019
+			
+		
+	
+	"DFW Airport Aircraft Noise," https://www.dfwairport.com/aircraftnoise/, 2019. Accessed: 2019-04-14.
+
+
+
+	
+		A Terminal Area Analysis of Continuous Ascent Departure Fuel Use at Dallas/Fort Worth International Airport
+		
+			KeenanRoach
+		
+		
+			JohnRobinson
+		
+		10.2514/6.2010-9379
+	
+	
+		10th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference
+		
+			American Institute of Aeronautics and Astronautics
+			2010
+			9379
+		
+	
+	Roach, K., and Robinson, J., "A terminal area analysis of continuous ascent departure fuel use at Dallas/Fort Worth international airport," 10th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference, 2010, p. 9379.
+
+
+
+	
+		Natural gamma aeroradioactivity map of the Fort Worth-Dallas area, Texas
+		10.3133/gp696
+		
+		
+			2019
+			US Geological Survey
+		
+	
+	"FAA Helicopter Routes Map for Dallas-Fort Worth Area," http://aeronav.faa.gov/content/aeronav/heli_files/ PDFs/Dallas-Ft_Worth_Heli_7_P.pdf, 2019. Images produced by the U.S. Government and in the public domain.
+
+
+
+	
+		The Generic Resolution Advisor and Conflict Evaluator (GRACE) for Detect-And-Avoid (DAA) Systems
+		
+			MichaelAbramson
+		
+		
+			MohamadRefai
+		
+		
+			ConfesorSantiago
+		
+		10.2514/6.2017-4485
+	
+	
+		17th AIAA Aviation Technology, Integration, and Operations Conference
+		
+			American Institute of Aeronautics and Astronautics
+			2017
+			4485
+		
+	
+	Abramson, M., Refai, M., and Santiago, C., "The Generic Resolution Advisor and Conflict Evaluator (GRACE) for Detect-And- Avoid (DAA) Systems," 17th AIAA Aviation Technology, Integration, and Operations Conference, 2017, p. 4485.
+
+
+
+	
+		HALEAKALĀ ON GOOGLE MAPS (SATELLITE VIEW)
+		
+			GoogleMaps
+		
+		10.2307/jj.2089642.6
+		97.0537879
+		
+	
+	
+		Aina Hanau / Birth Land
+		
+			University of Arizona Press
+			20210m/data=!3m1!1e3, 2019
+			
+		
+	
+	Google Maps, "Dallas/Fort Worth International Airport Area Satellite View," https://www.google.com/maps/@32. 9162036,-97.0537879,20210m/data=!3m1!1e3, 2019. Accessed: 2019-05-16.
+
+
+
+	
+		Eurocontrol Navigation System Proposal
+		10.1108/eb033549
+		
+	
+	
+		Aircraft Engineering and Aerospace Technology
+		0002-2667
+		
+			34
+			4
+			
+			2019
+			Emerald
+		
+	
+	"Eurocontrol Aircraft Performance Database," https://contentzone.eurocontrol.int/aircraftperformance/ details.aspx?ICAO=B738&ICAOFilter=B738, 2019. Accessed: 2019-04-14.
+
+
+				
+			
+		
+	

file108.txt

@@ -0,0 +1,620 @@
+
+	
+	
+		
+I. IntroductionThe airspace of the future is expected to support Unmanned Aircraft Systems (UAS) aircraft operations that are orders of magnitude higher than conventional aviation traffic the National Airspace System (NAS) handles today [1][2][3].An important question is if there is a traffic density at which the airspace becomes too complex to operate.This paper presents an approach to estimate the complexity of a given UAS traffic scenario with an associated traffic density.The availability of airspace to meet traffic demand in a safe and efficient manner is central to airspace operations both today and in the future with increasing number of unmanned and manned vehicles sharing the airspace with commercial air traffic.Measures of airspace complexity are used by air traffic management to schedule flights and resolve conflicts.New measures of airspace complexity are needed to make traffic flow decisions as controller workload limitations are enhanced or removed in certain parts of the airspace by increased automation.In conventional aviation, air traffic complexity is evaluated from controller and pilot workload [4][5][6][7].Monitor Alert Parameter (MAP) [8], the maximum number of aircraft an Air Traffic Control (ATC) controller can simultaneously handle, is an example.Another example is Dynamic Density (DD) [9,10], a weighted summation of factors that affect the air traffic complexity.The complexity metrics are defined based on an assumption of a structured airspace and Air Traffic Management (ATM) that includes controller displays, sectors and airways [11][12][13].Fast-time and real-time simulation methods [14] are then used to evaluate a given traffic scenario.Intrinsic metrics have also been developed to estimate complexity in a sector independent of controller workload.Delahaye et.al. [15,16] proposed a geometrical approach based on the properties of relative positions and relative speeds of aircraft in a sector to obtain time histories of traffic divergence, convergence, and sensitivity.They also developed entropy based velocity vector field methods [16,17] to compute complexity maps for given traffic scenario snapshots.Future UAS traffic and its management would differ from conventional traffic in several ways.First, the high number of proposed operations suggests a need to shift from a human to an automated controller, negating the use of cognitive measures.Second, future operations may be free flight by nature i.e. using a predefined on-board conflict resolution model, they may prefer responsibility for determining their course independent of a global plan or system [2,18].Furthermore, a good and quick approximation of UAS traffic complexity, without the need for a full-scale simulation, would support a real-time assessment of traffic scenarios [19], re-planning of flight routes and schedules to alleviate traffic bottlenecks, and mitigation of operation risk.A set of different complexity metrics can together also help classification of traffic scenarios for traffic management studies.This paper extends an earlier approach by Xue [20] that introduced a scenario complexity metric based on the number of potential conflicts weighted by the associated conflict resolution cost.In this work, an impedance-based complexity metric is proposed.When the number of conflicts is high, a scenario is expected to be complex.However, it may not necessarily be so if those conflicts are isolated.The pattern of traffic flow also contributes to the complexity of the scenario, if the aircraft begin to impede the conflict resolutions of others in the airspace.The proposed impedance metric therefore accounts both for the number of conflicts and the free space available around aircraft in conflict.Furthermore, it simplifies the complexity computation process.First, the metric is tuned for a given type of aircraft and conflict resolution method, using high-fidelity simulation data as baseline.Then the impedance for new scenarios can be computed assuming the aircraft as point masses, without having to simulate the vehicle dynamics or actual conflict resolution maneuvers for each of the scenarios.A high-fidelity fast-time simulator, Fe 3 [21], was used to simulate over a thousand randomly generated scenarios and measure the baseline for complexity.The impedance metric was then evaluated separately for the same scenarios and compared against the baseline data and the metric from previous study.The results showed that the correlation between the proposed impedance metric is better than the previous metric.It also produces impedance maps showing regions of high complexity which provides better spatial information for managing air traffic.The rest of the paper is organized in the following way.The test scenario description, and the metric and its evaluation in detail are presented in Section II.Detailed results and discussion are presented in Section III.Summary of the paper, as well as identification of directions for further research is presented in Section IV.
+II. MethodologyIn this paper, the focus is on estimating the complexity of operations in a two-dimensional horizontal portion of airspace that has no constraints such as controlled airspace, temporary flight restrictions, geo-fences or terrain.In the next two sections, the test scenarios and the evaluation of the impedance metric are described in detail.
+A. Test ScenariosThe test scenario generation remains the same as used in [20].To evaluate the complexity metrics, random scenarios with a large variety of complexities were generated.Then the metric referred to as "number of resolution maneuvers" was evaluated in the Fe 3 [21] simulator.The high-fidelity simulator uses dynamic models of aircraft and a pairwise conflict resolution method that employs a combination of speed and direction changes to simulate the trajectories and encounters of aircraft typical to a real scenario.The aircraft are therefore actually diverting during close encounters instead of a prescribed course change.Using the simulation-generated measurements as the baseline for the complexity of the scenario, the proposed complexity metric, Impedance (I), was then analyzed and compared using statistical methods.Note, the number of conflict resolution maneuvers measures the resolution moves issued during the simulation.Since the time step size in Fe 3 is 0.5 seconds, the number of conflict resolution maneuvers also reflects the resolution duration.For the generated scenarios, several criteria were used to ensure high traffic intensity and comparability in scenarios.First, a 1.3x1.3nautical mile region was defined (shown as the red box in Fig. 1), and all flights were required to go through the predefined region with origin and destination outside of the region.Second, at most one turning point was allowed other than the origin and destination in a flight plan.Third, all flights were set to depart within a five-minute window.Lastly, this study focused on low-altitude small UAS traffic.Hence, the target ground speeds of all flights were set in the range of 5 meters per second to 20 meters per second.Fig. 1 shows a sample scenario with 30 vehicles, where the circle, cross, and diamond markers represent origins, destinations, and mid-points, respectively.In [20], the number of aircraft in these scenarios was varied from 5 to 50 (or in density from 3 to 30 vehicles/nmi 2 ).Fig. 2 shows the percentage of scenarios with and without conflicts during the process of generating scenarios.When the traffic density increases, the likelihood of having conflicts increases and reaches 100 % at approximately 15 Fig. 1 A Sample Scenario with 30 flights vehicles/nmi 2 .Additionally, scenarios with aircraft densities from 50 to 100 were also generated in increments of 2. Consequently, a total of 1045 scenarios were created and used.From 5 to 50, 20 scenarios were generated at each level of density and from 52 to 100, 5 scenarios were generated at every alternate level of density.Scenarios without conflicts are defined as having zero scenario complexity based on the proposed metric.Therefore, only the scenarios with potential conflicts are used in experiments.
+Fig. 2 Likelihood of Conflicts at Different Density Levels
+B. The Impedance Metric and Its EvaluationTo evaluate the complexity of a scenario, first a notion of conflict is defined.Two aircraft are assumed to be in conflict at a given time if they are within a distance h sep = k.D wc of each other, where D wc is the well-clear distance (arbitrarily chosen at 50 feet or 15.24 meters) and k is a rational number (≥ 1) multiplication factor.h sep is referred to as Conflict Distance in this paper.This approach is used to calibrate the Impedance evaluation to the region of influence of the conflict resolution method that would have been used in an actual traffic simulation.For a given scenario, the Impedance metric is computed as follows:Let R be the pre-defined region of interest in two-dimensional Euclidean space, i.e.R ∈ R 2 .Grid the region into square cells C xy with a side length l, where x is the row number and y is the column number.Each cell has n adjacent cells C axy , where n ∈ [3,5,8] depending on the location of the cell (corner, edge, interior) in the grid.The letter a is used to denote adjacent.At each instant of time t i in the entire duration of the scenario, compute the aircraft occupancy graph/map O t i =[O xy,t i ], where O xy,t i is the number of aircraft in a cell at that time.Also, at each t i , compute the aircraft conflict graph/map C t i ,c =[C xy,t i ,c ], the set of all cells that have at least one conflict in them.For each C xy,t i ,c , let m of the adjacent cells C axy,t i ,c have at least one aircraft in them over the next dt seconds.This can be obtained for each adjacent cell by summing its occupancy, O xy,t j for t j ∈ [t i + 1,t i + dt].The Impedance of a Cell in space (location xy) at time t i , I xy,t i = m/n.Thus at each time instant t i , there is a colored grid/map produced, called the Impedance MapI t i =[I xy,t i ],where the color of a cell indicates its impedance I xy,t i ∈ [0, 1].Now, to get a single snapshot of the region, the time slices need to be collapsed over the entire period of the scenario.This produces the Impedance graph/map of the Scenario I xy .An example is shown in Fig. 3.This can be done by either taking the time mean or a percentile value of each cell's impedance.To understand the severity of impedance in each cell over the entire scenario, in this study, the p th percentile value of I xy,t is used to collapse the graphs in time.Finally, to get a single impedance number for the whole scenario, both the space and time dimensions need to be collapsed.To do this, compute the percentage of cells in the time collapsed map with I xy ≥ P , where P is the chosen impedance threshold.This gives the Impedance, I for the entire scenario.For example, suppose choosing the 99 th percentile value for time collapse and an impedance threshold, P = 20%, results in an impedance value of 0.3 for a scenario.This can be interpreted as -30% of the region has conflicts that are impeded by nearby aircraft in one-fifth (20%) of the vicinity, 1% of the time.In other words, conflicts in almost one-third of the region are impeded in the scenario.The metric computation uses two parameters: the conflict distance parameter k and the time window parameter dt.A set of traffic scenario complexities evaluated in a high-fidelity simulator like Fe 3 , with a given conflict resolution model, is used to tune the impedance metric parameters to achieve the maximum correlation.Then, the tuned parameters can be used to evaluate the complexity for any new scenario (comprising the same type of aircraft and the same conflict resolution model) without the need for high-fidelity simulations.The metric captures the effect of aircraft dynamics and resolution models in its parameters.For example, an aircraft with low maneuverability will need more space and the conflict resolution method will have a high conflict distance when evaluating the baseline data in Fe 3 .That in turn means the Impedance metric tuned for that type of aircraft and conflict resolution method will have a different value of k and dt where it is most correlated with the baseline data.In other words, the tuning of the Impedance metric parameters is tied to the type of aircraft and conflict resolution method that will be used in the real scenarios.
+C. AssumptionsIn addition to the scenario assumptions stated earlier, the chosen values of different parameters are defined as follows:-The cell edge length l = 100m (0.054nmi).-For each scenario, impedance is computed by varying the k value between 1 and 5, to ensure that the conflict distance is less than the cell edge length.-The time window dt is varied between 3 seconds and 17 seconds with 2-second increments.Since each aircraft flies at a speed between 5 to 20 meters per second, an aircraft will leave a cell anytime between the next time step to 20 seconds at most.Hence, time window values are chosen between those numbers, ignoring the smallest and largest values.On average it will take about 5 seconds to reach an adjacent cell.-The percentile number p = 99.9.This is done to capture the worst impedances observed at every cell over the entire duration of the scenario.-The impedance threshold P is varied from 10% to 80%.In other words, if 10% impedance is considered bad, having a third aircraft in the vicinity is considered bad and every moderate to bad cell will contribute to the scenario complexity.This is typically what might be considered bad in a realistic scenario today.On the other hand, if only 80% or more impedance is considered bad, only the worst of the bad cells will contribute.This could be the case for a highly futuristic scenario where multi-aircraft conflicts are operationally acceptable.-Since this study focuses on scenario intrinsic complexity, uncertainties on wind, communication, navigation, and surveillance were not included in the simulations.After the impedance metric for each scenario is evaluated, the Pearson method is used to compute the correlation between the impedance measures and the number of resolution maneuvers (baseline) for each scenario.This varies as a function of the chosen k, P and dt values and is discussed under results.Furthermore, since the Pearson method is designed for checking linear correlations, a maximal correlation method [22] is also used to capture any non-linear association, and the Alternative Conditional Expections (ACE) method implemented in Matlab is used to compute such maximal correlations.
+III. ResultsImpedance maps were generated for each scenario at different conflict distances and time windows.For any given time window, two common trends were observed.For a fixed conflict distance, h sep , as the traffic density was increased, the spread and value of the impedance increased (the cells became yellower/brighter) (Fig. 4).For a fixed traffic density, as the conflict distance was increased, again the spread and value of impedance increased (Fig. 5).However, the effect was less pronounced.This indicates that the metric is more sensitive to the traffic density than the conflict distance.For computing the final impedance number for each scenario, the cell impedance threshold was varied from 10% to 80%.This was repeated at different conflict distances.For each time window, to determine the best combination of conflict distance and cell impedance threshold, correlations were computed between the impedance numbers and the baseline data.In general, as the conflict distance was increased, the peak correlation was observed at higher impedance thresholds.It was found that a conflict distance, h sep of 75 feet and a threshold, P of 10% had the best correlation coefficient when compared with the number of resolution maneuvers from the actual simulation, irrespective of the time window.As the time window was increased, the correlation improved up to a time window of 9 seconds, and then deteriorated.The best Pearson correlation observed was 0.9207 for the 9-second time window (Fig. 6).The correlation obtained for the same using the ACE method was 0.9325.The corresponding best correlation coefficients for the weighted conflict complexity metric from our earlier work were 0.9 and 0.913, respectively [20].The impedance metric therefore performed better.
+A. DiscussionThe impedance metric serves two purposes.First, the space and time collapsed impedance of a scenario provides a single complexity number in the usual sense of measuring airspace complexity.It captures the impact of both number of conflicts, and the relative spatial distribution of aircraft in conflict with respect to other aircraft in vicinity, which could impede the performance of the conflict resolution strategy.Second purpose is the impedance map, which provides visual information to identify the hot spots: regions with limited conflict resolution capability in the scenario.This is useful in not only flagging a scenario as too complex but also pin-pointing where the problem is.Consequently, more informed air traffic management decisions can be taken.Suppose an arbiter runs a scenario and the impedance is above a threshold, then the maps show her where the hot spots are, and she could either deny the whole scenario, or just the flights which go through that area, or provide a reroute to flights going through hot spots, and so on.
+IV. ConclusionsIn this paper, an impedance-based metric was introduced to represent the complexity of a given unmanned aircraft system traffic scenario.There were 1045 scenarios analyzed, and their impedance metric was computed and compared against the baseline data produced from high-fidelity simulation.The metric was evaluated for varying conflict distances and traffic scenarios.It was found that the proposed metric had a high correlation of 0.92 (Pearson) and 0.9325(ACE) Additionally, the metric provides a way to account for both the number of aircraft and the traffic flow pattern.The impedance maps produced as part of the impedance computation process identified areas of concern in a given scenario.Such information may be helpful for developing traffic management strategies such as adjusting and re-planning only flights that pass through the most impeded areas.The air traffic services in a UAM environment may be provided by one or more operators.Each operator needs real-time tools to assess the safety and efficiency of operations and make adjustments to changing traffic demands.The impedance metric provides a tool to identify hot spots, regions with limited conflict resolution capability, in the airspace operations.It can therefore also be used to assess if and how the hot spots vary with uncertainties like sudden changes in demand, wind speed variations and low visibility.Similarly, it could be used to reallocate demand to maintain safety.The results presented in this paper investigated the top 0.1% impedances in each cell for a scenario.In other words, the metric is only studying the worst 1 minute for every 1000 minutes in an area.Other levels of impedance can be explored further and tested against the baseline.Also, for this analysis, a fixed grid cell edge length was assumed.Grids with varying edge lengths could also be explored.Finally, this work extends our earlier work that introduced a weighted conflict-based scenario complexity metric.The impedance metric performed marginally better than the weighted conflict metric.This is part of an extended effort to develop complexity metrics that can be computed in real time without the need for scenario simulation.This therefore can be used for jointly classifying a scenario as acceptable, unacceptable or acceptable with changes made to flight plans that pass through high complexity regions and then applying necessary air traffic management strategies.Fig. 33Fig. 3 The Impedance Map of the Sample Scenario with 30 flights at conflict distance, h sep = 60.96m, for a time window, dt = 5 sec
+Fig. 44Fig. 4 Impedance maps with varying traffic density and fixed conflict distance, h sep = 45.72m for a time window, dt = 9 sec
+Fig. 55Fig. 5 Impedance maps with varying conflict distance, h sep and fixed traffic density for a time window, dt = 9 sec
+Fig. 66Fig. 6 Correlation between Impedance and Number of Resolution Maneuvers (baseline) as a function of conflict distance and impedance threshold at dt = 9 sec
+
+
+		
+		
+			
+
+				
+
+
+	
+		A ground-delay-based approach to reduce impedance-based airspace complexity
+		
+			VishwanathBulusu
+		
+		
+			RSengupta
+		
+		
+			ZLiu
+		
+		10.2514/6.2021-2340
+	
+	
+		AIAA AVIATION 2021 FORUM
+		
+			American Institute of Aeronautics and Astronautics
+			2016
+		
+	
+	Bulusu, V., Sengupta, R., and Liu, Z., "Unmanned Aviation: To Be Free or Not To Be Free? A Complexity Based Approach," 7th International Conference on Research in Air Transportation, 2016.
+
+
+
+	
+		A threshold based airspace capacity estimation method for UAS traffic management
+		
+			VishwanathBulusu
+		
+		
+			ValentinPolishchuk
+		
+		10.1109/syscon.2017.7934758
+	
+	
+		2017 Annual IEEE International Systems Conference (SysCon)
+		
+			IEEE
+			2017
+			
+		
+	
+	Bulusu, V., and Polishchuk, V., "A threshold based airspace capacity estimation method for UAS traffic management," 2017 Annual IEEE International Systems Conference (SysCon), IEEE, 2017, pp. 1-7.
+
+
+
+	
+		Capacity Estimation for Low Altitude Airspace
+		
+			VishwanathBulusu
+		
+		
+			ValentinPolishchuk
+		
+		
+			RajaSengupta
+		
+		
+			LeonidSedov
+		
+		10.2514/6.2017-4266
+	
+	
+		17th AIAA Aviation Technology, Integration, and Operations Conference
+		
+			American Institute of Aeronautics and Astronautics
+			2017
+			4266
+		
+	
+	Bulusu, V., Polishchuk, V., Sengupta, R., and Sedov, L., "Capacity estimation for low altitude airspace," 17th AIAA Aviation Technology, Integration, and Operations Conference, 2017, p. 4266.
+
+
+
+	
+		Estimation of European Airspace Capacity from a Model of Controller Workload
+		
+			ArnabMajumdar
+		
+		
+			WashingtonOchieng
+		
+		
+			JohnPolak
+		
+		10.1017/s037346330200190x
+	
+	
+		Journal of Navigation
+		J. Navigation
+		0373-4633
+		1469-7785
+		
+			55
+			3
+			
+			2002
+			Cambridge University Press (CUP)
+		
+	
+	Majumdar, A., Ochieng, W., and Polak, J., "Estimation of European airspace capacity from a model of controller workload," Journal of Navigation, Vol. 55, No. 03, 2002, pp. 381-403.
+
+
+
+	
+		En-route sector capacity estimation methodologies: An international survey
+		
+			ArnabMajumdar
+		
+		
+			WashingtonYottoOchieng
+		
+		
+			JamesBentham
+		
+		
+			MartynRichards
+		
+		10.1016/j.jairtraman.2005.05.002
+	
+	
+		Journal of Air Transport Management
+		Journal of Air Transport Management
+		0969-6997
+		
+			11
+			6
+			
+			2005
+			Elsevier BV
+		
+	
+	Majumdar, A., Ochieng, W. Y., Bentham, J., and Richards, M., "En-route sector capacity estimation methodologies: An international survey," Journal of Air Transport Management, Vol. 11, No. 6, 2005, pp. 375-387.
+
+
+
+	
+		Airspace capacity estimation using flows and Weather-Impacted Traffic Index
+		
+			AlexanderKlein
+		
+		
+			LaraCook
+		
+		
+			BryanWood
+		
+		
+			DavidSimenauer
+		
+		10.1109/icnsurv.2008.4559188
+	
+	
+		2008 Integrated Communications, Navigation and Surveillance Conference
+		
+			IEEE
+			2008
+			
+		
+	
+	Klein, A., Cook, L., Wood, B., and Simenauer, D., "Airspace capacity estimation using flows and weather-impacted traffic index," 2008 Integrated Communications, Navigation and Surveillance Conference, IEEE, 2008, pp. 1-12.
+
+
+
+	
+		Capacity Estimation for Airspaces with Convective Weather Constraints
+		
+			JimmyKrozel
+		
+		
+			JosephMitchell
+		
+		
+			ValentinPolishchuk
+		
+		
+			JosephPrete
+		
+		10.2514/6.2007-6451
+	
+	
+		AIAA Guidance, Navigation and Control Conference and Exhibit
+		
+			American Institute of Aeronautics and Astronautics
+			2007
+		
+	
+	Krozel, J., Mitchell, J., Polishchuk, V., and Prete, J., "Airspace capacity estimation with convective weather constraints," AIAA Guidance, Navigation, and Control Conference, 2007.
+
+
+
+	
+		Applications of a Macroscopic Model for En Route Sector Capacity
+		
+			JerryWelch
+		
+		
+			JohnAndrews
+		
+		
+			BrianMartin
+		
+		
+			EricShank
+		
+		10.2514/6.2008-7221
+	
+	
+		AIAA Guidance, Navigation and Control Conference and Exhibit
+		Barcelona, Spain
+		
+			American Institute of Aeronautics and Astronautics
+			2007
+			138
+		
+	
+	Welch, J. D., Andrews, J. W., Martin, B. D., and Sridhar, B., "Macroscopic workload model for estimating en route sector capacity," Proc. of 7th USA/Europe ATM Research and Development Seminar, Barcelona, Spain, 2007, p. 138.
+
+
+
+	
+		AIR TRAFFIC MANAGEMENT
+		
+			IVLaudeman
+		
+		
+			SShelden
+		
+		
+			RBranstrom
+		
+		
+			CBrasil
+		
+		10.1201/9781482267952-27
+	
+	
+		Contemporary Ergonomics 1998
+		
+			CRC Press
+			1998
+			
+		
+	
+	Laudeman, I. V., Shelden, S., Branstrom, R., and Brasil, C., "Dynamic density: An air traffic management metric," 1998.
+
+
+
+	
+		Airspace complexity and its application in air traffic management
+		
+			BSridhar
+		
+		
+			KSSheth
+		
+		
+			SGrabbe
+		
+	
+	
+		Europe Air Traffic Management R&D Seminar
+		
+			
+			1998
+		
+	
+	2nd USA/
+	Sridhar, B., Sheth, K. S., and Grabbe, S., "Airspace complexity and its application in air traffic management," 2nd USA/Europe Air Traffic Management R&D Seminar, 1998, pp. 1-6.
+
+
+
+	
+		The Complexity Construct in Air Traffic Control: A Review and Synthesis of the Literature
+		
+			RHMogford
+		
+		
+			JGuttman
+		
+		
+			SMorrow
+		
+		
+			PKopardekar
+		
+		
+			1995
+		
+	
+	Tech. rep., DTIC Document
+	Mogford, R. H., Guttman, J., Morrow, S., and Kopardekar, P., "The Complexity Construct in Air Traffic Control: A Review and Synthesis of the Literature." Tech. rep., DTIC Document, 1995.
+
+
+
+	
+		Practice for Application of Federal Aviation Administration (FAA) Federal Aviation Regulations Part 21 Requirements to Unmanned Aircraft Systems (UAS)
+		
+			PKopardekar
+		
+		10.1520/f2505-07
+	
+	
+		Federal Aviation Administration
+		
+			ASTM International
+			2000
+		
+	
+	Dynamic density: A review of proposed variables
+	Kopardekar, P., "Dynamic density: A review of proposed variables," FAA internal document. overall conclusions and recommendations, Federal Aviation Administration, 2000.
+
+
+
+	
+		Airspace complexity measurement: An air traffic control simulation analysis
+		
+			PHKopardekar
+		
+		
+			ASchwartz
+		
+		
+			SMagyarits
+		
+		
+			JRhodes
+		
+	
+	
+		International Journal of Industrial Engineering: Theory, Applications and Practice
+		
+			16
+			1
+			
+			2009
+		
+	
+	Kopardekar, P. H., Schwartz, A., Magyarits, S., and Rhodes, J., "Airspace complexity measurement: An air traffic control simulation analysis," International Journal of Industrial Engineering: Theory, Applications and Practice, Vol. 16, No. 1, 2009, pp. 61-70.
+
+
+
+	
+		Air traffic predictability framework – Development, performance evaluation and application
+		
+			GonzaloTobaruela
+		
+		
+			PeterFransen
+		
+		
+			WolfgangSchuster
+		
+		
+			WashingtonYOchieng
+		
+		
+			ArnabMajumdar
+		
+		10.1016/j.jairtraman.2014.04.001
+	
+	
+		Journal of Air Transport Management
+		Journal of Air Transport Management
+		0969-6997
+		
+			39
+			
+			2012
+			Elsevier BV
+		
+	
+	Tobaruela, G., Majumdar, A., and Ochieng, W. Y., "Identifying Airspace Capacity Factors in the Air Traffic Management System," Proceedings of the 2nd International Conference on Application and Theory of Automation in Command and Control Systems, 2012, pp. 219-224.
+
+
+
+	
+		Airspace Congestion Metrics
+		
+			DanielDelahaye
+		
+		
+			StéphanePuechmorel
+		
+		10.1002/9781118743805.ch7
+	
+	
+		Modeling and Optimization of Air Traffic
+		Napoli, Italy
+		
+			John Wiley & Sons, Inc.
+			2000
+			
+		
+	
+	3rd USA/
+	Delahaye, D., and Puechmorel, S., "Airspace Complexity: Towards Intrinsic Metrics," 3rd USA/Europe Air Traffic Management R&D Seminar, Napoli, Italy, 2000.
+
+
+
+	
+		Air Traffic Complexity Map based on Non Linear Dynamical Systems
+		
+			DanielDelahaye
+		
+		
+			StephanePuechmorel
+		
+		
+			JohnHansman
+		
+		
+			JonathanHiston
+		
+		10.2514/atcq.12.4.367
+	
+	
+		Air Traffic Control Quarterly
+		Air Traffic Control Quarterly
+		1064-3818
+		2472-5757
+		
+			12
+			4
+			
+			2004
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	Delahaye, D., Puechmorel, S., Hansman, J., and Histon, J., "Air Traffic Complexity based on Non Linear Dynamical Systems," Air Traffic Control Quarterly, Vol. 12, No. 4, 2004, pp. 367-388.
+
+
+
+	
+		Describing Air Traffic Complexity Using Mathematical Programming
+		
+			MariyaIshutkina
+		
+		
+			EricFeron
+		
+		
+			KarlBilimoria
+		
+		10.2514/6.2005-7453
+	
+	
+		AIAA 5th ATIO and16th Lighter-Than-Air Sys Tech. and Balloon Systems Conferences
+		
+			American Institute of Aeronautics and Astronautics
+			2005
+		
+	
+	Ishutkina, M. A., and Feron, E., "Describing Air Traffic Complexity Using Mathematical Programming," AIAA 5th Aviation, Technology, Integration, and Operations Conference (ATIO), 2005.
+
+
+
+	
+		Cooperative and non-cooperative UAS traffic volumes
+		
+			VishwanathBulusu
+		
+		
+			RajaSengupta
+		
+		
+			ValentinPolishchuk
+		
+		
+			LeonidSedov
+		
+		10.1109/icuas.2017.7991506
+	
+	
+		2017 International Conference on Unmanned Aircraft Systems (ICUAS)
+		
+			IEEE
+			2017
+		
+	
+	Bulusu, V., Sengupta, R., Sedov, L., and Polishchuk, V., "Cooperative and Non-Cooperative UAS Traffic Volumes," International Conference on Unmanned Aircraft Systems ICUAS, 2017.
+
+
+
+	
+		A Throughput Based Capacity Metric for Low-Altitude Airspace
+		
+			VishwanathBulusu
+		
+		
+			RajaSengupta
+		
+		
+			EricRMueller
+		
+		
+			MinXue
+		
+		10.2514/6.2018-3032
+	
+	
+		2018 Aviation Technology, Integration, and Operations Conference
+		Atlanta, GA
+		
+			American Institute of Aeronautics and Astronautics
+			2018
+		
+	
+	Bulusu, V., Sengupta, R., Mueller, E. R., and Xue, M., "A Throughput-Based Capacity Metric for Low-altitude Airspace," AIAA Aviation Forum, Atlanta, GA., 2018.
+
+
+
+	
+		Scenario Complexity for Unmanned Aircraft System Traffic
+		
+			MinXue
+		
+		
+			MinhDo
+		
+		10.2514/6.2019-3513
+	
+	
+		AIAA Aviation 2019 Forum
+		
+			American Institute of Aeronautics and Astronautics
+			2019
+			3513
+		
+	
+	Xue, M., and Do, M., "Scenario Complexity for Unmanned Aircraft System Traffic," AIAA Aviation 2019 Forum, 2019, p. 3513.
+
+
+
+	
+		Fe<sup>3</sup>: An Evaluation Tool for Low-Altitude Air Traffic Operations
+		
+			MinXue
+		
+		
+			JosephRios
+		
+		
+			JosephSilva
+		
+		
+			ZhifanZhu
+		
+		
+			AbrahamKIshihara
+		
+		10.2514/6.2018-3848
+	
+	
+		2018 Aviation Technology, Integration, and Operations Conference
+		Atlanta, GA
+		
+			American Institute of Aeronautics and Astronautics
+			2018
+		
+	
+	Xue, M., Rios, J., Silva, J., Ishihara, A., and Zhu, Z., "Fe3: An Evaluation Tool for Low-Altitude Air Traffic Operations," AIAA Aviation Forum, Atlanta, GA., 2018.
+
+
+
+	
+		Estimating Optimal Transformations for Multiple Regression and Correlation
+		
+			LeoBreiman
+		
+		
+			JeromeHFriedman
+		
+		10.1080/01621459.1985.10478157
+	
+	
+		Journal of the American Statistical Association
+		Journal of the American Statistical Association
+		0162-1459
+		1537-274X
+		
+			80
+			391
+			
+			1985
+			Informa UK Limited
+		
+	
+	Breiman, L., and Friedman, J. H., "Estimating optimal transformations for multiple regression and correlation," Journal of the American statistical Association, Vol. 80, No. 391, 1985, pp. 580-598.
+
+
+				
+			
+		
+	

file109.txt

@@ -0,0 +1,100 @@
+
+	
+	
+		
+BACKGROUNDIn support of the Dynamic Interface Modeling and Simulation System (DIMSS) task in the Navy's Joint Shipboard Helicopter Integration Process (JSHIP) program, a real-time, piloted simulation experiment was conducted in the Vertical Motion Simulator (VMS) facility at NASA Ames Research Center.The purpose of the experiment was to develop and evaluate the capability of simulation to conduct dynamic interface testing, in order to establish the operational envelope for helicopters landing on ships in various wind conditions and sea states.The experiment simulated a UH-60 helicopter landing on an LHA ship, and incorporated an unsteady (timevarying) ship airwake model computed by Computational Fluid Dynamics (CFD) techniques.
+THE VERTICAL MOTION SIMULATORThe VMS is the world's largest R&D motion-base flight simulator.To study a variety of different aircraft, the VMS uses interchangeable cabs.For this particular study, an interchangeable cab was created to emulate the UH-60 helicopter.Barco rear-projection displays were installed to give a wide-angle out-the-window view approximating that seen from the cockpit of a Black Hawk, and a close representation of its instrument panel was generated using computer graphics.A DEC Alpha simulator host computer ran the Open VMS operating system.The simulation also included an Evans and Sutherland ESIG 4530 computer-generated "out-the-window" display system; graphical displays generated by Silicon Graphics computers, simulating the aircraft instruments; a hydraulic control loader system to simulate the control inceptors; and an ASTi sound simulation system.For this study, two other items were integrated into the simulator: a CATI PCbased XIG "out-the-window" display system, for comparison with the ESIG 4530; and a dynamic seat.The Evans and Sutherland ESIG 4530 was upgraded to include an "ocean wave" simulation capability.The waves were specified in terms of modal period and significant wave height.A graphical model of an LHA ship was developed for the simulation, and a real-time version of the CARDEROCK Ship Motion Program (SMP) was integrated into the simulator host computer to provide realistic ship motion.A Landing Safety Enlisted (LSE) man, with limited hand-signaling capability, was also programmed into the ESIG.
+UH-60 MODELThe UH-60 helicopter was simulated using a bladeelement model originally developed by Sikorski Aircraft and documented under contract from NASA. 1 The helicopter model consists of a main rotor model, aerodynamic models of the fuselage, horizontal and vertical tail surfaces, and tail rotor, as well as a simulation of the engine, drive train, flight controls, and landing gear.The blade-element main rotor model used five "equal annuli" segments on each of the four blades.All calculations for the helicopter flight model, including the airwake, were performed at an update rate of 100 Hz.
+CFD AIRWAKE MODELThe CFD airwake model consisted of a matrix of time histories at each of 56,661 grid vertices in the region around the ship.The airwake model data were developed by NAVAIR's Advanced Aerodynamics Branch, at Patuxent River Naval Air Station, using a Navier-Stokes formulation of the viscid flow equations in the vicinity of the ship. 2 This simulation produced time-varying values for the three components of the airwake velocity at each of the spatial points.Each of these time histories consisted of 30 seconds of data, sampled every 0.1 second, for a total of 300 points.Therefore, the total number of data values was 50,994,900.
+STORING THE DATAEach of the floating-point data values in the CFD airwake model required 4 bytes of storage in the computer.This would constitute a total file size of nearly 204 Megabytes if the data were stored as a binary file.However, binary files are not computerportable.In order to make the data computer-portable, a special format was developed that would be readable on any system, but would be the same size as a binary file.The data were inspected, and it was determined that the values were always less than 300 feet per second.It was also decided that an accuracy of 0.01 feet per second would be acceptable.Therefore, if the values were multiplied by 100 and truncated to an integer, the result would always be less than 32767.This meant that the data could be scaled and stored in memory as a signed 4-byte integer; and it could be written to a disk file under FORTRAN Z4 format, consisting of four ASCII characters.Thus, the data could be transported in files of the same size as binary files; but, since these files actually consisted of ASCII characters, they would be totally computer-portable.Another issue was the media on which to store the data.Compact disks (CD-ROM's) were the media of choice, since several sets of data could be stored on each one.The following compact disc specifications were adopted to minimize the difficulties in transferring the data:(1) The media type should be CD-R, not CD-RW.(2) The logical format should be ISO 9660, but may use the Joliet extensions (long file names).(3) The CD should not be CD-XA format.
+LOADING THE DATADue to the size of the data files, it was found that special techniques had to be implemented in order to minimize the time required to change data files.This was necessary because the experimental test plan called for frequent changes of wind azimuth (requiring a change of the data file).The technique used was developed specifically for the Open VMS operating system.It is based on the use of a Global Data Section, which is a section of memory that is allocated and named, and filled with data to be used by other programs.A pre-processor was written that would read the data, convert it from INTEGER*4 to REAL*4 and scale it by a factor of 100, and then output it to a file in the special binary format required for the Global Data Section.Whenever a different wind azimuth was desired, a simulation engineer would run a program that would create a Global Data Section using the desired data file.A command issued within the simulation would map an array in the code to the data in the Global Data Section.By using this technique, the time to change from one azimuth to another was reduced from about 20 minutes to a matter of seconds.Although this technique is dependent on specific Open VMS utilities, similar techniques could be developed for other operating systems.
+COORDINATESIn developing and implementing the airwake model and integrating it with the simulation, a number of different coordinate systems are involved:
+CFD Wind CoordinatesThe CFD data were generated using a set of coordinates originating at deck level, on the centerline of the ship, at the most forward part of the bow.The x-axis points aft, parallel to the centerline; the y-axis is toward the starboard side of the ship; and the z-axis is upward, perpendicular to the deck.This forms a right-handed coordinate system.The velocity components are positive when they blow along the positive axes; in other words, a wind blowing from bow to stern, port to starboard, and upward, has all positive components.
+CFD Grid CoordinatesThe CFD grid coordinates are different from the coordinates in which the velocities are represented.The grid coordinates also originate at deck level at the bow on the centerline; however, the coordinates form a left-handed system: the I-axis points aft, the J-axis points to port, and the K-axis is upward.The CFD grid is non-uniform; that is, the blocks of the grid have different sizes, depending on the location.Basically, the grid blocks get larger as the distance from the ship gets larger, as shown in Figure 1.The Earth Axes used in the simulation are local North, East and Down.Their origin, initialized to an arbitrary location relative to the ship, is used for the location of the aircraft.
+Ship AxesThe ship axes used in the simulation originate at the C.G. of the ship, pointing forward, starboard, and downward.
+Aircraft AxesThe aircraft axes originate at the aircraft C.G., pointing forward, right, and downward.
+Blade AxesThe helicopter blade axes are the axes in which the data are used in the rotor model.These axes are tangential, radial and perpendicular to the blade.For the airwake code, the transformation to this axis system was simplified by ignoring flapping and lagging angles.
+INTEGRATING THE AIRWAKE MODEL INTO THE SIMULATIONFor the airwake velocity components to affect the simulated aircraft, the velocity components need to be turned into forces and moments at the center-of-gravity (C.G.) of the vehicle.To do this, the airwake components were calculated at each of the aerodynamic centers of the helicopter.These consist of the fuselage aerodynamic center, the horizontal and vertical tail, the tail rotor, and at each of the main rotor blade segments.In order to look up the airwake velocity components, it was first necessary to locate each of the aircraft aerodynamic centers in the grid coordinates.The airwake data were extracted from the database at eight points adjacent to the aerodynamic center.The values were interpolated to the current time (modulo 30), and then interpolated on the spatial coordinates.Then the velocity components were related to the proper axis system.In order to keep the number of table lookups reasonable for real-time computation, a different algorithm was used to interpolate to the blade-elements of the main rotor.For this, the five points extracted from the database were the rotor hub, and the center of the outermost segment of each blade, as shown in Figure 2.
+Figure 2. Main Rotor Data Lookup PointsThe airwake velocity components at each of the other blade elements were found by interpolating along the blades.In this way, the total number of table lookups was kept to 432.This consists of one for each of the 5 points on the main rotor, plus one for each of the other 4 aerodynamic centers (giving 9 locations); times 2 for the number of time points to be interpolated for each point; times 8 for the number of adjacent points for the spatial interpolation; times 3 for the number of velocity American Institute of Aeronautics and Astronautics components at each point.If the data were to be looked up for each of the twenty blade elements, it would have been necessary to perform 1152 lookups.Although the CFD data had been computed with the ship stationary, it was allowed to steam along a straight course in the simulation.In order to simulate this situation correctly, it was necessary to separate the ambient wind from the airwake.This was done by solving for the ambient wind which, combined the ship speed and direction, would yield the desired wind-overdeck speed and azimuth corresponding to one of the sets of CFD data.The ambient wind was then added into the calculation of the airspeed of the aircraft, and the wind-over-deck was subtracted from the airwake data.The resulting airwake velocity perturbation components at each of the aerodynamic centers were then transformed to the proper coordinates (usually aircraft body axes; except that for the main rotor, blade axes were needed), and then added to the local velocity used to calculate the aerodynamics at that location.Another option that was provided was the capability to simulate wind-over-deck speeds different from the CFD data.All the CFD data had been run at a speed of 30 knots.In order to simulate different wind speeds, it was necessary to scale both the magnitude and frequency spectrum (or time).Scaling the magnitude was simply a matter of multiplying the airwake velocity components by the ratio of the wind-over-deck speed divided by the nominal speed (30 knots).In order to scale the frequency spectrum, the independent time variable used to look up the data in the time history was scaled by the same factor.
+SIMULATION VERIFICATIONThe airwake velocity components integrated into the simulation were verified by comparison with the raw CFD data at specified locations.The total effect was evaluated by experienced UH-60 pilots in the simulator, and was found to be generally realistic and to produce a workload representative of a shipboard landing.One pilot said, "Better than a generic model," and another said, "...better than any training simulator."Some negative comments were also received, but they can be explained by some of the known errors that were present in the airwake model at the time of the evaluation, but that have since been corrected.
+FREQUENCY ANALYSISIn order to better understand the effects produced by the introduction of the time-varying airwake into the UH-60 math model, frequency analysis was performed using a software tool known as Comprehensive Identification from Frequency Responses 3 or CIFER®.Three cases were simulated: one in still air, one with 30 knots of wind-over-deck (but without airwake turbulence), and one with 30 knots of wind-over-deck and airwake turbulence.Each case with wind was run at the two wind-over-deck azimuths of 0 and 60 degrees.Each simulation run was 30 seconds in length, in order to utilize the full 30 seconds of airwake data, thus preserving the low frequency data.Each run was made with the aircraft C.G. at an altitude of 10 feet over the deck of the LHA, hovering over landing spot #5 -just to port of the forward part of the "island" (superstructure) of the ship.For each run, a software flag was set to disable the integration of acceleration to velocity and velocity to position in the equations of motion.This effectively "froze" the aircraft in space, so that the forces and moments exerted on the aircraft by the airwake, and their resulting accelerations, could be analyzed without spurious transition to different locations during the run.For each of these runs, the ship was stationary, with a heading of 0 degrees (North).The aircraft also had a heading of 0 degrees, with zero ground speed (hovering over the deck).Figures 3 through 14 are Power Spectral Density (PSD) plots of the linear and rotational accelerations.Each plot shows three cases.The solid line shows the still air case, the dotted line represents the case with wind but not airwake, and the dashed line shows the case with the airwake.The series of plots on the left is for a wind azimuth of 0 degrees; the plots on the right are for a wind azimuth of 60 degrees.Note the peaks at 108 and 216 radians per second.These are the fundamental and second harmonic of n/rev: there are 4 blades, and the rotation rate is 27 radians per second (257.8rpm).In each case, the effect of adding steady wind does not significantly affect the frequency or energy content in the acceleration responses, except for some additional low frequency energy in some cases.The addition of airwake turbulence is significant in all cases, however, producing a substantial increase in total energy over a wide bandwidth.This effect contributes to pilot workload when landing on a ship in windy conditions.One pilot wrote, "Workload frequency (and magnitude, to a lesser extent) increased markedly in all axes to counteract turbulence."American Institute of Aeronautics and Astronautics
+RECOMMENDATIONSIn order to validate this model, a series of flight tests should be conducted, in which the helicopter is mounted on a stationary test stand attached to the deck of a ship.Instrumentation should include measurements of the forces and moments acting on the airframe and test stand.The tests should be conducted in a variety of wind-over-deck conditions, with the controls fixed, rotor turning at normal RPM, and the collective set so that the total lift force approximates the aircraft weight.Frequency analysis could then be used to compare the test data with a simulation of the same conditions.The volume in which the airwake is computed is necessarily bounded in order to keep the data storage requirements reasonable.In order to avoid transients at the boundary, the lookup algorithm held the function values at the boundary for positions outside the volume.Over the deck, the top of the volume was at about 51 feet above the deck, extending to 300 feet at 1000 feet from the ship.Although the experimental test plan called for the pilots to stay within the boundaries, they occasionally exceeded these limits in their familiarization flights.Since the volume over the deck was limited to 51 feet above deck height (but the island extends to about 80 feet above the deck), the pilots noticed that they did not reach a freestream airflow in a vertical ascent past the top of the island.In order to make the simulation more realistic, a heuristic fadeout should be applied to the airwake turbulence outside of the volume in which the CFD data are defined.
+CONCLUSIONSSignificant techniques to simulate a helicopter flying through ship airwakes have been developed and demonstrated in piloted simulation.Subjective pilot evaluation has shown that the techniques produce a realistic environment, with appropriate increases in pilot workload.Frequency analysis of acceleration histories show that the simulated airwake produces much more frequency content and total energy than steady simulated winds.Figure 1 .1Figure 1.Non-uniform CFD Grid
+c) 20012001American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.
+c) 20012001American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)1 Sponsoring Organization.
+Figure 3 .Figure 6 . 1 FrequencyFigure 9 .Figure 13 .361913Figure 3. Rolling Acceleration -Wind Azimuth 0 Degrees Figure 4. Rolling Acceleration -Wind Azimuth 60 Degrees
+		
+		
+			
+
+				
+
+
+	
+		UH-60A Black Hawk Engineering Simulation Program: Volume I -Mathematical Model
+		
+			JJHowlett
+		
+	
+	
+		NASA CR-166309, United Technologies, Sikorski Aircraft
+		Stratford, CT
+		
+			December 1981
+		
+	
+	Howlett, J.J., "UH-60A Black Hawk Engineering Simulation Program: Volume I -Mathematical Model," NASA CR-166309, United Technologies, Sikorski Aircraft, Stratford, CT, December 1981.
+
+
+
+	
+		Time-accurate computational simulations of an LHA ship airwake
+		
+			SusanPolsky
+		
+		
+			ChristopherBruner
+		
+		10.2514/6.2000-4126
+		AIAA-2000-4126
+	
+	
+		18th Applied Aerodynamics Conference
+		
+			American Institute of Aeronautics and Astronautics
+			August, 2000
+		
+	
+	Polsky, S., and Bruner, C., "Time-Accurate Computational Simulations of an LHA Ship Airwake", AIAA-2000-4126, August, 2000.
+
+
+
+	
+		Frequency-Response Method for Rotorcraft System Identification: Flight Applications to BO 105 Coupled Rotor/Fuselage Dynamics
+		
+			MarkBTischler
+		
+		
+			MavisGCauffman
+		
+		10.4050/jahs.37.3
+	
+	
+		Journal of the American Helicopter Society
+		j am helicopter soc
+		2161-6027
+		
+			37
+			3
+			
+			July 1992
+			American Helicopter Society
+		
+	
+	Tischler, M.B., and Cauffman, M.G., "Frequency- Response Method for Rotorcraft System Identification: Flight Applications to BO-105 Coupled Rotor/Fuselage Dynamics," Journal of the American Helicopter Society, Vol. 37, No. 3, pp. 3-17, July 1992.
+
+
+				
+			
+		
+	

file110.txt

@@ -0,0 +1,334 @@
+
+	
+	
+		
+In order to develop an aerodynamic database representing a tail-less vehicle, the CFD (computational fluid dynamics) team re-calculated the data using the same geometry but with the tail removed.Arbitrary, undefined control devices (which could be split-ailerons or other asymmetric drag devices) were assumed to replace the yawing moment of the rudder.These surfaces were not modeled in the CFD analysis; rather, the yawing moment coefficient due to the yaw control devices was arbitrarily set to a value that would yield approximately the same yaw control power that the rudder had produced.The flight control system was then modified to augment the directional stability of the tail-less aircraft, and the gains were re-optimized.The aircraft math model was then tested in the Vertical Motion Simulator (VMS) facility at NASA-Ames.This not only showed the quick turnaround possible with the RITE process, but also demonstrated the feasibility of the tail-less re-entry vehicle concept, and provided insight into the required control effectiveness of the yaw control devices.
+RAPID INTEGRATION TEST ENVIRONMENTThe RITE (Rapid Integration Test Environment) process was developed at Ames Research Center to promote rapid turnaround in the aircraft design cycle. 1 In this process, a multi-disciplinary integrated team developed the design using design optimization tools, calculated the aerodynamic data using CFD techniques, designed a flight control system architecture, and optimized the control gains using an off-the-shelf control design software tool.Then the vehicle was simulated in the VMS facility.
+A SHARP RE-ENTRY VEHICLEThe CTV8 vehicle was a modification of the final geometry developed in the RITE-3 project. 2 It is a crew transfer concept vehicle, smaller than the Space Shuttle and intended only for transport of personnel to and from the International Space Station, not for cargo delivery.The CTV8, like its predecessors, incorporates SHARP (Slender Hypersonic Aerothermodynamic Research Probes) technology. 3,4,5By using ultra high temperature ceramics (UHTC) for the leading edges, the leading edges can be made sharper than is the case in current re-entry vehicle design, which yields a higher hypersonic lift-to-drag ratio, thus allowing a larger potential landing footprint.The CTV8 design is shown in Figure 1.
+Figure 1 The CTV8 DesignA modification was made to the control surface configuration of the vehicles tested in RITE-3.In those simulations, it was found that the split-rudder used as a speed brake caused excessive pitch-up moment, which made it difficult to control when the speed brakes were opening.To mitigate this problem, upper and lower body flaps were added for use as speedbrakes.Each of the flaps produced a smaller pitching moment due to their proximity to the plane of the center-of-gravity of the vehicle, and simultaneous deflection of the upper and lower flaps causes their pitching moments to partially cancel each other.
+AERODYNAMIC MATH MODELA simplified aerodynamic math model of the vehicle was developed using computational fluid dynamics simulations 6 .The methods used included both the Navier-Stokes (viscous) and Euler (inviscid) formulations of the flow equations. 7,8A vortex lattice method was used to calculate the dynamic stability derivatives. 9The resulting file was then uploaded to an Internet-based data management system, to allow easy access by all interested parties to the project.The aerodynamic data tables were converted to the format required by the Function Table Processor (FTP) used in the NASA-Ames Vertical Motion Simulator facility.The data tables were then downloaded to the simulation host computer, and processed by the FTP.The FTP compiles the aerodynamic data into code that provides efficient table lookup with linear interpolation for real-time simulation.
+FLIGHT CONTROLSAs in the previous CTV simulations, the flight control system was developed using SimuLink and the CONDUIT® control optimization tools. 10like most other control optimization tools, CONDUIT® accepts flying qualities specifications defined by the user, and attempts to optimize the control gains to meet those specifications.This provides a user-friendly environment, and allows vehicles having different aerodynamic characteristics and/or different control system architectures to be optimized to meet a common set of specifications.The pitch control system used the same Nz-Q architecture as the CTV simulations in the previous year's RITE experiment (see Figure 2).This blended feedback system, previously used in the HL-20 (a reentry vehicle design concept investigated at NASA-Langley), provides an approximate glideslope angle rate command by scaling the normal acceleration by the inverse of the airspeed, and augments the pitch damping with pitch rate feedback.The roll control system provided augmented roll damping by feeding back roll rate to the ailerons (see Figure 3).The aileron command was scheduled inversely with dynamic pressure to maintain relatively constant performance throughout the approach.An "aileron fail" mode was provided, as well, to demonstrate the capability of the RITE process to assist in the development of fault-tolerant flight controls.This mode used the elevators differentially for roll control, as well as symmetrically for pitch control.This capability was implemented in an aileron-elevon control mixer (see Figure 4).
+Figure 4 Aileron-Elevon MixerThe yaw control system used in the previous RITE/CTV simulations had used sideslip rate (beta dot) feedback (see Figure 5).This was found to provide excellent tracking over the runway in the presence of gusty winds.However, in order to do that, the control system caused the aircraft to make rapid yaw corrections in response to the gusts.The astronaut pilots felt that this would be unacceptable, since it would be disorienting to a "de-conditioned" pilot returning from a long space mission.Therefore, alternatives were developed.One of these was to use a complimentary filtering technique to combine the low frequency components of the sideslip rate with the high frequency components of the yaw rate.This worked well; but, somewhat surprisingly, did not seem to have any real advantage over a classical yaw damper, consisting of a simple washed-out yaw rate feedback to the rudder.These different mechanizations were implemented by varying the feedback gains in the yaw control system.In each mechanization, the rudder command was scheduled inversely with dynamic pressure in order to maintain constant performance.
+Figure 5 Yaw Control SystemAs with the previous simulations, a speed control system was included.This system modulated the speedbrake deflection in order to provide an airspeed hold capability (Figure 6).Because the vehicle configuration used upper and lower body flaps as speedbrakes, it was expected that the pitching moment changes produced by speedbrake deflection would be negligible.However, when the astronaut pilots began to fly the simulation, they discovered an objectionable tendency of the aircraft to overshoot the desired attitude in the pre-flare maneuver.It was determined that this was caused by the changing pitching moment produced by the speedbrakes, which were closing because of the increasing drag due to the change in attitude.This was mitigated by scheduling the lower body flap as a function of the average deflection of the upper body flaps in such a way as to minimize the net pitching moment.This was successful, and the tendency to overshoot in the pre-flare was barely detectable with the modified schedule.It should be noted that no design tradeoff was done in this study, so no conclusions have been reached regarding whether a tail-less design would really be better than the more conventional design with a vertical stabilizer and rudder.This study was just a first look at the tail-less option, to see if it would be potentially controllable, and if reasonable handling qualities were potentially achievable.In order to provide an aerodynamic model of the modified vehicle, the CFD team developed a design in which the only difference was that the vertical tail (including both the vertical stabilizer and rudder) was removed (see Figure 7).Due to lack of time and funds, the vehicle geometry was not optimized as a tail-less design.Also, the new yaw control surfaces required to replace the rudder were not designed, nor were they included in the CFD analysis.New CFD simulation runs were made, using the baseline CTV8 configuration with the vertical tail removed, without yaw control devices, and the data were integrated into the piloted simulation.
+Figure 7 The Tail-Less CTVOf course, such a design requires an active stability augmentation system (SAS) to provide adequate weathercock stability.This means that the yaw SAS must operate full-time, and must have adequate redundancy to mitigate the possibility of failure.For the purposes of this preliminary study, it was assumed that suitable yaw control devices (such as "split ailerons" or other differential drag devices) could be developed, having sufficient yaw control power to provide the needed stability and control.Then, if the vehicle could be made stable and controllable, a first cut approximation of the necessary yaw control power could be specified to the designers as a requirement.In order to provide weathercock stability, a sideslip angle feedback path was added to the directional control system, using proportional plus integral compensation (see Figure 8).The architecture of the flight controls for the other axes was not changed; however, the gains for all axes were re-optimized, since the aerodynamic data had changed.
+Figure 8 Modified Yaw Control SystemThe resulting vehicle simulation turned out to have very good handling qualities.In addition, it was determined that the new vehicle also worked well with the "failed aileron" flight control system, using the elevators differentially for roll control, as well as symmetrically for pitch control, as was done with the nominal CTV8 configuration.This was good news, since it meant that the tail-less vehicle would not require ailerons for roll control.That meant that yaw control devices could be installed in place of the CTV8's ailerons.This would, however, remove some control redundancy from the design.Or, if "split ailerons" were used to provide both roll and yaw control, then the elevators could still provide control redundancy as on the CTV8.
+THE SIMULATION EXPERIMENTThe tail-less CTV configuration was simulated in the Vertical Motion Simulator at NASA-Ames Research Center.The simulator cockpit was configured as it would be for the Space Shuttle simulation (Figure 9).The simulation was compared with simulations of the CTV8 and the Space Shuttle.Both CTV configurations were found to be much more maneuverable in roll than the Space Shuttle, and to have excellent handling qualities in all axes.
+Figure 9 Simulator Cab Interior
+CONCLUSIONSThe Virtual Flight -RITE process resulted in a rapid design cycle.The tail-less CTV was developed and tested in piloted flight simulation in less than one week from the time that the decision was made to do it.The simulation experiment showed that a tail-less vehicle similar to the SHARP CTV conceptual designs developed at NASA-Ames would not be difficult to control, and could potentially be made to have excellent handling qualities.It was found that yaw control devices with approximately the same yaw moment control power as the rudder on the CTV8, driven by actuators with time constants of 0.025 seconds, would be adequate to stabilize the tail-less version.However, no analysis was performed to estimate the control power that could be achieved with physically realizable control surfaces.It was also demonstrated that adequate roll control performance could be achieved using the elevators as elevons, so that the ailerons could be replaced by differential drag yaw control devices.
+RECOMMENDATIONSSince the aerodynamics of the assumed yaw control devices were never calculated, any future work on such a vehicle should include the design of possible yaw control devices and calculation of their effect on the aerodynamics.Also, since the CTV8 was never optimized to be a tail-less configuration, optimization should be carried out.The mass distribution for the tailless vehicle should also be determined, and new moments of inertia should be calculated.This would also allow determination of any potential weight savings.Finally, a design tradeoff study should be conducted to compare the costs and benefits of a tailless configuration to the standard configuration with a tail.Figure 22Figure 2 Pitch Command Generator
+Figure 33Figure 3 Roll Command Generator
+Figure 66Figure 6 Speed Control System
+11Pitch RateLead LagQb GainLon StickStick ShapingStick Gain-+ +XElev Gain+ +AutoTrimElev CmdLowDelta NzXLead LagNz GainPassVtGain SchedDyn PressElevator Gain Sched
+Reference Dyn Press Yaw Rate Yaw Rate Gain Washout Filter Beta Dot Rudder Pedals Beta Dot Gain Rudder Pedal Gain + + + Low Pass Filter Kdr Dyn Press Rudder Actuator Rudder
++10 Non-linear Gearing Lt. Upper Body Flap Actuator Rt. Upper Body Flap Actuator Lt. Upper Body Flap Speedbrake CommandEquiv Airspeed+ -Kp+ +Proportional + IntegralCompensationH > 1200VholdKi1/sH < 120015%-0.4+ +Lower Body Flap ActuatorLower Body Flap-10/+3010.-30/Rt. UpperBody Flap
+Reference Dyn Press Rt. Yaw Actuator Rt.Yaw Device Lt. Yaw Actuator Lt.Yaw Device Yaw Rate Yaw Rate Gain Washout Filter Beta Dot Beta Dot Gain Low Pass Filter Rudder PedalsBetaBeta Gain+ + +Rudder Pedal Gain+KyawDyn Press
+		
+		
+
+			
+ACKNOWLEDGEMENTSThe author would like to acknowledge the following people for their significant contributions to this work: Jorge Bardina, Kenny Cheung, Susan Cliff, Arsenio Dimanlig, Ron Gerdes, Veronica Hawke, Jeff Homan, Dave Kinney, Julie Mikula, Joe Ogwell, Chun Tang, Alex Te, Mark Tischler,and Dan Wilkins, as well as the astronaut pilots who participated in the flight simulation tests: Eric Boe, Ken Ham, Charlie Hobaugh, Scott "Doc" Horowitz, Greg "Ray Jay" Johnson, Steve Lindsey, Barry Wilmore, and George Zamka.
+			
+
+			
+
+				
+
+
+	
+		Rapid Integration Test Environment: An Integrated Process for Aircraft Design
+		
+			JohnBunnell
+		
+		
+			JulieMikula
+		
+		10.2514/6.2002-4479
+		AIAA-2002-4427
+	
+	
+		AIAA Modeling and Simulation Technologies Conference and Exhibit
+		Monterey, CA
+		
+			American Institute of Aeronautics and Astronautics
+			August 2002
+		
+	
+	Bunnell, J.W., and Mikula, J.A., "Rapid Integration Test Environment: An Integrated Process for Aircraft Design," AIAA-2002-4427, Monterey, CA August 2002.
+
+
+
+	
+		Vehicle Design of a Sharp CTV Concept Using a Virtual Flight Rapid Integration Test Environment
+		
+			FannyZuniga
+		
+		
+			SusanCliff
+		
+		
+			DavidKinney
+		
+		
+			VeronicaHawke
+		
+		
+			ChunTang
+		
+		
+			StephenSmith
+		
+		10.2514/6.2002-4881
+		AIAA-2002-4881
+	
+	
+		AIAA Atmospheric Flight Mechanics Conference and Exhibit
+		Monterey, CA
+		
+			American Institute of Aeronautics and Astronautics
+			August 2002
+		
+	
+	Zuniga, F.A., Cliff, S.E., Kinney, D.J., Hawke, V.M., and Tang, C.Y., "Vehcile Design of a Sharp CTV Concept Using a Virtual Flight Rapid Integration Test Environment," AIAA-2002-4881, Monterey, CA, August 2002.
+
+
+
+	
+		A reusable space vehicle design study exploring sharp leading edges
+		
+			JamesReuther
+		
+		
+			DavidKinney
+		
+		
+			StephenSmith
+		
+		
+			DeanKontinos
+		
+		
+			PeterGage
+		
+		
+			DavidSaunders
+		
+		10.2514/6.2001-2884
+		AIAA-2001-2884
+	
+	
+		35th AIAA Thermophysics Conference
+		
+			American Institute of Aeronautics and Astronautics
+			June 2001
+		
+	
+	Reuther, J.J., Kinney, D.J., Smith, S.C., Kontinos, D.A., Saunders, D., and Gage, P., "A Reusable Space Vehicle Design Study Exploring Sharp Leading Edges," AIAA-2001-2884, June 2001.
+
+
+
+	
+		Conceptual design of a 'SHARP' - CTV
+		
+			DavidKinney
+		
+		
+			JeffBowles
+		
+		
+			LilyYang
+		
+		
+			CathyRoberts
+		
+		10.2514/6.2001-2887
+		AIAA-2001-2887
+	
+	
+		35th AIAA Thermophysics Conference
+		
+			American Institute of Aeronautics and Astronautics
+			June 2001
+		
+	
+	Kinney, D.J., Bowles, J.V., Yank, L.H., and Roberts, C.D., "Conceptual Design of a 'SHARP' CTV," AIAA-2001-2887, June 2001.
+
+
+
+	
+		Temperature constraints at the sharp leading edge of a Crew Transfer Vehicle
+		
+			DeanKontinos
+		
+		
+			KenGee
+		
+		
+			DineshPrabbu
+		
+		10.2514/6.2001-2886
+		AIAA-2001- 2886
+	
+	
+		35th AIAA Thermophysics Conference
+		
+			American Institute of Aeronautics and Astronautics
+			June 2001
+		
+	
+	Kontinos, D.A., Gee, K., Prabhu, D.K., "Temperature Constraints at the Sharp Leading Edge of a Crew Transfer Vehicle," AIAA-2001- 2886, June 2001.
+
+
+
+	
+		Spatial Convolution Neural Network for Efficient Prediction of Aerodynamic Coefficients
+		
+			TRajkumar
+		
+		
+			JBardina
+		
+		10.2514/6.2021-0277.vid
+	
+	
+		Proceedings of FLAIRS 2002
+		FLAIRS 2002Florida
+		
+			American Institute of Aeronautics and Astronautics (AIAA)
+			2002
+		
+	
+	Rajkumar, T., and Bardina, J., "Prediction of Aerodynamic Coefficients Using Neural Network for Sparse Data," Proceedings of FLAIRS 2002, Florida, 2002.
+
+
+
+	
+		Unsteady aerodynamic simulation of static and moving bodies using scalable computers
+		
+			RobertMeakin
+		
+		
+			AndrewWissink
+		
+		10.2514/6.1999-3302
+		AIAA-99- 3302
+	
+	
+		14th Computational Fluid Dynamics Conference
+		
+			American Institute of Aeronautics and Astronautics
+			
+		
+	
+	Meakin, R.L., and Wissink, A.M., "Unsteady Aerodynamic Simulation of Static and Moving Bodies Using Scalable Computers," AIAA-99- 3302.
+
+
+
+	
+		Robust and efficient Cartesian mesh generation for component-based geometry
+		
+			MJAftosmis
+		
+		
+			MJBerger
+		
+		
+			JEMelton
+		
+		10.2514/3.13918
+		AIAA 97-0196
+	
+	
+		AIAA Journal
+		AIAA Journal
+		0001-1452
+		1533-385X
+		
+			36
+			
+			January 1977
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	Aftmosis, M.J., Berger, M. J., and Melton, J.E., "Robust and Efficient Cartesian Mesh Generation for Component-Based Geometry," AIAA 97-0196, January 1977.
+
+
+
+	
+		A Generalized Vortex Lattice Method for Subsonic and Supersonic Flow Applications
+		
+			LRMiranda
+		
+		
+			RDElliot
+		
+		
+			WMBaker
+		
+		NAS1-12972
+		
+			December 1977
+		
+	
+	NASA CR-2866
+	Miranda, L.R., Elliot, R.D., and Baker, W.M., "A Generalized Vortex Lattice Method for Subsonic and Supersonic Flow Applications," NASA CR- 2866, Contract No. NAS1-12972, December 1977.
+
+
+
+	
+		A multidisciplinary flight control development environment and its application to a helicopter
+		
+			MBTischler
+		
+		10.1109/37.777786
+	
+	
+		IEEE Control Systems
+		IEEE Control Syst.
+		1066-033X
+		1941-000X
+		
+			19
+			4
+			
+			August 1999
+			Institute of Electrical and Electronics Engineers (IEEE)
+		
+	
+	Tischler, M.B., et al, "A Multidisciplinary Flight Control Development Environment and Its Application to a Helicopter," IEEE Control Systems Magazine, Vol. 19, No. 4, pg.22-33, August 1999.
+
+
+
+	
+		Investigation of the launch pad abort capabilities of the HL-20 lifting body
+		
+			EBJackson
+		
+		
+			RobertRivers
+		
+		
+			RajivChowdhry
+		
+		
+			WARagsdale
+		
+		
+			DavidGeyer
+		
+		10.2514/6.1993-3690
+	
+	
+		Flight Simulation and Technologies
+		Langley Research Center, Hampton, VA
+		
+			American Institute of Aeronautics and Astronautics
+			July 1992
+		
+	
+	Jackson, E.B., Cruz, C.L., and Ragsdale, W.A., "Real-Time Simulation Model of the HL-20 Lifting Body," NASA TM-107580, Langley Research Center, Hampton, VA, July 1992.
+
+
+				
+			
+		
+	

file111.txt

@@ -0,0 +1,426 @@
+
+	
+	
+		
+BACKGROUNDThe traditional process for aircraft design is sequential, where each step is completed before the next one begins.This facilitates scheduling each of the facilities (such as wind tunnels and simulators).However, it also guarantees that the knowledge gained during the simulation tests will not be used to refine the aerodynamic configuration -at least not until a subsequent design cycle, or a later model is designed.This is because there is no room in the process for lessons learned in the flight simulation phase to be fed back to the vehicle designers in time to make a difference.Some iteration may be done on the flight control system design during the flight simulation, but usually not on the aerodynamic shape of the vehicle.In order to solve this problem, the simulation phase of the process must be introduced earlier in the design cycle.It is this goal that has driven the RITE project team to develop a new, integrated process, in which pilots (in this case, astronaut-pilots) have an input early in the cycle by evaluating a flight simulation of the design before the design has been finalized.Recent advances in computer speed, computational fluid dynamics technology, and modern control techniques, have made it possible to rapidly make changes to the design, calculate new math model parameters, re-optimize control gains, and integrate the new data into a flight simulator.This allows the simulator test results to be fed back to the vehicle designer, and a modified design to be created and retested during the simulation test period.
+THE RITE PROCESSThe RITE process is an extension of the traditional design and analysis process in that it adds the dimension of piloted flight simulation to the decision making process.In this environment, design cycle times are shortened by using a number of modern techniques, including codes that facilitate rapid development of parametric geometries and the resulting surface and volume grids of vehicle designs; an integrated information system to allow rapid distribution of data; and control design tools to allow rapid re-optimization of the control system parameters when the vehicle design is changed.These capabilities allow design modifications to be accomplished rapidly during piloted flight simulation testing.The RITE process, like traditional development, begins with a conceptual design of the aircraft.The design is American Institute of Aeronautics and Astronautics drawn using a graphic design tool, then lofted in CAD software, and the aerodynamic characteristics of the design are determined using a combination of computational methods and wind tunnel tests.The aerodynamic model is then developed in a form usable in a real-time flight simulator.If the aircraft were powered, a simplified engine model would be developed.Since the particular aircraft studied in this project does not have an engine, this step was bypassed.Next, a flight control system must be developed, and the gains optimized for desired performance.Then the various parts of the model are integrated into the flight simulator, and a simulation experiment is conducted to evaluate the total vehicle performance.Several iterations may be needed to refine the flight controls, after which tests may be performed to determine the handling qualities of the vehicle.The results of these tests are then fed back to the designers, who now have an opportunity to improve the design.CFD simulations are once again used to calculate the aerodynamics of the modified design, the control gains are re-optimized, and the modified vehicle is tested again in the flight simulator.This process is shown in Figure 1.
+VEHICLE DESIGN AND OPTIMIZATIONIn order to develop this process, a conceptual aircraft design of interest to NASA-Ames researchers was chosen as a test case.This aircraft was a Crew Transfer Vehicle (CTV). 1,2A CTV is a re-entry vehicle that could be used to return astronauts to earth following a space station mission -or in the case of a mission abort, possibly due to a launch vehicle failure.This particular vehicle incorporated sharp leading edges, made of high-temperature ceramics, in order to improve the hypersonic lift-to-drag (L/D) ratio. 3The improved L/D would give the vehicle the capability to land within a larger footprint on the earth, thus giving the crew more options. 4e baseline aircraft design for the project, designated V-7, was developed by the Systems Analysis Branch of the Aeronautical Projects and Programs Office at NASA-Ames Research Center.Five modifications were made to this baseline, and each was tested in the Vertical Motion Simulator (VMS) with a pilot-in-theloop.Five of these configurations are shown in Figure 2, where the baseline is shown in the center as CTV0.In the figure, the models are color coded by pressure coefficient at Mach 0.3, at an angle-of-attack of 10 degrees.
+Figure 2. The Aircraft ConfigurationsThere are subtle differences in the configurations shown in Figure 2. Wing twist and camber were modified to produce the CTV1 configuration.In CTV2, the concavity of the upper surface was eliminated.In an attempt to stabilize the Dutch roll mode, more dihedral was added to the configuration of CTV3. 5 CTV4 was developed using an optimization code to vary the wing twist, dihedral and sweep. 6The CTV5 configuration (not shown) was developed during the simulation period, using feedback from the piloted tests.These models were designed at Mach 6 and Mach 0.3 using unsteady aerodynamic shape optimization.
+AERODYNAMIC DATA GENERATIONComputational fluid dynamics (CFD) simulations and wind tunnel tests were used to develop mathematical models of the aerodynamics of several variations of the conceptual vehicle.
+Computational Fluid DynamicsSeveral forms of computational fluid dynamics codes were used in the RITE process.These included a vortex lattice method, as well as both the Euler (inviscid) and Navier-Stokes formulations of the flow equations. 7,8,9e vortex lattice method, simplest and fastest of the computational methods employed, was used to obtain preliminary estimates of the aerodynamics.This method was also used to develop approximations to the dynamic derivatives (such as roll moment due to roll rate), since computing them with the more sophisticated CFD techniques would have been very costly and timeconsuming.The Euler method is computationally faster, and therefore cheaper, than the Navier-Stokes method.However, the Navier-Stokes formulation is usually more accurate, especially when flow separation is a factor.Both methods were used: the Navier-Stokes formulation was used to compute the "clean" aerodynamics (without control surface deflections), and the Euler method was used to compute the control effectiveness and the ground effect model.
+Wind Tunnel TestingOne problem with wind tunnel testing is the delay caused by the need to construct a physical model of the aircraft.This delay was minimized by using a stereolithography technique to create the model.This process uses two computer-aimed lasers, shining into a vat of resin, which cause the resin to harden where the lasers intersect.The hardened resin forms a model for use in the wind tunnel.Two such models were tested manufactured and tested in the Ames 32 inch by 48 inch atmospheric low speed wind tunnel. 10This manufacturing technique is most useful for small-scale, low-speed tests where structural loads on the model are minimized.For tests demanding greater structural strength, other rapid-prototyping techniques are now available to hasten model fabrication.
+DATA TRANSFERAn internet-based data management system was used to allow all members of the design/test team ready access to the aerodynamic data during the development of the mathematical model.The data were then converted to the Function Table Processor (FTP) format used in the VMS simulation facility.The Function Table Processor compiles function table data into a run-time database, with linear interpolation.This database system allows up to seven independent variables, can either use equally spaced or arbitrary breakpoints, and provides a number of features that enhance real-time computational efficiency.
+FLIGHT CONTROLSIn order to test an aircraft design in a real-time, piloted flight simulator, a flight control system model is required.SimuLink® and the CONDUIT® control design tools were used to facilitate the development of suitable control laws to complete the mathematical model of the vehicle.CONDUIT® provides a relatively user-friendly environment for optimizing control system gains to meet flying qualities specifications defined by the control engineers. 11This was essential to the RITE process, as it allowed rapid re-optimization to account for changes to the aerodynamic design.The flight control system design began using a slightly modified version of the HL-20 flight controls. 12The pitch control system, shown in Figure 3, utilized an Nz-Q command, with a blend of pitch rate and normal acceleration feedback.This approximates a flight path command, since the airspeed is held nearly constant.
+Figure 3. Pitch Control SystemThe roll command system, shown in Figure 4, used roll rate damping, together with bank angle command (from the guidance system) and bank angle feedback.
+Figure 4. Roll Control SystemThe yaw control system, shown in Figure 5, originally used washed-out yaw rate feedback to augment the yaw damping.However, it was found that the Dutch roll mode was not sufficiently well damped, so an alternative system, consisting of inertial sideslip and sideslip rate feedback, was tried.This system worked very well, and it was found that the sideslip gain could be set to zero.The resulting system, using only inertial sideslip rate feedback, had the interesting property of automatically compensating for side gusts.
+Figure 5. Yaw Control SystemIn order to provide an airspeed hold function, a split rudder was used as a speed brake.The speed control system, using equivalent airspeed feedback with proportional plus integral compensation, is shown in Figure 6.This system worked well to control airspeed, but introduced objectionable pitch transients in the flare maneuver.
+FLIGHT SIMULATIONA real-time, piloted flight simulation was conducted in the VMS, using astronauts as the subject pilots to test the conceptual vehicle designs in approach and landing tasks on the Kennedy Space Center runway, and the results were fed back to the designers.The various CTV configurations were compared to both the Space Shuttle and the HL-20, a re-entry vehicle concept previously studied at NASA-Langley Research Center. 12The simulator cab was configured exactly as it normally is when simulating the Space Shuttle for astronaut training (Figure 7).
+Figure 7. Simulator Cab InteriorThe Space Shuttle Orbiter simulation was used as a calibration point for the pilots.It was found that there was a significant difference between the handling qualities ratings given by astronauts who had piloted the Space Shuttle on an actual orbital mission, versus those who had been trained in the simulators but had not yet flown the real Space Shuttle.It was found useful to have every pilot fly and rate the Orbiter simulation, (which all had flown in their training) as this provided insight into their ratings of the CTV configurations.A suite of tools known as the Virtual Laboratory (Figure 8) allowed members of the design/test team to participate in the simulation experiments in real time from a remote site. .
+Figure 8. The Virtual LaboratoryThe RITE process then allowed the designers to make changes based on the simulation results, and to perform evaluations with the integrated modifications.Significantly, within a few weeks the flying qualities were improved from barely controllable to excellent.
+LESSONS LEARNEDThere were several different categories of lessons learned during this experiment: information about the specific aircraft; information about the class of aircraft; information about the experiment; and information about the process.About the Specific Configurations Each of the configurations tested was determined to be acceptable, with Level I flying qualities, after proper optimization of the control system.However, some configurations required less control activity than others.This could indicate that those configurations might be able to use less powerful actuators, with consequent weight savings.More details on the results pertaining to each of the configurations have been published in another paper. 5out the Class of Aircraft The CTV is a lifting-body re-entry vehicle, a class of aircraft that usually has a low L/D compared to winged aircraft.It also has no engine, so power cannot be used to control rate of descent at touchdown.This complicates the landing task, and there is no possibility for a go-around.Therefore, the L/D in ground effect is critical to pilot's ability to control the rate of descent at touchdown.Since go-around is not possible, and the pilot is returning from the physical and mental stress of a space mission (possibly aborted), the flying qualities of the re-entry vehicle must be excellent.The pilot-astronauts who participated in this study are among the best pilots in the world, and they are well trained to fly the Space Shuttle.As expected, their performance with the Shuttle landing task was consistently excellent.Nevertheless, they consistently gave the Shuttle mediocre handling qualities ratings, and said that the next generation of re-entry vehicles must have better flying qualities.It was also found that the guidance information on the Head-Up Display (HUD) was critical to the pilot's ability to meet the touchdown performance criteria.Since there is no engine in this vehicle, the trajectory must be accurately followed in order to control touchdown point, rate of descent, and landing airspeed.
+About the ExperimentSince L/D is such an important factor in landing, in simulation it was found that the ground effect model is critical to the pilot's ability to control rate of descent at touchdown.It therefore had a large effect on pilot ratings for the landing task.Initially, there was no plan to develop a new set of ground effect data, but rather to use the ground effect model from the HL-20 simulation for all configurations (except the Space Shuttle).When the astronauts discovered that they were having difficulty controlling the descent rate at touchdown, they suggested two experiments to determine the cause of the problem.First, the ground effect model was "turned off" in the simulation math model.The pilots found that the vehicle was equally difficult to land either way, implying that there was very little effect from the ground effect model.Next, they suggested a test in which the L/D was set to an arbitrarily high value of 6 for the entire run.This was done by setting the drag calculated by the simulation math model equal to the lift divided by 6, without regard for whether such a high value of L/D would be attainable with this type of vehicle.This test resulted in a simulated aircraft that was easy to land.Based on these tests, it was postulated that the difficulty in controlling rate of descent at touchdown was due to lack of a good ground effect model.Runs were made rapidly, using unstructured grid based Euler CFD methods, to generate a more realistic ground effect model.The new ground effect data were found to produce a greater L/D, which improved the touchdown performance significantly.Another finding from the experiment was that the split rudder used for speedbrakes in the simulation produced too much nose-up pitching moment.This required it to be opened slowly and to remain at a fixed deflection when the aircraft was near the runway.Therefore, this speedbrake mechanization was not useful for manual control inputs.Finally, the Rotational Hand Controller used to fly the simulated aircraft caused difficulty for the flight control optimization.This device is the same one that is used in the Space Shuttle.It was developed for maneuvering in space, but the astronauts don't like it for approach and landing.In addition, since there is no Military Specification for such a control inceptor, controller gains had to be determined by trial and error.
+About the ProcessThe process worked well, but it could be improved by having a more systematic test procedure.The different configurations were tested in the simulator by having the pilot perform landings, either straight-in with no winds, offset laterally, or straight-in with gusty winds.The pilot then gave Cooper-Harper handling quality ratings. 13This procedure showed how good each configuration was for the landing tasks, compared with the other configurations.But it did not provide any indication of how the configuration might be improved unless the pilot happened to make some observation (as they did in the case of ground effect) concerning the aerodynamic cause of deficiencies.So, a more systematic procedure should be developed that would have more probability of pointing out areas for possible improvements to the aerodynamic configuration.
+FEEDBACK FROM SIMULATIONDuring the simulation, feedback was provided to the design and CFD teams regarding a number of issues.In one case, when the configuration was first flown, it exhibited excessive adverse yaw.When this characteristic was described to the CFD team, they were able to identify and correct an error that had been made in creating the data.The simulation then showed minimal yaw due to roll.
+American Institute of Aeronautics and AstronauticsAnother issue that was fed back to the CFD team was the ground effect problem.The team was able to rapidly generate ground effect data for the baseline CTV configuration, using the vortex lattice method, and this resulted in significantly better Cooper-Harper ratings.One astronaut commented, "With the new ground effect, my touchdown speeds are slower and my sink rate is slower.Overall, it would be easier."During the experiment, a new configuration (CTV5), based on feedback from the piloted simulation, was implemented.New CFD data were calculated, function tables were generated, the control gains were reoptimized, and the simulation was flown again to evaluate the performance of the new configuration.The original plan was to try to do the re-design over a weekend, and test it in the simulator during the following week.Due to scheduling priorities, it was actually flown for the first time on Thursday, but the complete cycle required less than four days of work.This demonstrated the capability of fast turnaround.The new configuration also incorporated body flaps that could be used as speedbrakes, in an attempt to reduce the pitching moment due to speedbrakes.This new feature was not tried, however, due to lack of time.
+RECOMMENDATIONSSince the RITE process is intended to provide a design team with guidance about how to improve its design, it would be advantageous to use a methodical test procedure that would show what could be improved.In this experiment, the test pilots were able to discover some potential improvements.However, there was no systematic procedure to look for areas of potential improvement to the aerodynamics of the vehicle.Some method should have been devised to systematically investigate what aerodynamic changes to the vehicle might improve its performance.With hindsight, it is postulated that a procedure could be developed in which incremental changes to the various aerodynamic tables could be programmed into the simulation math model code.By evaluating the effect of each incremental change, data could be produced that would show the design team where improvements might be made.For example, an increment could be added to the roll damping in the rolling moment equation.If this hypothetical aerodynamic data set produced a better Cooper-Harper rating, that could suggest that the design might be improved by increasing the roll damping.In fact, such a procedure was used to some extent to investigate the ground effect problem in this experiment, by arbitrarily varying the L/D of the simulated vehicle, as described previously.The approach proved to be very useful, and will be incorporated into future RITE experiments.
+CONCLUSIONSThis project has demonstrated the feasibility of the Rapid Integration Test Environment process for aircraft design.The strength of the RITE process is that it allows the vehicle designers to get feedback about the vehicle configuration before the design must be finalized.It also provides a way for pilots to be involved in the design process.This approach has a number of benefits to the design process.First, it facilitates the rapid discovery of any errors that may have occurred in the calculation of the aerodynamic data.Second, it aids in the identification of any other math model deficiencies.Third, it provides insight into the handling qualities of the design, and allows the designers to make improvements and tradeoffs.For these reasons, the RITE process should become the standard for aircraft design.Figure 1 .1Figure 1.The RITE Process
+AIAA Modeling and Simulation Technologies Conference and Exhibit 5-8 August 2002, Monterey, California AIAA 2002-4479 2
+		
+		
+
+			
+ACKNOWLEDGEMENTSThe authors would like to acknowledge the following people for their significant contributions to this work: Fanny Zuniga, Dave Kinney, Steve Smith, Veronica Hawke, Chun Tang, Susan Cliff, Jorge Bardina, Joe Ogwell, and Dan Wilkins, as well as the test pilots and astronauts who participated in the flight simulation tests.
+			
+
+			
+
+				
+
+
+	
+		A reusable space vehicle design study exploring sharp leading edges
+		
+			JamesReuther
+		
+		
+			DavidKinney
+		
+		
+			StephenSmith
+		
+		
+			DeanKontinos
+		
+		
+			PeterGage
+		
+		
+			DavidSaunders
+		
+		10.2514/6.2001-2884
+		AIAA-2001-2884
+	
+	
+		35th AIAA Thermophysics Conference
+		
+			American Institute of Aeronautics and Astronautics
+			June 2001
+		
+	
+	Reuther, J.J., Kinney, D.J., Smith, S.C., Kontinos, D.A., Saunders, D., and Gage, P., "A Reusable Space Vehicle Design Study Exploring Sharp Leading Edges," AIAA-2001-2884, June 2001.
+
+
+
+	
+		Conceptual design of a 'SHARP' - CTV
+		
+			DavidKinney
+		
+		
+			JeffBowles
+		
+		
+			LilyYang
+		
+		
+			CathyRoberts
+		
+		10.2514/6.2001-2887
+		AIAA 2001-2887
+	
+	
+		35th AIAA Thermophysics Conference
+		
+			American Institute of Aeronautics and Astronautics
+			June 2001
+		
+	
+	Kinney, D.J., Bowles, J.V., Yank, L.H., and Roberts, C.D., "Conceptual Design of a 'SHARP'- CTV," AIAA 2001-2887, June 2001.
+
+
+
+	
+		Temperature constraints at the sharp leading edge of a Crew Transfer Vehicle
+		
+			DeanKontinos
+		
+		
+			KenGee
+		
+		
+			DineshPrabbu
+		
+		10.2514/6.2001-2886
+		AIAA-2001- 2886
+	
+	
+		35th AIAA Thermophysics Conference
+		
+			American Institute of Aeronautics and Astronautics
+			June 2001
+		
+	
+	Kontinos, D.A., Gee, K., Prabhu, D.K., "Temperature Constraints at the Sharp Leading Edge of a Crew Transfer Vehicle," AIAA-2001- 2886, June 2001.
+
+
+
+	
+		Crew Transfer Vehicle trajectory optimization
+		
+			DavidSaunders
+		
+		
+			GaryAllen, Jr.
+		
+		
+			PeterGage
+		
+		
+			JamesReuther
+		
+		10.2514/6.2001-2885
+		AIAA-2001-2885
+	
+	
+		35th AIAA Thermophysics Conference
+		
+			American Institute of Aeronautics and Astronautics
+			June 2001
+		
+	
+	Saunders, D., Allen, G. Jr., Gage, P., Reuther, J.J., "Crew Transfer Vehicle Trajectory Optimization," AIAA-2001-2885, June 2001.
+
+
+
+	
+		Vehicle Design of a Sharp CTV Concept Using a Virtual Flight Rapid Integration Test Environment
+		
+			FannyZuniga
+		
+		
+			SusanCliff
+		
+		
+			DavidKinney
+		
+		
+			VeronicaHawke
+		
+		
+			ChunTang
+		
+		
+			StephenSmith
+		
+		10.2514/6.2002-4881
+		AIAA-2002-4881
+	
+	
+		AIAA Atmospheric Flight Mechanics Conference and Exhibit
+		Monterey, CA
+		
+			American Institute of Aeronautics and Astronautics
+			August 2002
+		
+	
+	Zuniga, F.A., Cliff, S.E., Kinney, D.J., Hawke, V.M., and Tang, C.Y., "Vehicle Design of a Sharp CTV Concept Using a Virtual Flight Rapid Integration Test Environment," AIAA-2002-4881, Monterey, CA, August 2002.
+
+
+
+	
+		Aerodynamic Shape Optimization Using Unstructured Grid Methods
+		
+			SusanCliff
+		
+		
+			ScottThomas
+		
+		
+			TimothyBaker
+		
+		
+			AntonyJameson
+		
+		
+			RaymondHicks
+		
+		10.2514/6.2002-5550
+		AIAA-2002-5550
+	
+	
+		9th AIAA/ISSMO Symposium on Multidisciplinary Analysis and Optimization
+		Atlanta, GA
+		
+			American Institute of Aeronautics and Astronautics
+			September 2002
+		
+	
+	Cliff, S.W, Thomas, S.D, Baker, T.J., Jameson, A., and Hicks, R.M., "Aerodynamic Shape Optimization Using Unstructured Grid Methods," AIAA-2002-5550, Atlanta, GA, September 2002.
+
+
+
+	
+		A Generalized Vortex Lattice Method for Subsonic and Supersonic Flow Applications
+		
+			LRMiranda
+		
+		
+			RDElliot
+		
+		
+			WMBaker
+		
+		NAS1-12972
+		
+			December 1977
+		
+	
+	NASA CR-2865
+	Miranda, L.R., Elliot, R.D., and Baker, W.M., "A Generalized Vortex Lattice Method for Subsonic and Supersonic Flow Applications," NASA CR- 2865, Contract No. NAS1-12972, December 1977.
+
+
+
+	
+		Robust and efficient Cartesian mesh generation for component-based geometry
+		
+			MJAftosmis
+		
+		
+			MJBerger
+		
+		
+			JEMelton
+		
+		
+			MJAftosmis
+		
+		
+			MJBerger
+		
+		
+			JEMelton
+		
+		10.2514/6.1997-196
+		AIAA 97-0196
+	
+	
+		35th Aerospace Sciences Meeting and Exhibit
+		
+			American Institute of Aeronautics and Astronautics
+			January 1977
+		
+	
+	Aftosmis, M.J., Berger, M.J., and Melton, J.E., "Robust and Efficient Cartesian Mesh Generation for Component-Based Geometry," AIAA 97-0196, January 1977.
+
+
+
+	
+		Unsteady aerodynamic simulation of static and moving bodies using scalable computers
+		
+			RobertMeakin
+		
+		
+			AndrewWissink
+		
+		10.2514/6.1999-3302
+		AIAA-99- 3302
+	
+	
+		14th Computational Fluid Dynamics Conference
+		
+			American Institute of Aeronautics and Astronautics
+			
+		
+	
+	Meakin, R.L., and Wissink, A.M., "Unsteady Aerodynamic Simulation of Static and Moving Bodies Using Scalable Computers," AIAA-99- 3302.
+
+
+
+	
+		Low speed aerodynamics and landing characteristics of Sharp-class Crew Transfer Vehicle concepts
+		
+			StephenSmith
+		
+		
+			JamesReuther
+		
+		
+			DavidKinney
+		
+		
+			DavidSaunders
+		
+		10.2514/6.2001-2888
+		AIAA-2001-2888
+	
+	
+		35th AIAA Thermophysics Conference
+		
+			American Institute of Aeronautics and Astronautics
+			June 2001
+		
+	
+	Smith, S., Reuther, J., Kinney, D., and Saunders, D., "Low Speed Aerodynamics and Landing Characteristics of Sharp-Class Crew Transfer Vehicle Concepts," AIAA-2001-2888, June 2001.
+
+
+
+	
+		A multidisciplinary flight control development environment and its application to a helicopter
+		
+			MBTischler
+		
+		10.1109/37.777786
+	
+	
+		IEEE Control Systems
+		IEEE Control Syst.
+		1066-033X
+		1941-000X
+		
+			19
+			4
+			
+			August 1999
+			Institute of Electrical and Electronics Engineers (IEEE)
+		
+	
+	Tischler, M. B., et al, "A Multidisciplinary Flight Control Development Environment and Its Application to a Helicopter," IEEE Control Systems Magazine, Vol. 19, No. 4, pg. 22-33, August 1999
+
+
+
+	
+		Investigation of the launch pad abort capabilities of the HL-20 lifting body
+		
+			EBJackson
+		
+		
+			RobertRivers
+		
+		
+			RajivChowdhry
+		
+		
+			WARagsdale
+		
+		
+			DavidGeyer
+		
+		10.2514/6.1993-3690
+	
+	
+		Flight Simulation and Technologies
+		
+			American Institute of Aeronautics and Astronautics
+			
+		
+	
+	Jackson, E.B., Cruz, C.I., and Ragsdale, W.A., "Real-Time Simulation Model of the HL-20
+
+
+
+	
+		Prospective Futures of Civilian Air Transportation
+		
+			DennisMBushnell
+		
+		10.30919/es8d565
+	
+	
+		Engineered Science
+		Eng. Sci.
+		2576-988X
+		2576-9898
+		
+			July 1992
+			Engineered Science Publisher
+			Langley Research Center, Hampton, VA
+		
+	
+	Lifting Body
+	Lifting Body," NASA TM-107580, Langley Research Center, Hampton, VA, July 1992.
+
+
+
+	
+		The Use of Pilot Rating in the Evaluation of Aircraft Handling Qualities
+		
+			GECooper
+		
+		
+			RPHarper
+		
+		
+			Jr
+		
+		NASA TN D-5153
+		
+			April 1969
+		
+	
+	Cooper, G.E., and Harper, R.P., Jr., "The Use of Pilot Rating in the Evaluation of Aircraft Handling Qualities," NASA TN D-5153, April 1969.
+
+
+				
+			
+		
+	

file112.txt

@@ -0,0 +1,238 @@
+
+	
+	
+		
+negatively impacts the NAS by increasing controller workload along with aircraft fuel burn and emissions.Departuretime approval requests (APREQs) provide center-approved departure times to allow for smooth stream insertion.For many years APREQs have used land-line voice communications.Each day en-route centers send a generalinformation message to certain towers indicating that Call-For-Release (CFR) is required for departures to specific destinations.When an affected flight is ready to depart, the control tower traffic manager calls the adjacent en-route center to request approval for a time that reflects the best estimate of when the flight will be able to depart.The center traffic manager enters the requested time in TBFM and responds with a departure time predicted to enable the flight to fit into the overhead stream of traffic.Tower controllers maneuver the aircraft on the airport surface to meet the time; the FAA considers the eventual release compliant if the departing aircraft's takeoff rotation is within two minutes prior to one minute after the approved time.ATD-2 draws from prior NASA research geared toward replacing CFR with electronic center-tower coordination for APREQ scheduling.Most recently, the NASA Precision Departure Release Capability [2,4,5] led to the FAA's Integrated Departure Arrival Capability (IDAC) implemented within TBFM.IDAC includes departure-demand monitoring, slot identification, and semi-automatic and automatic modes for requesting release times from towers equipped with the Integrated Departure Scheduling Tool (IDST) [6].The tower component of the ATD-2 surface system, called the Surface Trajectory-Based Operations (STBO) Client, encompasses the IDST functionality.In addition, the STBO Client provides the capability to leverage a surface traffic schedule and airline-provided Earliest Off-Block Times (EOBTs) to calculate the Earliest Feasible Takeoff Time (EFTT) that the tower traffic manager should request for a given flight.Thus, release-time requests made using the STBO Client can consider not only slotavailability in the overhead stream, but also the feasibility of departing at the requested time.CLT Tower traffic managers began using the fielded ATD-2 system to electronically negotiate APREQ times with Washington Center (ZDC) in November 2017.For the initial 41-day introductory period from 23 November 2017 to 2 January 2018, electronic coordination was used for more than half of eligible flights, and ZDC traffic managers approved electronic requests, on average, in less than one minute [7].These data also showed that average compliance with electronically negotiated release times and the average tactical delay assigned did not differ significantly from those of release times coordinated using CFR.In addition, traffic managers also used electronic negotiation to reschedule release times.The present research extends the analysis in Ref. [7] to examine APREQs during daily operations at CLT from 1 January 2018 to 28 February 2019.During this period the ATD-2 system underwent numerous enhancements, and in October 2018 operations expanded to include electronic APREQ negotiation with Atlanta Center (ZTL).In addition to providing comparisons with prior results, this paper examines APREQ rescheduling and compliance improvements.The paper first provides background on APREQs at CLT and electronic APREQ negotiation using the STBO Client.It then presents the results of the analysis, followed by conclusions and topics for future investigation.
+II. BackgroundThe ATD-2 system became operational at CLT in September 2017.The CLT airport surface layout is shown in Fig. 1.During typical operations runway 18R/36L is dedicated to arrivals, 18C/36C is primarily dedicated to departures, and 18L/36R serves both arrivals and departures.Surface traffic management challenges at CLT stem from limited ramp area, the dual-use runway, and arrivals taxiing across the dedicated departure runway.Construction on the 5/23 runway has prevented southconverging operations since May 2018.A key IADS information-sharing and coordination focus area entails linking Traffic Management Initiatives (TMIs) developed using TFMS to the TBFM scheduling capabilities.TMIs implemented to manage demandcapacity imbalances include ground delay programs, ground stops, required re-routes, miles-in-trail restrictions, Expect Departure Clearance Times (EDCTs), and APREQs [8].Like APREQs, EDCTs are controlled departure times, but EDCTs are imposed NAS-wide by the FAA Command Center, and have a larger compliance window, from five minutes earlier to five minutes later than the assigned departure time.Tactical departure scheduling via APREQs is particularly important at CLT owing to its location underneath busy overhead traffic streams entering ZDC and ZTL airspace.ZDC and ZTL impose daily APREQ restrictions on CLT departures to busy airports such as the New York metroplex airports and Atlanta.Complying with approved release times helps the CLT departures merge smoothly into packed traffic flows (Fig. 2).The STBO Client (Fig. 3) supports APREQ TMIs through specialized display of relevant information on its runway timelines.Similar to IDST, STBO Client timelines depict green and red areas that reflect where slots are available and unavailable, respectively, in the relevant center's TBFM schedule (Fig. 4).This helps the tower maintain awareness of the center's demand and request release times the center is likely to approve.Moreover, because the timelines also show runway demand, including arrivals, the tower traffic manager can request release times that account for other surface traffic management considerations, potentially increasing the likelihood of compliant releases.Timeline symbology for aircraft subject to APREQs also indicates whether semi-automatic or automatic electronic-negotiation modes are available, or whether circumstances dictate the use of CFR.The principal difference between semi-automatic and automatic coordination is that under automatic mode TBFM/IDAC automatically sends an approved release time back to the tower STBO Client without input from the center traffic manager.In either mode, the tower traffic manager can right-click the flight data tag on the STBO Client timeline and choose one of two release-time request methods from a context menu: 'Select Slot on Timeline' or 'Request Release Time.'The former enables the traffic manager to then click within the red/green area of the timeline to transmit a requested time to the center, while the latter directs STBO to automatically choose an EFTT and request it.Active requests are indicated with a yellow arrow next to the flight's data tag.At the center, TBFM/IDAC produces an audible alert and highlights the flight on the TBFM timeline.Under semi-automatic mode, the center traffic manager can adjust the requested time before sending an approved time to the tower.The timeline symbology for a flight changes to reflect receipt of an approved release time.If the approved time differs from the originally requested time, the STBO Client produces both audible and visual alerts.The tower traffic manager can acknowledge the new time and clear the alert symbol by clicking it or selecting a context-menu item.To be compliant APREQ flights must depart the runway within a compliance window from two minutes earlier to one minute later than the approved release time.Some flights may be subject to both EDCT and APREQ restrictions; the STBO Client also shows EDCT compliance windows for selected flights (Fig. 5), so that requested times can also honor the EDCT compliance window.Once a flight has an approved release time (or times), the STBO Client colorcodes the labels at the end of the flight's data tag according to the flight's projected compliance (see Ref. [7]).The compliance indications aid the tower traffic manager in identifying flights that may benefit from a rescheduled release time.Circumstances may also arise in which a specific flight may be excluded from an APREQ restriction, or have a previously approved release time removed (referred to as a 'free release'); the STBO Client also supports these operations.The ATD-2 deployment at CLT began with a focus on efficiency and predictability improvements in airport surface and departure operations enabled by data integration and sharing, surface movement scheduling, and tactical departure scheduling.Additional system-integration elements, including integration with Advanced Electronic Flight Strips (AEFS), the aforementioned introduction of IDAC at ZTL, extending the scope of 'prescheduling' operations with ZTL, and surface scheduler improvements have all contributed toward improved APREQ management: AEFS automatically shows APREQ release times to tower controllers, IDAC at ZTL further reduces the need for CFR operations, and scheduler enhancements improve pushback-time advisories.Prescheduling refers to assigning release times based on a flight's airline departure time, rather than waiting until the pilot calls to indicate the flight is ready to push back from the gate; ZTL has implemented prescheduling operations with CLT for many years.The analysis presented in the following section highlights some of these impacts.
+III. Field-Data AnalysisATD-2's data-integration focus has yielded a rich, electronically-logged data set covering the January 2018 through February 2019 study period.This section first generally describes CLT operations and data included for analysis, then presents a series of results pertinent to electronic APREQ negotiation.
+A. CLT OperationsThe raw data for the 423-day study period include 627,516 CLT flight operations (313,984 arrivals and 313,532 departures).To focus the analysis on normal operations, calendar days with departure counts outside [1.5 * interquartile range] were identified; removing those 21 days from consideration leaves 402 days encompassing 303,729 departures.Table 1 describes the distribution of departures per day for the reduced data set used for subsequent analyses.CLT is a major hub for American Airlines (AAL); most CLT departures are operated by AAL and its regional carriers.CLT flights to Atlanta, Newark, LaGuardia, and John F. Kennedy airports are subject to APREQ restrictions throughout each day, with other major destinations including Chicago O'Hare, Washington Dulles, and Philadelphia also frequently subject to APREQs.All are among the top ten most frequent destinations of CLT departures.Scheduled AAL operations at CLT are organized into banks, which leads to periods of surface congestion interspaced with lulls.Fig. 6 shows the departure-bank structure reflected in the departure runway-utilization local time, summed over all days in the 402-day data set.
+B. APREQ and EDCT DeparturesFrom January 2018 through February 2019, there were 32,337 flights (10.6% of all departures) with controlled release times due to APREQs, EDCTs, or both.Of these, 26,733 flights (8.8% of all departures) were subject to APREQ restrictions, including those also subject to EDCTs.Fig. 7 shows the counts and proportions of controlled departures in each category.Overall, 82.6% of controlled departures were APREQ flights.More APREQ flights were negotiated with ZDC (61.1%) than with ZTL (38.9%).The larger proportion of APREQ flights negotiated with ZDC reflects the large number of flights to the U.S. Northeast that are subject to daily APREQ restrictions from ZDC.Fig. 8 shows the airport configuration in use at takeoff during each month of the study period for 26,436 APREQ flights (98.9% of all APREQ flights) for which these data were available.The south-converging ('South_Conv') flows utilizes runway 23 for arrivals, which adds complexity to surface traffic management.Due to the 5/23 construction noted above, the predominant airport configuration for the latter part of the study period was the more standard north-flow, with 36C and 36R used for departures.The north-flow configuration affords more room for APREQ flights to wait on the airport movement area away from ramp-area congestion; by contrast, the south-flow runways (18C and 18L) are considerably closer to the ramp area surrounding the main terminal building near the top of Fig. 1.Flights through ZDC are likely to use the eastern departure runways (18L/36R), while flights through ZTL are likely to depart from the western departure runway (18C/36C).Overall, 22.8% of APREQ flights used runway18L and 36.3%used 36R, while 15.8% used 18C and 25.1% used 36C.The higher utilization of 18L and 36R again reflects the typical use of the eastern runway for ZDC APREQs.
+C. Electronic APREQ Coordination and Release Time Request MethodsTo perform electronic APREQ negotiation, the center traffic manager must first enable it in TBFM IDAC by specifying whether semi-automatic or automatic mode should be used, or whether CFR is required.When semiautomatic or automatic mode is available, tower traffic managers have the option to select the desired release time manually via the 'Select Slot on Timeline' (SSOT) method or allow the STBO Client to automatically request a release time using 'Request Release Time' (RRT).Figs. 9 and 10 show the methods used by CLT tower traffic managers to negotiate release times in semi-automatic or automatic mode with ZDC and ZTL, respectively.'OFF' indicates electronic negotiation was turned off, so that CFR was required.Center release-mode data became available in March 2018, and except for a few test periods beginning in July 2018, ZTL only used CFR prior to October 2018 when IDAC was officially introduced there.Semi-automatic mode was used predominately at both centers, indicating a desire on the part of center traffic managers to approve requested release times manually; however, both centers increased the use of automatic mode toward the end of the study period.Anecdotal evidence suggests some center traffic managers may prefer the flexibility to add slack to schedules under certain circumstances (e.g., if they anticipate unscheduled flights or expect flights will require scheduling soon); this requires semi-automatic mode.The ZTL data also depict the introduction of prescheduling, in which the ATD-2 system automatically requests release times for ATL flights.All of the ZTL 'Request Release Time' usage in automatic mode stems from prescheduling; all but a small fraction stems from prescheduling in semi-automatic mode (note large proportions of 'SEMI, RRT' and 'AUTO, RRT' in Fig. 10).The ZDC release methods (Fig. 9), on the other hand, directly reflect user preference, indicating increased use of the 'Request Release Time' method than during the introductory period for electronic APREQ negotiation described in Ref. [7].The use of CFR even when semi-automatic or automatic modes were available may indicate some discussion about particular APREQ flights was warranted.Traffic managers resorted to CFR less frequently in recent months.
+D. APREQ ReschedulingTower traffic managers may request a new release time for a previously scheduled APREQ flight if it appears the flight will be unable to comply with its current release time, or if the opportunity arises to meet an earlier time and incur less delay.Of the 26,733 APREQ flights, tower traffic managers renegotiated release times for 6,936 flights (25.9%) and the rescheduling process led to a new release time for 6807 flights (25.5%).Removing release-time-difference outliers beyond [1.5 * interquartile range] yields 6,373 flights with new release times.Table 2 describe release-time differences (final release timeinitial release time) for these flights, so that a positive difference indicates the flight was  rescheduled to a later time.1,935 of these flights (30.4%) were rescheduled to an earlier time (mean= -431.6 s; SD=306.1 s); for those with initial and final release times both negotiated via IDAC, the total delay savings over the study period was 73.8 hrs.Data for comparison with rescheduling using CFR are unavailable.Another possible reason to reschedule an APREQ is to better ensure EDCT compliance.However, on a percentage basis, APREQ flights that were also subject to an EDCT were rescheduled approximately as often as APREQ flights not subject to an EDCT (26.9% vs. 25.8%,respectively).In some circumstances center traffic managers may simply release a flight that is nominally subject to an APREQ restriction.So-called 'free releases' occurred for 269 APREQ flights (1%) during the study period.
+E. APREQ Aircraft LocationsThe ATD-2 surface system records the estimated 'surface state' of flights, which can be used to identify where APREQ flights were during the APREQ negotiation process.Tower traffic managers are expected to request a release time for flights after the pilot calls ready and before the aircraft has initiated the pushback operation (i.e., while the aircraft is still at the gate in the 'SCHEDULED' state).Excluding prescheduled flights that are always at the gate when prescheduling occurs, Fig. 11 shows that tower traffic managers received the majority of initial release times before the flight started taxiing.An apparent trend toward obtaining release times later, during pushback, may actually reflect enhancements made to the ATD-2 system that results in earlier detection of pushback events from surfacesurveillance data.Fig. 12 shows the surface states of APREQ flights upon receipt of renegotiated release times.The majority of release times are rescheduled while flights are in the 'TAXI_OUT' state in the active movement area prior to reaching the runway queue.It is possible that rescheduling of flights in 'TAXI_OUT' or 'IN_QUEUE' states is triggered based on the STBO Client's projected compliance information.The 'IN_QUEUE' state may reflect aircraft that are actually parked out of the main runway queue.Data from the later months in the study period show an increased number of flights had renegotiated release times, in part due to increased rescheduling via IDAC at ZTL.The reduced number of rescheduled APREQ flights in September 2018 warrants further investigation.Table 3 depicts the initial and final surface states for the 6,213 APREQ flights that were not prescheduled, but were later rescheduled.The greatest proportion (24.4%) first had a release time negotiated at the gate ('SCHEDULED') and then renegotiated in the movement area ('TAXI_OUT').13.3% of flights were assigned updated release times in the ramp area, whereas 14% were in the runway queue.11.8% of flights registered a rescheduled release time prior to pushback.The table shows that most flights were scheduled at the gate, but possible compliance issues that warranted rescheduling did not arise until flights attained the 'TAXI_OUT' or 'IN_QUEUE' states.The STBO Client compliance projections are likely to be more accurate by this time.
+F. APREQ Assigned DelaysThe delay assigned to APREQ flights was computed using the last-updated airline expected departure time ('Ltime') for correspondence with Ref. [7], as shown in Eq. ( 1):
+APREQ delay = Final approved release time -Ltime(1)  25,573 APREQ flights (95.7%) have a valid Ltime; removing delay-value outliers beyond [1.5 * inter-quartile range] yields 24,267 values covering 90.8% of all APREQ flights.Fig. 13 shows the resulting APREQ-delay histogram with one-minute bins for release times negotiated with ZDC and ZTL.The ZDC APREQ-delay distribution (N=14,693; mean=23.9mins; SD=8.3 mins) is similar to the ZTL APREQ-delay distribution (N=9,462; mean=22.1 mins; SD=8.3 mins).The slightly higher mean delay for ZDC may reflect the larger number of APREQs through ZDC to the northeast U.S. The overall APREQ-delay distribution has a mean of 23.2 mins (SD=8.4mins), which corresponds closely to the APREQ-delay distribution for the introductory electronic-negotiation period [7].A trend of slightly lower median delays for ZTL APREQ flights than ZDC APREQ flights, with comparable variation, holds when examining the data along several dimensions.APREQ flights that are also subject to EDCT restrictions show slightly higher median APREQ delay (Fig. 14; whisker end-points are at [1.5 * inter-quartile range] from the box edges); this may indicate later release-times are more commonly requested to meet EDCT restrictions.There is no apparent difference in APREQ delay by departure bank or release-time request method, with slightly lower median APREQ delays incurred by ZTL flights.The data also show slightly higher median delays for rescheduled APREQ flights compared non-rescheduled APREQ flights, in accordance with the tendency to reschedule APREQ flights to a later release time.Median assigned delay tends to be lower for the runways not typically used for departures to the respective centers, a possible effect of APREQs assigned during lower-traffic periods when departure-traffic direction is less critical.Flights that received an approved release time prior to taxing also incurred less median APREQ delay.For brevity, plots of these results are not shown.
+G. APREQ ComplianceOverall, 17,854 APREQ flights (66.8 %) were compliant with their assigned departure release times (within two minutes before and one minute after the assigned time).Fig. 15 shows the monthly compliance percentage over the study period.A trend toward improved compliance is evident, with monthly compliance reaching 71.8% in January 2019.To confirm the trend, Fig. 16 shows average compliance computed using a rolling window over 10,000 individual APREQ flights and smoothed by taking every 100 values.Fig. 16 shows a clear trend toward increasing compliance that extends to the end of the study period.Compliance was also examined along various dimensions in a manner similar to APREQ delay, considering the same 24,267 APREQ flights that remain after removing delay outliers.Compliance for ZTL flights is generally slightly higher than for ZDC flights.This trend holds, for example, when examining non-rescheduled versus rescheduled flights; otherwise there is no apparent difference in compliance.Electronic release-time request methods also show a limited positive effect on the release-time-compliance percentage over CFR (Fig. 17), similar to the results in Ref. [7].One case in which the compliance percentage for ZTL flights was lower than that of ZDC flights was for APREQ flights that were also subject to an EDCT restriction (Fig. 18).Median delay values were slightly higher for such flights, as shown in Fig. 14.The compliance percentage was also lower for ZTL flights that used runways not typically assigned to ZTL flights (Fig. 19).Finally, Fig. 20 depicts average APREQ compliance by APREQ delay grouped in five-minute bins (axis labels indicate APREQ delay was less than or equal to the labeled value).Again compliance appears relatively insensitive to the amount of assigned delay, with ZTL enjoying a slight advantage in compliance.Compliance was worst for flights with delays of five minutes or less.Taken together, these findings indicate the main drivers of APREQ compliance lie elsewhere, potentially in the context of flight-specific surface operations.
+H. Approval Response TimesAn important advantage of electronic release-time negotiation is the time savings relative to CFR [7].Using IDAC message data available from 13 February 2018 to the end of the study period, approval response times for electronically negotiated release times were computed as the time difference between a request message and the corresponding approval message for a particular flight.Response times and associated electronic release-time request method were obtained for 12,241 APREQ flights (45.8% of all APREQ flights); for rescheduled flights, the computed response time is that of final renegotiation.The overall median response time was 9 secs.Fig. 21 depicts the response-time distributions for each center and release-time request type (whisker end-points are at [1.5 * inter-quartile range] from the box edges).Median response times are slightly lower for 'Request Release Time' requests for both centers.Median response times are slightly lower for ZTL than for ZDC, with slightly lower variation-another factor that could impact observed compliance.As discussed in Ref. [7], response times are consistently better than CFR response times that might be experienced during busy periods, which can exceed five minutes.
+IV. Conclusions and Further ResearchThis paper documents ATD-2 electronic APREQ negotiation in daily operations at CLT over fourteen months.The analysis indicates that field traffic managers are consistently exercising capabilities provided by the STBO Client and TBFM IDAC to good effect.Electronic departure-approval requests from CLT to ZDC and ZTL have largely supplanted CFR.In addition, compliance is improving, supported in part by the capability to reschedule release times electronically.Detailed examination of assigned APREQ delays suggests that the delay assigned to a flight via a given release time is not obviously affected by bank, restrictions, release-time request method, or other factors.It may therefore depend primarily on the demand at the stream-insertion points used as scheduling points by each center.Rescheduling APREQ flights typically results in slightly higher median delay, but also provides delay savings for a sizable proportion of flights.Renegotiating release times may also contribute to improved compliance by providing more achievable release times.The results also suggest an APREQ flight's bank, the method used to negotiate its release time electronically, and whether it was also subject to an EDCT or had its release time rescheduled do not significantly impact APREQ compliance-nor does the amount of assigned delay.This may indicate a variety of specific contextual factors related to surface traffic movement, pilot response, and situation awareness and skill of CLT Tower controllers also play a significant role.Normal use of a mainly departure-only runway, coupled with the capability to easily renegotiate release times for flights projected to miss their assigned times, may bolster the compliance of ZTL APREQ flightswhich in turn has contributed to improved overall compliance since the introduction of IDAC at ZTL.Overall, these promising findings support future, broader deployments of similar capabilities because no specific systematic factors appear to negatively impact compliance with electronically negotiated release times.The ATD-2 ground system has clearly contributed to streamlining release-time requests and improving APREQ compliance during the operational period examined here.STBO Client features, including projected compliance indications, EDCT compliance windows, and APREQ exclusions, likely provide incremental advantages that are difficult to discern at the aggregate level.Electronic release-time negotiation also provides significant time savings in approving release times for both tower and center traffic managers, consistent with the response-time data in Ref. [7].Additional research is needed to examine the impacts of specific ATD-2 enhancements on APREQ compliance.For example, automatically propagating approved release times to AEFS's flight strips may afford tower controllers advanced notice, and enable them to better formulate plans for managing APREQ flights.Future research will examine additional effects of APREQ compliance on other important IADS metrics (e.g., arrival-time compliance), and apply more sophisticated analyses to determine the contributions of specific ATD-2 system enhancements on compliance improvements.Fig. 11CLT airport surface layout.
+Fig. 2 Fig. 3 STBO23Fig. 2 Non-compliant APREQ departures resulting in excessive vectoring in ZDC airspace (left) versus smooth stream-insertion of compliant APREQ departures (right).
+Fig. 44Fig. 4 Timeline with available slots for selected APREQ aircraft shown in green and unavailable slots shown in red vertically in the middle.Predicted arrivals are shown in gray.
+Fig. 6 Fig. 767Fig. 6 CLT departure banks.
+Fig. 88Fig. 8 APREQ flights by airport configuration.
+Fig. 99Fig. 9 ZDC release-request when automatic or semi-automatic negotiation mode was available.
+Fig. 1111Fig. 11 Surface states of non-prescheduled APREQ flights at initial release-time approval.
+Fig. 1313Fig. 13 APREQ assigned delay computed using Eq.(1) by negotiation center.
+Fig. 1414Fig. 14 APREQ assigned delay by center per restriction category.
+Fig. 1616Fig. 16 Rolling window of average compliance with dates when cumulative numbers were reached.
+Fig. 1919Fig. 19 APREQ compliance by center per runway.Fig. 20 APREQ compliance by center per assigned APREQ delay in five-minute bins.
+Table 1 CLT Departures Per Day.1Mean (Std. Dev.)755.5 (52.3)Min.593Median764Max.865
+Table 2 Rescheduled release time difference (s).2Mean (Std. Dev.) 306.1 (645.0)Minimum-14221 st Quartile-120Median3003 rd Quartile698Maximum2129
+Table 3 Surface-state combinations at initial and final release-time approval.3Surface State on Initial Release-Time Approval (Non-prescheduled APREQ flights)Final Surface State(RescheduledAPREQ Flights)SCHEDULED PUSHBACK RAMP_TAXI_OUT TAXI_OUTIN_QUEUESCHEDULED731 (11.8%)11 (0.2%)4 (0.1%)3 (0.0%)-PUSHBACK367 (5.9%)259 (4.2%)-1 (0.0%)-RAMP_TAXI_OUT826 (13.3%)243 (3.9%)129 (2.1%)--TAXI_OUT1517 (24.4%)485 (7.8%)129 (2.1%)129 (2.1%)1 (0.0%)IN_QUEUE868 (14.0%)345 (5.6%)88 (1.4%)57 (0.9%)20 (0.3%)
+		
+		
+
+			
+AcknowledgmentsThis research was supported by the NASA ATD-2 project, Al Capps, Project Lead.Thanks go to the many dedicated researchers and practitioners who have supported the ATD-2 field demonstration.
+			
+
+			
+
+				
+
+
+	
+		Comparing European ATM master plan and the NextGen implementation plan
+		
+			DavidBatchelor
+		
+		10.1109/icnsurv.2015.7121357
+		
+	
+	
+		2015 Integrated Communication, Navigation and Surveillance Conference (ICNS)
+		
+			IEEE
+			10 April, 2019
+		
+	
+	Federal Aviation Administration, "NextGen Priorities Joint Implementation Plan," URL: https://www.faa.gov/nextgen/media/NG_Priorities_Joint_Implementation_Plan.pdf [retrieved 10 April, 2019].
+
+
+
+	
+		Airspace Technology Demonstration 2 (ATD-2) Phase 1 Concept of Use (ConUse)
+		
+			YJung
+		
+		NASA TM-2018- 29770
+		
+			2018
+		
+	
+	Jung, Y., et al., "Airspace Technology Demonstration 2 (ATD-2) Phase 1 Concept of Use (ConUse)," NASA TM-2018- 29770, 2018.
+
+
+
+	
+		Operational Impact of the Baseline Integrated Arrival, Departure, and Surface System Field Demonstration
+		
+			ShivanjliSharma
+		
+		
+			AlCapps
+		
+		
+			ShawnEngelland
+		
+		
+			YoonJung
+		
+		10.1109/dasc.2018.8569828
+	
+	
+		2018 IEEE/AIAA 37th Digital Avionics Systems Conference (DASC)
+		London
+		
+			IEEE
+			2018
+		
+	
+	Sharma, S., Capps, A., Engelland, S., and Jung, Y., "Operational Impact of the Baseline Integrated Arrival, Departure, and Surface System Field Demonstration," 37th IEEE Digital Avionics Systems Conference, IEEE, London, 2018.
+
+
+
+	
+		Impact of Departure Prediction Uncertainty on Tactical Departure Scheduling System Performance
+		
+			AlanCapps
+		
+		
+			EdwardWalenciak
+		
+		
+			ShawnEngelland
+		
+		10.2514/6.2012-5674
+	
+	
+		12th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference and 14th AIAA/ISSMO Multidisciplinary Analysis and Optimization Conference
+		Indianapolis
+		
+			American Institute of Aeronautics and Astronautics
+			2012
+		
+	
+	Capps, A., Walenciak, E., and Engelland, S., "Impact of Departure Prediction Uncertainty on Tactical Departure Scheduling System Performance," 12th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference and 14th AIAA/ISSMO Multidisciplinary Analysis and Optimization Conference, Indianapolis, 2012.
+
+
+
+	
+		
+			SEngelland
+		
+		
+			ACapps
+		
+		
+			KDay
+		
+		
+			MKistler
+		
+		
+			FGaither
+		
+		
+			GJuro
+		
+		NASA TM-2013-216533
+		Precision Departure Release Capability (PDRC) Final Report
+		
+			2013
+		
+	
+	Engelland, S., Capps, A., Day, K., Kistler, M., Gaither, F., and Juro, G., "Precision Departure Release Capability (PDRC) Final Report," NASA TM-2013-216533, 2013.
+
+
+
+	
+		Scheduling and Delivering Aircraft to Departure Fixes in the NY Metroplex with Controller-Managed Spacing Tools
+		
+			EricChevalley
+		
+		
+			BonnyParke
+		
+		
+			JoshKraut
+		
+		
+			NancyBienert
+		
+		
+			FaisalOmar
+		
+		
+			EverettPalmer
+		
+		10.2514/6.2015-2428
+	
+	
+		15th AIAA Aviation Technology, Integration, and Operations Conference
+		Dallas
+		
+			American Institute of Aeronautics and Astronautics
+			2015
+		
+	
+	Chevalley, E., Parke, B., Kraut, J., Bienert, N., and Omar, F., "Scheduling and Delivering Aircraft to Departure Fixes in the NY Metroplex with Controller-Managed Spacing Tools," 15th AIAA Aviation Technology, Integration, and Operations Conference, Dallas, 2015.
+
+
+
+	
+		Evolution of Electronic Approval Request Procedures at Charlotte Douglas International Airport
+		
+			LindsayStevens
+		
+		
+			ToddJCallantine
+		
+		
+			RobertStaudenmeier
+		
+		10.1109/dasc.2018.8569318
+	
+	
+		2018 IEEE/AIAA 37th Digital Avionics Systems Conference (DASC)
+		London
+		
+			IEEE
+			2018
+		
+	
+	Stevens, L., Callantine, T., and Staudenmeier, R., "Evolution of Electronic Approval Request Procedures at Charlotte Douglas International Airport," 37th IEEE Digital Avionics Systems Conference, London, 2018.
+
+
+
+	
+		Aggregate Statistics of National Traffic Management Initiatives
+		
+			JosephRios
+		
+		10.2514/6.2010-9382
+	
+	
+		10th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference
+		Fort Worth
+		
+			American Institute of Aeronautics and Astronautics
+			2010
+		
+	
+	Rios, J., "Aggregate Statistics of Traffic Management Initiatives," 10th AIAA Aviation Technology, Integration, and Operations Conference, Fort Worth, 2010.
+
+
+				
+			
+		
+	

file113.txt

@@ -0,0 +1,347 @@
+
+	
+	
+		
+A. Pushback Start TimeThe purpose of the pushback start measure was to assess the uncertainty of the predicted pushback time.The SDSS component of the PDRC prototype system begins computing surface trajectories and takeoff time predictions as soon as a flight plan is received for a given flight.These surface trajectories begin with at the departure gate position and with an estimated pushback (i.e.OUT) time.Multiple sources of pushback time estimates were available to the PDRC system during the evaluation.These include the filed flight plan time, pushback times received directly from an interface with a major air carrier, and a secondary source of pushback data from a commercial flight data service.The primary focus of this research was on the accuracy of pushback received directly from the air carrier.
+Measurement ApproachThe pushback prediction used in this analysis was recorded immediately prior to the actual pushback event.Truth data used to assess the predictive accuracy of the pushback event was the actual out event data received from a gate docking system at DFW airport.This system uses video surveillance of the gate to detect when the flight has moved one meter, at which point it sends notification of the event.The actual OUT data were supplied by the airline system.Estimated OUT times from the air carrier were compared with actual OUT times from the carrier for all flights from the May 30, 2012 through June 22, 2012.The data used were from air carrier predictions that immediately preceded the actual OUT occurrence, generally within 10 seconds prior to actual OUT event.All flights which did not have an actual OUT or predicted OUT from the airline were removed from the sample.Given the objective was to evaluate the general uncertainty associated with this event and observers were not available to capture details on each OUT event occurrence, the outliers were removed from the sample by removing 1.5 times the interquartile range (IQR).The IQR method of filtering was used due to asymmetric distributions for some departure events in this research.The remaining set of data contained 30,792 departing flights from DFW with air carrier data.
+Uncertainty ResultsThe results of the pushback time measure are illustrated in Fig. 2. The mean for this sample was -52 seconds, indicating that flights pushback earlier than their last predicted OUT time from the carrier.After eliminating outliers, the median for this sample was 0 seconds and the standard deviation was 148 seconds.Fifty-three (53) percent of flights in this sample push back earlier or exactly on time with their estimated pushback time.Eighty-six (86) percent of flights push back within one minute late, and ninety-two (92) percent of flights push back prior to 2 minutes late.For tactical departure scheduling purposes, flights that arrive at the spot early pose no challenge for scheduling into the overhead stream.While this is an encouraging statistic, 2,355 flights in this sample pushed back greater than two minutes later than their predicted time.This is significant because if any of these flights were tactically scheduled prior to pushback, it is unlikely the flights could be expedited to depart within the current-day tactical window of two minutes early through one minute late (-2/+1).A notable observation is that the data received from the airline to report actual OUT time is currently at the resolution of one minute.For error measurement purposes, the actual OUT was compared to the predicted OUT estimate without rounding or truncations of the data.A later version of the airline data feed will contain the actual OUT in seconds level precision.
+B. Pushback DurationThe purpose of the pushback duration measure was to assess the uncertainty associated with the range of the pushback event duration.Currently, PDRC's SDSS component accounts for pushback duration with an adaptable "pushback buffer" value.This value may be tailored to account for gate location and ramp geometry.The pushback duration results are utilized in later sections of this paper to determine the pushback duration prediction error.The pushback duration prediction error is the difference in actual pushback duration from the mean pushback duration.
+Measurement ApproachOften the pushback is thought of as an instantaneous event in time, when in reality it has a duration which is dependent upon a number of factors like jet blast policy for an airport and gate geometry.The pushback event was defined to be the number of seconds from the OUT event discussed in the previous section until the flight begins forward motion under its own power.Observations indicate that nearly all flights at DFW employ the use of a tow tractor, known as a 'tug'.Truth data used to measure pushback duration are from manual observations made by test personnel.Manual observations were collected for 194 flights departing various gates during the months of May through July 2012.The observer of the pushback event had either direct line of site visibility to the flight or live video camera with pan/tilt/zoom control capability.The start and end of the pushback event were recorded at second's level precision.For 138 of the events, the time at which the tug was disconnected was also captured.
+Uncertainty ResultsThe results of the pushback duration measure are depicted in Fig. 3.The sample taken had a mean value of 202 seconds, with a median of 189 seconds.The sample demonstrated a significant amount of variation with a standard deviation of 73 seconds, a minimum pushback time of 77 seconds and a maximum pushback time of greater than 7 minutes.The pushback duration data best fit a lognormal distribution.This lognormal distribution is not surprising given this measurement is of time durations which are all positive and independent of one another.Gate-specific pushback duration variability can occur due to limited space for the tug to push the flight into regions of the ramp taxi area, or areas in which the jet blast from engine start may not be allowed.Due to this geometry, the tug may be required to push the flight back then subsequently pull the flight to a different location prior to disconnecting and aircraft engine start.Significant pushback duration variance existed by air carrier as well with the highest carrier average pushback of 246 seconds and the lowest 148 seconds.
+C. Ramp Taxi DurationThe purpose of the ramp taxi duration measure was to assess the uncertainty associated with the range of ramp taxi time duration.Currently PDRC's SDSS component models ramp movement as a constant-speed taxi from the gate location to the spot.The ramp taxi duration results are utilized in later sections of this paper to determine the ramp taxi prediction error.The ramp taxi prediction error is the difference in actual ramp taxi time versus the predicted ramp taxi time.
+Measurement ApproachThe ramp taxi event was defined as the number of seconds from forward motion in the ramp area until the aircraft reaches the spot.The primary source of truth data used to assess ramp taxi uncertainty was from manual observations by test personnel with visual access to the pushback event mentioned in the previous section through the flight's arrival at the spot.Manual observations were collected for 189 flights at various DFW gates during the months of May through July 2012.The start and end of the pushback event were recorded at seconds level precision.
+Uncertainty ResultsThe results of the ramp taxi measure are depicted in Fig. 4. The mean ramp taxi time for the sample was 85 seconds with a median of 81 seconds.Some variation was noted with a standard deviation of 40 seconds, and a minimum of 5 seconds with a maximum of over 5 minutes.The distance from the ramp taxi start location to the spot was an important consideration.The average ramp taxi speed was computed for each flight by dividing the ramp taxi duration by the ramp taxi distance.The mean and median ramp taxi speed for all flights was 8 knots.A significant amount of ramp taxi speed variation existed within the sample with a standard deviation of 3 kts, a high of 20 kts and a low of 3 kts.The variation was more evident amongst aircraft type than by air carrier, with the lowest average ramp taxi speed of 6.5 kts by the Boeing 737 series and the highest average ramp taxi speed of 8 kts by the McDonnell Douglas MD-80 series.
+D. Spot Crossing DurationThe purpose of the spot crossing duration measure was to measure uncertainty associated with the range of times aircraft were held at the spot prior to entering the airport movement area (AMA).Currently PDRC's SDSS component has the ability to add delay for spot crossing to deconflict with other flights, however, this was not used in the PDRC operational evaluation.Therefore, any time spent waiting at the spot is considered to be prediction error.
+Measurement ApproachThe spot crossing duration was defined as the number of seconds that the flight waited at the airport spot prior to entering the AMA.Manual observations were collected for 190 flights during the months of May through July 2012.
+Uncertainty ResultsThe results of the spot crossing duration times are depicted in Fig. 5. Approximately 81% of flights at DFW did not stop at the spot prior to entering the AMA.This can be seen in Fig. 5 by the large number of aircraft which had 0 to 10 seconds delay.When flights do stop at the spot, there is generally a small time expense to this action and the average wait is 29 seconds.Significant variation exists with a minimum of 0 seconds and max of 100 seconds for this sample.Many of the longer wait times at the spot can be explained by already present transiting flight on the AMA taxiway.Another situation that may explain the non-zero wait time is ground controller delay in contacting the flight and issuing a clearance to enter the AMA.
+E. Airport Movement Area Taxi DurationThe purpose of the AMA taxi duration measure was to assess the uncertainty of the taxi time duration in the FAA-controlled airport movement area.Currently PDRC's SDSS component models AMA movement as a constant-speed taxi from the spot to the runway departure queue following a node-link surface trajectory.AMA taxi prediction error is any error generated by the PDRC system using the current day algorithms and AMA taxi decision trees available for this prediction.
+Measurement ApproachThe AMA taxi time duration is defined as the amount of time from entering the AMA to the point at which the flight enters the departure queue.The AMA taxi time measure intentionally excludes uncertainty associated with the departure queue itself.Truth data used for AMA taxi time was obtained from post-analysis routines of PDRC output.This logic determined the entry into the AMA as well as the time the flight entered the departure queue.The AMA entry and departure queue time were compared to create the actual ramp taxi time for each flight.The PDRC AMA taxi time prediction was obtained on each flight by assuming a constant AMA taxi speed of 17 knots over the distance between the spot and departure queue which is used in SDSS.For each AMA taxi time prediction, the predicted taxi time to the departure queue was compared with the actual.Outliers were eliminated from the data sample using 1.5 times the IQR.The remaining sample of 46,325 flights from June and July 2012 are discussed in the next section.
+Uncertainty ResultsThe results of the airport movement taxi time statistics are depicted in Fig. 6.The overall mean AMA taxi time prediction error is 25 seconds, while the median error is 23 seconds.A positive error indicates the tendency of the PDRC system is to under predict the amount of time it takes for the flight to taxi from the spot to the runway threshold.Under predicting the AMA taxi time is undesirable in tactical departure scheduling because it may allow insufficient time for the local controller to stage the flight in the departure queue and control the flight to meet its coordinated departure time.The absolute AMA taxi error from the PDRC system was 35 seconds with a median absolute error of 29 seconds.The average absolute error is approximately 16% of the size of the flight's total AMA taxi time duration.Variance also existed in the AMA taxi prediction error with a standard deviation of 38 seconds, a minimum of 76 seconds early and a maximum taxi error of 131 seconds late.The variance is most prominent when viewing the data by air carrier and aircraft type.The top carrier and aircraft type combination have an average of 54 seconds taxi average error, while the lowest have an -13 seconds average taxi error.This suggests that the prediction error could be reduced by using different taxi speeds based upon air carrier and aircraft type.
+F. Takeoff Clearance Reaction Time UncertaintyThe purpose of this measure was to assess the uncertainty associated with the range of times of pilot throttle-up response to ATC takeoff clearance.Currently PDRC's SDSS component does not explicitly model this portion of the departure.The takeoff clearance reaction time results are utilized in later sections of this paper to determine the prediction error that would exist assuming the mean clearance reaction time for all flights.
+Measurement ApproachTakeoff clearance reaction time was defined as the time from ATC issuance of the 'Cleared for Takeoff' directive to the point the flight begins its takeoff roll.This time, as well as the takeoff roll duration, was not part of the surface system's prediction.Thus, an estimate was required prior to communicating the predicted wheels OFF time to the downstream decision support system in PDRC.The clearance reaction time measure required undelayed ATC voice clearance instructions as well as direct line of site to the departing flight to determine when it began its takeoff roll.A total of 108 flights were observed from the DFW Center tower during May-July 2012.
+Uncertainty ResultsThe results of the takeoff clearance reaction time measure are depicted in Fig. 7.An interesting observation was that approximately 35% of flights did not stop on the runway threshold but rather continued directly into their takeoff roll.For these flights, the departure clearance was given before the runway hold line or immediately upon entering the runway.These flights are captured in the first bin of the histogram in Fig. 7.The mean and median for the sample taken were both 6 seconds.The data sample also demonstrated some variance with a standard deviation of 5 seconds, a minimum of 0 and a maximum of 25 seconds.Using the mean and median times for the entire sample do not give the best indication of how long it took for the time it took the pilot to react to the departure clearance given this includes flights that did not stop at all.For those flights that did stop on the runway threshold, the mean and median clearance reaction time was 9 seconds.
+G. Takeoff roll durationThe purpose of the takeoff roll measure was to assess uncertainty that exists in the range of times from a flight's start of roll to the time at which the rear wheels were OFF the airport surface.An estimate of this measure was required given the downstream decision support system's ascent model begins at wheels OFF time and location.PDRC's SDSS component uses an adaptable value for takeoff roll duration.Currently, this value is the same for all aircraft types.The takeoff roll prediction error discussed in this paper is the difference in actual takeoff roll versus the mean takeoff roll.
+Measurement ApproachDirect observation of takeoff duration was selected over surface data analysis techniques to ensure the accuracy of the start of roll and rear wheel liftoff.This measure required direct line of site to the departing flight in order to determine when the flight began its takeoff roll and wheels were off of the airport surface.One hundred ninety one (191) flights were observed from DFW center tower during 2011 and 2012 PDRC evaluations.
+Uncertainty ResultsThe results of the takeoff roll duration measure are depicted in Fig. 8.Both the mean and median takeoff time duration for the sample taken was 38 seconds.Some variation was noted with a standard deviation of 7 seconds, a minimum of 18 seconds and a maximum of 55 seconds.Analytical models for calculating the distance for takeoff roll exist in the literature. 5These models generally require knowledge of takeoff weight, prevailing winds on the surface or other variables that are not currently available to the PDRC system in real time.Given this data access limitation, as well as a focus on takeoff duration rather than takeoff roll length, sampling of actual takeoff roll duration was utilized to formulate an average takeoff time duration.However, assuming a significant sample size it may be possible to narrow the variation by use of empirical data.In the sample of data collected for this research, the average takeoff duration for a McDonnell Douglas MD-80 aircraft was 41 second, compared to an average takeoff time for an Embraer ERJ 145 of 32 seconds.
+H. TRACON transit timeThe purpose of the TRACON transit time measure was to assess uncertainty that exists in the transit between the wheels off event and crossing the departure fix on the boundary of TRACON and Center airspace.The following paragraphs describe improvements to TRACON transit time predictions that were incorporated into PDRC's TMA/EDC component and used during the operational evaluation.Currently, departure logic in PDRC's TMA/EDC component predicts that the flight will fly an adaptable number of nautical miles in the direction of departure and then acquire the first departure fix in the departure route.Analysis of the DFW departure data revealed that the first fix was significantly downstream in the aircraft's route of flight.Due to this, the en route transit time predicted by TMA/EDC assumed that the flight would head directly toward this fix instead of capturing the nominal waypoints along the RNAV departure route.Figure 9a illustrates the horizontal profile of the TMA/EDC predictions which are representative of the current operational system logic.In this diagram, DFW is the green dot and DARTZ (red X) is the first fix in TMA's estimated route.The thick, dull gray line is the actual track.The various colored lines which extend from the gray route are the TMA/EDC provided estimated routes at that point in time.In an ideal scenario, these lines would overlay the thick gray route.In order to provide a more accurate route for PDRC evaluation, several potential solutions were analyzed.The solution selected was to create more specific departure routing which includes the expected TRACON departure fixes from the RNAV departure route.The routing assignment in adaptation was linked to the departure runway, which is automatically passed to TMA/EDC from PDRC's SDSS component.Figure 9b provides a graphical view of PDRC predictions of a DFW departure after implementing this solution.As illustrated the predictions and the actual tracks align closely.The previous figures plot the route geometry before and after but do not reflect the impact to the estimate times-ofarrival to the departure fix.To determine this, the predicted and actual transit times were compared for a sample of 109 flights from DFW to IAH. Figure 10 plots the difference between the actual time of flight and the TMA predicted time of flight to the departure point using the routing change illustrated in Fig 9b .The actual transit time was defined as the time between the first radar track and the time when the flight crosses the departure fix.The data in this figure are stratified by departure runway.The points shown in blue diamonds were measures of TMA Estimated Time-of-Arrival (ETA) error to the meter fix before the routing solution was implemented.The same measure was taken for a sample of 53 flights after the routing solution previously mentioned was added to the system.The results are illustrated as green triangles in Fig. 10.North flow departures time prediction error to Runway 35 prior to the routing solution showed a mean error of 176 seconds, while after the solution this was reduced to a mean error to 28 seconds.Runway 36 demonstrated similar improvement with a mean error of 235 seconds prior to the routing solution that was reduced to 62 seconds after the solution.
+Measurement ApproachThis measurement considered the transit between the wheels off event and crossing the departure fix on the boundary of TRACON and Center airspace.The truth data used to assess this uncertainty was the actual OFF time as derived from the surface system and actual departure fix crossing time derived from airborne surveillance from TRACON and Center data sources.TRACON transit time prediction error is any error associated with the PDRC system using the current day algorithms used in the field evaluation.
+Uncertainty ResultsThe improvements mentioned in the previous section were incorporated into PDRC and used during the operational evaluation.Ninety two (92) flights from the PDRC evaluation were analyzed.Figure 11 provides a breakout of the size and frequency of TRACON transit time error.The transit time error is measured in absolute values to prevent aircraft that had negative flight time error (transit time lower than predicted) from biasing the results.For the PDRC scheduled flights during the operational evaluation, the mean absolute error was 25 seconds with a median TRACON transit time error of 21 seconds.Some variation was observed with a standard deviation of 20 seconds, a low error of 1 seconds and high of 122 seconds.The flights with the highest TRACON transit time error were scrutinized.Amongst these flights was an aircraft scheduled to depart on Runway 35L.This runway prediction was supplied by the surface system based upon statically adapted rules.Later, the flight changed to departure from Runway 36R.The runway the flight was scheduled with added approximately 40 seconds of additional transit time.This example highlights the need for obtaining the correct runway assignment prior to the en route scheduling process.
+I. Center Transit TimeThe purpose of the Center transit time measurement was to assess the uncertainty in transit between departure fix crossing and meter point crossing events.
+Measurement ApproachThe departure fix is located on the boundary between TRACON and Center airspace, while the departure meter points are generally located on the neighboring Center boundary.Truth data used to assess this uncertainty were derived departure fix crossing time and meter point crossing times from airborne surveillance.Those flights which had a change to the meter point assignment after scheduling were removed from the sample to eliminate flights which changed intent between the time the flight was scheduled and crossing of the meter point.Center transit time prediction error is any error associated with the PDRC system using the current day algorithms used in the field evaluation.
+Uncertainty ResultsThe results of the Center transit time uncertainty measure are illustrated in Fig. 12.The mean Center transit time error for all PDRC scheduled flights was 49 seconds with a median error of 32 seconds.Note that the mean error is approximately twice as high as the TRACON transit time error despite the fact that the flight distance for these two measures are approximately the same.A significant factor is this error is that the PDRC evaluation used the EDC component of the TMA decision support tool which did not present times to the sector controllers' scopes.Thus, sector controllers made sequencing choices independent of the PDRC guidance which introduces individual sequencing preferences into the uncertainty.Manual observation of PDRC scheduled flights and discussions with Center Traffic Managers also revealed other factors, including significant speed fluctuations in the overhead stream, flights that cut corners off of the nominal route, pop-up flights that were scheduled after the PDRC scheduled flights, and altitude error.In the case of altitude error, the primary challenge was that the TMA/EDC system has no knowledge of the Letter of Agreement between Fort Worth Center and Houston Center in which aircraft are provided to Houston at flight level 290 if Houston is in East flow and flight level 310 if Houston is in West flow.Without the crossing altitude information TMA/EDC is left to speculate that the flight will cross at their filed flight plan altitudes which could have significant differences in wind speed and/or could take some time to maneuver to.
+III. System PerformanceTo determine system performance, PDRC research utilized a combination of quantitative metrics and qualitative feedback from operational personnel and subject matter experts.The focus of this section is on the objective metrics that were used to assess PDRC system performance.
+A. OFF Time ComplianceOne objective of PDRC is to improve upon schedule compliance by reducing uncertainty that has been demonstrated in manual coordination. 3Thus, OFF time compliance is an important system metric for PDRC.
+Measurement ApproachThe approach used in this measurement was to leverage the highest precision measurement source available to evaluate compliance.For the baseline sample, a full year of OFF time compliance data were available from operational TMA/EDC recordings covering more than 400 scheduled flights from October 2010 until November 2011.Flights with strategic times (EDCTs) were removed from the compliance analysis presented here but utilized in other analyses.EDCTs were not counted in the primary compliance measure because they introduced variation due to procedural differences which were not the focus of this research.Briefly, the research team observed individual controllers following different procedures in situations where flights were subject to both a strategic and tactical TMI.The OFF time agreed upon between Center and Tower traffic managers was defined to be the coordinated OFF time, which was compared against the departure time as obtained from the departure message from en route automation.Given that this large sample of data covered a long duration in which unknown circumstances might have been involved without a PDRC observer to report them, the outliers outside of 1.5 times the IQR were removed.Measuring PDRC OFF time compliance was more straightforward than the baseline given firsthand knowledge of every scheduled flight.For example, one PDRC scheduled flight that was subject to an APREQ procedure was later expedited in order to prevent potential hail damage.At the point that verbal direction was given to expedite the flight, the APREQ time was no longer valid.However, no electronic commands were issued for this flight and had the team not been aware of this occurrence then the flight would otherwise have looked non-compliant.Flights which had both a strategic and tactical TMI were captured for analysis in PDRC, but like the baseline set they did not count toward OFF time compliance results.The measure of OFF time compliance for a PDRC flight is the coordinated OFF time versus the actual wheels off time as available in the PDRC.PDRC calculates the actual OFF by utilizing a detected start of takeoff roll and adapted roll duration.A total of 120 flights were scheduled by the PDRC system during the operational evaluation from May 30, 2012 through July 26, 2012.For a flight to count as a PDRC scheduled flight, both the surface and the Center traffic managers had to schedule the flight using the PDRC system and agreed upon scheduling procedure.
+ResultsThe distribution of PDRC OFF time compliance is illustrated in Fig. 13.The mean compliance was 33 seconds with a median of 37 seconds, indicating a slightly later actual OFF time than planned time on average.A fair amount of variance was exhibited in this sample as well, with a standard deviation of 63 seconds, a minimum of 135 seconds early and a maximum of 165 seconds late.The two flights with the highest OFF time error were both due to the flight not having its weight and balance numbers when it arrived at the runway threshold.Table 1 provides a comparison between the baseline OFF time compliance and the PDRC system OFF time compliance.The second column of this table has the PDRC OFF time compliance mean, median and standard deviation values.The baseline OFF time compliance is listed in the third column, and the last column indicates the estimated percentage of PDRC OFF time compliance compared to the baseline compliance.This improvement is characterized as a lower bound estimate due to the fact that outliers were removed from the baseline data set but not the PDRC data set.As previously discussed, outliers were removed given the PDRC team did not observe all flights in the one year sample.As the table indicates, PDRC scheduled flights demonstrated a significant improvement over baseline levels of OFF time compliance.
+B. Hit Slot Performance MeasureThe purpose of the hit slot performance measure was another way to assess how well the system delivered flights to the available airspace, or slot, they were originally scheduling into.
+ApproachThe purpose of the hit slot performance measure was to assess how well the system delivered flights to the available airspace, or slot, they were originally scheduling into.The hit slot measure compares the leading and trailing flight at the time of scheduling with those at the time the flight crosses the meter point in en route airspace.A detailed explanation of this measurement and the baseline hit slot results can be found in prior PDRC research. 1
+ResultsForty-one percent (41%) of PDRC scheduled flights hit the slot they were scheduled into.This represents only a modest improvement over the baseline level of hit slot performance of 39%. 1 The merge of PDRC scheduled flights into the overhead stream was analyzed to determine the primary reason for low hit slot performance.The most common cause for missing the scheduled slot was a change in the overhead stream from the time the flight was scheduled to the time the flight arrived at the meter point.Departure uncertainty from nearby airports as well as airborne flights that short cut their route contributed to the overhead stream uncertainty.
+IV. DiscussionThis section discusses inferences of system performance from individual departure predictive accuracy measures.
+A. Impact of Communication Uncertainty and APREQ Window Size on OFF Time complianceTable 1 of this paper indicated PDRC exhibited a significant improvement over baseline absolute OFF time compliance.The primary reasons for improved OFF time compliance using PDRC is inferred to be reduced communication uncertainty between the Tower and Center traffic managers and use of a PDRC-enabled target time rather than the standard APREQ window.This section provides a summary of information gained during the PDRC evaluation from direct observations and traffic management controller (TMC) interviews.Without the use of PDRC, the coordinated OFF time window for an APREQ is communicated verbally over facility inter-phone from the Center to the Tower.Currently, there is no set national standard for the time window to use in this communication or the phraseology to employ in this procedure.However, this time window is generally accepted as being 3 minutes and is structured to favor flights departing early rather than late.The general rule is to build a time window two minutes ahead (-2) and one minute behind (+1) the flight's desired OFF time.Observations of this verbal exchange and interviews of Center personnel indicate a significant degree of uncertainty in this communication.For example, a time of 17:25:26 Zulu from the TMA/EDC system is truncated and displayed in minutes level granularity to Center personnel as 1725.The Center TMC then verbally communicates the APREQ window which may be communicating a single value of 1725 or any number of variations in APREQ window size, bias and phraseology.After receiving the time window verbally from the Center TMU, the Tower TMC must interpret the information.In some cases, the TMC receiving the information may assume the beginning of the first minute to the end of the last minute given.In the case of the time communicated as 1724 to 1727, this interpretation would be 1724:00 to 1727:59, which would be a 3 minute and 59 second window.Other Tower TMCs indicate that they take the specified window to mean the beginning of the first minute to the beginning of the last minute given.Interviews with Tower personal revealed that another source of communication uncertainty is the location at which the flight is expected to be at this time window.In some cases, the Tower TMC assumed the location that the flight was expected at the negotiated time was when the flight was "tagged up", or the point at which TRACON surveillance was first received for the flight.However, the trajectory from the TMA/EDC system begins at wheels OFF from the airport surface.Observations of "tag up" as compared to actual OFF time indicated this duration has a 26 second mean with 25 second median.To reduce the complexity of this communication, the PDRC system automatically sent the time expected by en route automation at seconds level precision to the surface system.The Tower TMC then used the seconds level precision to communicate a minute's level precision value to the surface local controller.The local controller was asked to try to achieve wheels OFF at the exact minute communicated to the best of their ability rather than using the standard APREQ compliance window described above.
+B. Combined View of Departure ErrorEarlier sections of this document described departure event measurements of uncertainty for departure events analyzed in this research.However, a reasonable question may be, how well does the system perform in the presence of this uncertainty?To answer this question, it was necessary to analyze the level of predictive error associated with each departure event.In most cases, the PDRC prediction for the event was utilized to obtain this measure.In some cases, like in the hypothetical case of scheduling flights from the gate which was not performed during the PDRC evaluation, it was necessary to estimate the level of error using a reasonable approach at the prediction like those discussed in earlier sections of this paper.Figure 14 combines the departure prediction measures described in this research into a single diagram.This figure provides a view of the size and distribution of the current prediction error for each departure event in the presence of current day uncertainty.In the majority of cases, the PDRC prediction for the event is utilized to obtain this measure.In some cases, like in the hypothetical case of scheduling flights from the gate which was not performed during the PDRC evaluation, it was necessary to estimate the level of error using a reasonable approach at the prediction.The mean values of each measure are shown in bold near the red cross.The median value, upper quartile and lower quartile define the boundaries of the box structure in the box plot, and the 'whiskers' extend on both directions of the box to encompass the variance of the distribution without inclusion of the outliers.In the case of pushback start and center transit error, a portion of the box plot whisker cannot be seen because they go beyond the scale of the diagram.The variance in the pushback start error, pushback duration error, AMA taxi error and Center transit error are the largest among all departure events.Of those events, only the AMA taxi and the Center transit error are part of current day tactical departure scheduling.Both the pushback duration and pushback start error would need to be considered when performing tactical scheduling prior to the spot.Even assuming perfect OUT time compliance, the pushback duration event alone experiences variances that are high enough to prevent a flight from being able to make the current day tactical departure window which only allows flights to be one minute late.Additionally, while mean ramp taxi error is low based upon the average ramp taxi speed method utilized, the upper tail stretches nearly 70 seconds.If significant ramp taxi error occurred then this would leave little room for error for any remaining event predictions.However, the error measured in this analysis is absent air carrier efforts to meet a specified time.It is possible that with greater air carrier involvement flights could be expedite to meet an earlier time if required.Spot crossing duration has a positive overall mean of 6 seconds.While this value is low, the maximum spot crossing error was 100 seconds and spot crossing duration is an event that can add to overall system uncertainty in current day tactical departure scheduling.The AMA taxi represents a significant source of uncertainty in today's tactical departure scheduling process.The taxi time error has a number of error components, the most sizable of which are runway prediction error from the spot and predicted taxi speed from the spot.The size of the error of the clearance and the takeoff roll time are negligible in comparison with the other error sources once taken into account.If they were not taken into account however, then on average each flight would have 6 seconds of clearance error plus 38 seconds of roll time error, for a total of 44 seconds error on average.While this seems small, that equates to approximately 5nmi in the overhead stream (assuming 420 kts).TRACON transit time error has been reduced significantly with the use of more detailed TRACON routing in TMA/EDC as well as automatic utilization of the runway assignments passed from the surface system.However, the overall average of the error is positive.That is, the current system is consistently under predicting the transit time from OFF to the departure fix.For tactical scheduling purposes, it is better to over predict rather than under predict TRACON transit time if forced to choose between the two.Over predicting the transit time would allow the flight to be delayed to meet the time rather than accelerated.Center transit time has the largest positive error of any departure event measured.Bearing in mind that all the flights measured were during a volatile period in which an APREQ event was put in place to help manage, a certain degree of variance is not surprising.Additionally, this operational evaluation involved only outbound tactical departure scheduling 3 using the TMA/EDC decision support tool.Currently, the FAA operates TMA/EDC in an open-loop mode.Unlike arrival metering with TMA, TMA/EDC schedule times and sequence information are not displayed on sector controllers radar scopes.Center TMCs use TMA/EDC to manage constrained traffic flows to provide sector controllers with a workable traffic situation.Sector controllers solve the traffic puzzle with no knowledge of the TMA/EDC planned solution.Thus, differences between the TMA/EDC and controller solutions are to be expected.Given the timing associated with widespread deployment of a surface capability that could supply the OFF times required, it is likely better to assume a metering environment in which times are presented to the controllers with +/-30 seconds of error like those demonstrated by the Efficient Descent Advisor (EDA). 6
+C. Extending tactical departure scheduling to the gateThe cumulative departure scheduling error that may occur when scheduling from the gate was estimated by taking 1000 random draws, with replacement, from the error data samples described in previous sections of this paper.The error components of each departure event were accumulated to form the total surface departure error associated with each of the 1000 flights in the sample.If the total departure error was one minute or less, the flight was considered compliant with the surface OFF time.Total departure error that allowed a flight to be available earlier than their scheduled time was considered compliant for this measure given the flight could be delayed by ATC to meet the required time.The cumulative error estimate for PDRC during the operational evaluation was taken by using the takeoff roll time and clearance reaction time that were available during the evaluation, given these values were slightly different than those discussed in this research.As indicated in Table 2, scheduling from the spot using the levels of PDRC accuracy available during the PDRC operational evaluation yielded an estimated 71% departure compliance.That is, 71% of flights had one minute or less total surface error when scheduling from the spot.For comparison purposes, the actual percentage of flights in the PDRC operational evaluation with one minute or less surface error were determined to be 70%.Thus, the theoretical estimates and actual distribution for PDRC OFF compliance are very close.An estimate was taken of the improvement to OFF time compliance that could be achieved from implementing changes described in this research.Specifically, the prediction error associated with using the mean clearance reaction time, mean takeoff roll time and adding a 30 seconds buffer to remove late bias associated with current OFF time compliance.The estimated OFF time compliance when scheduling from the spot with these changes is 92%.OFF time compliance while scheduling from the gate was estimated by using the error components associated with pushback start, pushback duration and ramp taxi.Scheduling from the gate using the levels of PDRC accuracy available during the operational evaluation yielded an estimated 64% compliance.However, scheduling from the gate with the changes described in this research would yield an estimated 73% OFF time compliance.For comparison, OFF time compliance achieved nation-wide with manual scheduling process is approximately 69%. 1 It is important to note that this analysis did not consider active air carrier participation in the tactical departure scheduling process.Currently, the PDRC concept places no requirements on air carriers beyond passively providing gate assignment and pushback estimate information that already resides in air carrier systems.Other concepts such as Spot and Runway Departure Advisor (SARDA), 7,8 Collaborative Departure Queue Management (CDQM), 9 and Surface Collaborative Decision Making (Surface CDM) 10 assume an active air carrier role in departure scheduling to enable surface delays to be absorbed at the gate or in the ramp area.Combining active air carrier participation from these concepts with PDRC's integration between Tower and Center departure scheduling systems may enable the tactical departure scheduling horizon to be extended to the gate with satisfactory compliance.
+D. Removing Late Bias from Departure Fix complianceGiven only modest improvements to hit slot performance over the baseline measure despite PDRC's significant improvement to OFF time compliance, an alternative measure of system performance was developed.The purpose of the departure fix compliance performance measure was to assess how well the system delivered flights to the boundary between TRACON and Center airspace compared to the scheduled time.This measure was useful because it provided an interim point between wheels off and meter point crossing at which the performance of the system could be analyzed.In addition, the majority of airborne vectoring and speed controls occur after the departure fix which allows an objective metric to be obtained from operational data with fewer confounding influences than the hit slot measure.To determine if the departure fix compliance was a reliable performance measure, PDRC OFF time compliance and departure fix crossing errors were analyzed for correlation.The correlation coefficient for these two data sets was 0.937, which indicates a high correlation.This suggests that as OFF time uncertainty decreases, so does departure fix crossing uncertainty.In contrast, the correlation coefficient between PDRC OFF time compliance and meter fix compliance was only 0.16, indicating a low correlation.The departure fix compliance measure utilized in PDRC relied upon the coordinated departure time negotiated between the systems and TMA/EDC's estimate of time to fly to the departure fix.The two values were combined to form the scheduled departure fix crossing time.The scheduled crossing time was then compared against the actual crossing time to determine if it was compliant.The compliance standard used was the same as today's standard of two minutes early through one minute late (written as -120/+60).The results for departure fix compliance for PDRC scheduled flights are that 51% hit the -120/+60 second window.Given significant improvement to OFF time compliance demonstrated with PDRC as well as lower variance in the overall distribution, one might expect higher departure fix compliance than was demonstrated.Analysis of departure fix compliance revealed that the primary reason for lower than expected departure fix compliance is flights were slightly later than planned.This fact is not surprising given that the local controllers using the PDRC times were required to change their procedure to use a single minute time rather than the time window that they were accustomed to.To compensate for this, a slight modification can be made to the times that are communicated via PDRC.This modification can be implemented in the software communication layer between the Center system and the Tower system such that no changes to PDRC scheduling procedures would be required.Analysis of possible buffer values using PDRC scheduled flights indicated departure fix compliance as high as 86% may be possible with this simple software change.
+V. ConclusionsDuring the operational evaluation of Precision Departure Release Capability at DFW, OFF time compliance improved from an average absolute error of 108 seconds to less than 59 seconds.PDRC demonstrated greater predictability than the baseline sample by decreasing OFF time error from a standard deviation of 96 seconds to 40 seconds.Significant improvements to TRACON transit time predictions were achieved by including TRACON-specific routing in the horizontal profile and electronically supplying the airborne system with the departure runway assignment from the airport surface system.Despite these improvements, additional work is needed to reduce TRACON transit time error.Prediction errors associated with the departure events were utilized to estimate the OFF time associated with scheduling tactical departures from the gate.This estimate indicates that OFF time compliance of 73% of flights scheduled from the gate may be possible without requiring active airline involvement, which exceeds baseline tactical departure scheduling OFF time compliance.Figure 1 .1Figure 1.Tactical departure scheduling is affected by the cumulative uncertainty of numerous departure events.
+Figure 2 .2Figure 2. Comparison of airline provided actual OUT with airline provided estimated pushback times.
+Figure 3 .3Figure 3. Histogram showing pushback durations.
+Figure 4 .Figure 5 .45Figure 4. Ramp taxi time duration uncertainty.
+Figure 6 .6Figure 6.Airport movement area taxi prediction error assuming 17 knot taxi speed.
+Figure 7 .Figure 8 .78Figure 7. Scattergram showing duration from takeoff clearance to start of roll.
+Figure 9 .9Figure 9. TMA departure route predictions before and after implementation of an adaptationbased solution.
+Figure 11 .Figure 10 .1110Figure 11.Absolute Error of TRACON Transit Time
+Figure 12 .12Figure 12.Absolute Center transit time error.
+Figure 13 .13Figure 13.PDRC OFF time compliance distribution.
+Figure 14 .14Figure 14.Comparison of prediction error for the various departure events analyzed in the PDRC operational evaluation.
+Table 1 . PDRC absolute OFF time compliance compared with baseline. PDRC absolute OFF time compliance (sec)1Baseline OFF timeOFF time errorcompliance, outliersversus baselineremoved (sec)(lower bound %)
+Table 2 . Estimated OFF time compliance percentage using cumulative departure event error2Downloaded by NASA Ames Research Center on November 29, 2012 | http://arc.aiaa.org| DOI: 10.2514/6.2012-5674Scheduling from theScheduling from theScheduling from theScheduling from thespot using PDRCspot using changesgate using PDRCgate with changesaccuracy duringdescribed in thisaccuracy duringdescribed in thisoperational evaluationresearchoperational evaluationresearch71%92%64%73%
+			Downloaded by NASA Ames Research Center on November 29, 2012 | http://arc.aiaa.org| DOI: 10.2514/6.2012-5674
+		
+		
+
+			
+AcknowledgementsThe authors would like to acknowledge the essential support provided by FAA personnel at the Fort Worth Center Traffic Management Unit and Dallas/Fort Worth ATCT.Finally, we wish to thank our colleagues at NTX and NASA Ames whose support was critical to the success of PDRC research objectives.
+			
+
+			
+
+				
+
+
+	
+		Characterization of Tactical Departure Scheduling in the National Airspace System
+		
+			RichardCapps
+		
+		
+			ShawnEngelland
+		
+		10.2514/6.2011-6835
+	
+	
+		11th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference
+		Virginia Beach, VA
+		
+			American Institute of Aeronautics and Astronautics
+			Sep. 2011
+			
+		
+	
+	AIAA-2011-6835, 11th
+	Capps, A. and Engelland, S.A., "Characterization of Tactical Departure Scheduling in the National Airspace System," AIAA-2011-6835, 11th American Institute of Aeronautics and Astronautics (AIAA) Aviation Technology, Integration, and Operations (ATIO) Conference, Virginia Beach, VA, 20-22 Sep. 2011.
+
+
+
+	
+		Traffic Management Advisor Flow Programs: an Atlanta Case Study
+		
+			ShonGrabbe
+		
+		
+			BanavarSridhar
+		
+		
+			AvijitMukherjee
+		
+		
+			AlexMorando
+		
+		10.2514/6.2011-6533
+	
+	
+		AIAA Guidance, Navigation, and Control Conference
+		Portland, Oregon
+		
+			American Institute of Aeronautics and Astronautics
+			Aug. 2011
+			
+		
+	
+	Grabbe, S., "Traffic Management Advisor Flow Programs: an Atlanta Case Study", American Institute of Aeronautics and Astronautics (AIAA) Guidance, Navigation, and Control Conference, Portland, Oregon, 8-11 Aug. 2011.
+
+
+
+	
+		Trajectory-Based Takeoff Time Predictions Applied to Tactical Departure Scheduling: Concept Description, System Design, and Initial Observations
+		
+			ShawnEngelland
+		
+		
+			RichardCapps
+		
+		10.2514/6.2011-6875
+	
+	
+		11th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference
+		Virginia Beach, VA
+		
+			American Institute of Aeronautics and Astronautics
+			Sep. 2011
+			
+		
+	
+	AIAA-2011-6875, 11th
+	Engelland, S.A. and Capps, A., "Trajectory-Based Takeoff Time Predictions Applied to Tactical Departure Scheduling: Concept Description, System Design, and Initial Observations," AIAA-2011-6875, 11th American Institute of Aeronautics and Astronautics (AIAA) Aviation Technology, Integration, and Operations (ATIO) Conference, Virginia Beach, VA, 20-22 Sep. 2011.
+
+
+
+	
+		Benefit Assessment of Precision Departure Release Capability Concept
+		
+			KeePalopo
+		
+		
+			GanoChatterji
+		
+		
+			Hak-TaeLee
+		
+		10.2514/6.2011-6834
+	
+	
+		11th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference
+		Virginia Beach, VA
+		
+			American Institute of Aeronautics and Astronautics
+			Sep. 2011
+			
+		
+	
+	AIAA-2011-6834, 11th
+	Palopo, K., Chatterji, G., and Lee, H., "Benefit Assessment of the Precision Departure Release Capability Concept," AIAA- 2011-6834, 11th American Institute of Aeronautics and Astronautics (AIAA) Aviation Technology, Integration, and Operations (ATIO) Conference, Virginia Beach, VA, 20-22 Sep. 2011.
+
+
+
+	
+		Takeoff Performance Monitoring System display options
+		
+			DavidMiddleton
+		
+		
+			RaghavachariSrivatsan
+		
+		
+			LeePerson, Jr.
+		
+		10.2514/6.1992-4138
+		NASA TP- 3403
+	
+	
+		Flight Simulation Technologies Conference
+		
+			American Institute of Aeronautics and Astronautics
+			1994
+		
+	
+	Middleton, D.B., Srivatsan, R., and Person, L.H., Jr., "Flight Test of Takeoff Performance Monitoring System," NASA TP- 3403, 1994.
+
+
+
+	
+		Development and Testing of Automation for Efficient Arrivals in Constrained Airspace
+		
+			RCoppenbarger
+		
+		
+			GDyer
+		
+		
+			MHayashi
+		
+		
+			RLanier
+		
+		
+			LStell
+		
+		
+			DSweet
+		
+	
+	
+		27th International Congress of the Aeronautical Sciences (ICAS)
+		Nice, France
+		
+			Sep. 2010
+			
+		
+	
+	Coppenbarger, R., Dyer, G., Hayashi, M., Lanier, R., Stell, L., Sweet, D., "Development and Testing of Automation for Efficient Arrivals in Constrained Airspace," 27th International Congress of the Aeronautical Sciences (ICAS), Nice, France, 19- 24 Sep. 2010.
+
+
+
+	
+		A Concept and Implementation of Optimized Operations of Airport Surface Traffic
+		
+			YoonJung
+		
+		
+			TyHoang
+		
+		
+			JustinMontoya
+		
+		
+			GautamGupta
+		
+		
+			WaqarMalik
+		
+		
+			LeonardTobias
+		
+		10.2514/6.2010-9213
+	
+	
+		10th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference
+		Fort Worth, TX
+		
+			American Institute of Aeronautics and Astronautics
+			Sep. 2010
+			
+		
+	
+	Jung, Y. C., Hoang, T., Montoya, J., Gupta, G., Malik, W., and Tobias, L., "A Concept and Implementation of Optimized Operations of Airport Surface Traffic," 10th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference, Fort Worth, TX, 13-15 Sep. 2010.
+
+
+
+	
+		Tower Controllers' Assessment of the Spot and Runway Departure Advisor (SARDA) Concept
+		
+			THoang
+		
+		
+			YJung
+		
+		
+			JHolbrook
+		
+		
+			WMalik
+		
+	
+	
+		9th USA/Europe ATM R&D Seminar (ATM2011)
+		Berlin, Germany
+		
+			June 2011
+			
+		
+	
+	Hoang, T., Jung, Y., Holbrook, J., and Malik, W., "Tower Controllers' Assessment of the Spot and Runway Departure Advisor (SARDA) Concept," 9th USA/Europe ATM R&D Seminar (ATM2011), Berlin, Germany, 14-17 June 2011.
+
+
+
+	
+		Field test results of Collaborative Departure Queue Management
+		
+			ChrisBrinton
+		
+		
+			SteveLent
+		
+		
+			ChrisProvan
+		
+		10.1109/dasc.2010.5655527
+	
+	
+		29th Digital Avionics Systems Conference
+		Salt Lake City, Utah
+		
+			IEEE
+			Oct. 2010
+			
+		
+	
+	Brinton, C., Lent, S., and Provan, C., "Field Test Results of Collaborative Departure Queue Management," 29th Digital Avionics Systems Conference, Salt Lake City, Utah, 3-7 Oct. 2010.
+
+
+
+	
+		An Integrated Collaborative Decision Making and Tactical Advisory Concept for Airport Surface Operations Management
+		
+			GautamGupta
+		
+		
+			WaqarMalik
+		
+		
+			YoonJung
+		
+		10.2514/6.2012-5651
+	
+	
+		12th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference and 14th AIAA/ISSMO Multidisciplinary Analysis and Optimization Conference
+		
+			American Institute of Aeronautics and Astronautics
+			June 2012
+		
+	
+	FAA Air Traffic Organization Surface Operations Office
+	FAA Air Traffic Organization Surface Operations Office, "U.S. Airport Surface Collaborative Decision Making (CDM) Concept of Operations (ConOps) in the Near-Term -Application of Survace CDM at United States Airports," 15 June 2012
+
+
+				
+			
+		
+	

file114.txt

@@ -0,0 +1,279 @@
+
+	
+	
+		
+I. IntroductionASA"s current Integrated Arrival/Departure/Surface research portfolio includes integration of surface information with en route departure scheduling.The Precision Departure Release Capability (PDRC) activity is assessing the value of using surface trajectory-based takeoff (OFF) time predictions for departure scheduling.Companion papers 1,2 present a concept overview and results from benefits assessment studies.This paper describes the NAS shortfalls that PDRC technology seeks to address and assesses current PDRC levels of predictive accuracy against the current need.The document begins by describing a nation-wide survey of current tactical departure scheduling operations.Existing system shortfalls are then examined via a discussion of system performance along with the measurement approach and corresponding results.The shortfalls discussion is followed by a description of the current levels of OFF time prediction accuracy that can be obtained in the PDRC system today.The paper concludes with a discussion of sites most likely to benefit from PDRC technology.
+II. Current Day Tactical Departure SchedulingIn order to identify existing shortfalls which may be eliminated with reduced departure prediction uncertainty, it is necessary to have an understanding of the current day tactical departure scheduling process.This section covers the following five topics: 1) Tactical departure scheduling overview, 2) Current Inbound Tactical Departure Scheduling Capability, 3) Current Outbound Tactical Departure Scheduling Capability, and 4) Tactical versus strategic departure scheduling.
+A. Tactical Departure Scheduling OverviewTactical departure scheduling is the process used by ATC to regulate air traffic flow to eliminate local demand/capacity imbalances and satisfy local traffic management initiatives (TMIs).Tactical departure scheduling is not required during normal NAS operations as the airspace into which the flight is being released generally has sufficient capacity to accommodate the departure.However, during periods of high demand or low capacity for the airspace being scheduled into, tactical departure scheduling may be utilized.Tactical departure scheduling in the NAS today can be divided into two distinct tactical scheduling modes, which are outbound scheduling of departures from an airport within the departure Air Route Traffic Control Center (ARTCC, hereafter referred to as "Center") to a remote Center and inbound scheduling of departures into an arrival stream of a Traffic Management Advisor (TMA) metered airport.The inbound and outbound terms are generic labels for tactical departure scheduling functions provided by existing decision support tools (i.e.TMA scheduling, "internal" scheduling, "adjacent" scheduling, "coupled" scheduling, extended metering, etc.)The flight length associated with the tactical timeframe varies somewhat in the literature.The authors chose an upper bound of 90 minutes as the guideline for flight lengths subject to tactical departure scheduling.This flight length was chosen in part based upon information obtained from operational data usage of the decision support tools that support tactical departure scheduling.Figure 1 illustrates the relationship of the Dallas/Fort Worth (DFW) departure airport relative to arrival metering to Houston Intercontinental (IAH) airport.Given that DFW resides within the IAH metering freeze horizon and the limited airspace available to maneuver after departure prior to the outer meter arc, a high level of departure prediction accuracy is required.Later sections provide an estimate as to the level of predictive accuracy that is required.Call For Release (CFR) is a common tactical departure scheduling procedure which requires Air Traffic Control Tower (ATCT) personnel to call the Center Traffic Management Unit (TMU) for a scheduled departure time prior to releasing the aircraft for departure.The CFR procedure is applied to departing aircraft in order to ensure the demand placed on local airspace resources do not exceed the available capacity.In a CFR scenario it may or may not be necessary to delay the aircraft based upon the latest information available on the constrained flow at the time that an aircraft is ready to depart.The improved departure time compliance associated with the CFR procedure provides more accurate schedule predictions than are available via the aircraft"s filed flight plan departure time (also known as Predicted Departure Time or PTIME) or by use of Expect Departure Clearance Times (EDCTs).EDCT times are generated by Traffic Flow Management (TFM) as a part of the strategic departure scheduling system and are not intended for tactical use.Aircraft PTIMEs represent a starting point from which the departure planning process begins but are historically prone to OFF time uncertainty.The required departure compliance window for CFR aircraft varies somewhat by facility.Today, no nationwide guidance exists, but based upon information obtained from traffic managers, generally inter-facility agreements call for flights to depart within a three minute window.This three-minute window is generally structured to allow departure two minutes prior to, or one minute later than, the target coordinated departure time.The idea of allowing the aircraft to depart two minutes early is that it is easier to delay the aircraft to fit into the constrained flow than to accelerate the aircraft to meet its scheduled time.Figure 2 provides an illustration of nationwide departure time compliance comparison between estimation methods available to TMCs during the month of January 2011.January was selected for operational data analysis primarily due to the availability and completeness of the TMA operational data set during this time period.The values reported in Fig. 2 are the average absolute difference between the expected departure time and the actual departure time.The operational TMA data analyzed had information on aircraft PTIME, EDCT times, TMA times and actual departure times which were used for this nationwide departure time compliance analysis.An obvious difference exists in the departure time compliance between PTIME estimates, EDCT controlled times and CFR controlled times with the departure times coming from the CFR process providing the best compliance of the three.Using the CFR process during the month of January, approximately 69.2% of aircraft subject to CFRs in which TMA automation was utilized met the required -2/+1 window.In contrast, if EDCT times were required to meet a -2/+1 window the compliance would have been approximately 20.4 %.Using PTIME compliance this percentage would drop to only 4% of flights that met the -2/+1 window.
+B. Inbound Tactical Departure Scheduling CapabilityAs adjacent center metering has expanded the reach of TMA, the greatest need for departure scheduling capability has been for airports residing in another Center.Analysis of January 2011 operational data shows that 69.3% of all departure scheduling is performed from an origination Center that is different than the destination Center being scheduled into.The expanded scope of TMA usage is a factor to consider in analysis of tactical departure scheduling shortfalls, another factor is the effect that tactical departure scheduling capability has on the balance of delay that is assigned to the airborne stream versus airport surface.In December of 2005 a feature was added to TMA that allowed the TMC to determine whether or not departures should compete directly with active airborne flights.Prior to this feature, TMA always scheduled aircraft into the overhead stream in a manner that the departure had the same priority as airborne aircraft.The intent of this feature was to prevent airborne delays from reaching the point which it made it difficult for controllers to achieve the TMA meter crossing times.However,  the tradeoff associated with limiting the airborne delays is an increase in departure delays.When the TMC chooses to delay the airborne flow, the TMA system will treat the departing aircraft with equal priority as airborne aircraft and assign a delay to unfrozen aircraft in the metered airborne stream if needed.In this situation, TMA may delay both the airborne stream and assign a ground delay to the departing aircraft.Analysis of the current usage based upon data from January 2011 indicates that the large majority (92%) of flights scheduled in TMA took all of their tactical departure delay on the surface.The ability for the TMC to determine whether the aircraft tactical delay should be taken airborne, on the surface, or a combination of the two is complicated by uncertainty in the scheduling process.Analysis of tactical departures scheduled into the arrival TMA system during metering indicates that approximately 21% of all scheduled aircraft experience both a TMA assigned ground delay and TMA assigned airborne delay.To prevent aircraft that are assigned delay on the airport surface from being delayed again once they join the airborne flow, the TMC may "freeze" the aircraft into the airborne flow when scheduling in TMA.If the TMC selects this option when scheduling a tactical departure, the TMA system will freeze the aircraft"s scheduled time of arrival to the meter point thereby preventing any additional delay from being added to the aircraft once it becomes airborne.This feature allows the TMC to ensure the aircraft does not receive unplanned airborne delay; however, if the aircraft does not depart when expected and cannot achieve the time which is frozen into the arrival metering system"s schedule, then the space that was being reserved for this aircraft will go unutilized barring additional action by ATC to prevent this from occurring.Currently, 29% of departing flights that are scheduled into an arrival TMA system are scheduled frozen into the airborne flow: the remaining 71% of aircraft are allowed to adjust their position in the TMA arrival schedule upon first surveillance.An additional shortfall of the current day inbound tactical departure scheduling system occurs when the tactical departure delays become very large.This situation may require Air Traffic Control System Command Center (ATCSCC) involvement.In the large majority of cases the assigned ground and airborne delay are small (i.e. less than 5 minutes 73% of the time in TMA), however, cases do exist in which airborne and/or ground delay is in excess of one hour.In the month of January there were approximately 20 occurrences of TMA assigned ground delays in excess of one hour.The majority of the examples of large TMA assigned ground delay were to either New York Center or Atlanta Center metered airports.In many cases, flights with high TMA-assigned surface delay also received an airborne delay from the TMA system.These examples of high ground delay with airborne delay may lend insight into why into why sites like New York Center and Atlanta Center are top users of the "schedule frozen" option previously discussed.When high tactically-assigned ground delay occurs in the NAS, the ATCSCC may choose to implement an Airspace Flow Program (AFP) to regulate the flow of aircraft into the destination airport with the objective of reducing the TMA-assigned surface delays.The AFP scheduling scenario used for this purpose is unique in that it is designed to work in conjunction with the arrival TMA system; hence it is called a TMA Flow Program (TFP).The objectives of a TFP are to pre-condition the arrival stream such that TMA can utilize available space in the stream for tactical departure scheduling purposes.The boundaries of the TFP are set to be roughly contiguous with the arrival metering system"s freeze horizon and any airport with departures inside of this boundary are exempt from the program.Using a TFP the TFM suite of tools assigns a ground delay to aircraft bound for the metered airport which are located outside of the red circle shown in Fig. 3, while TMA assigns a tactical ground delay (and potentially airborne delay depending on TMC selection) for those aircraft bound to the metered airport located within the red circle.
+C. Outbound Tactical Departure Scheduling CapabilityIn addition to the TMA arrival metering system, the Enroute Departure Capability (EDC) is now part of the tactical departure scheduling decision support tools available to TMC personnel.The EDC system design re-uses a number of common components of the arrival TMA system like its adaptation data structure, route processing algorithms and trajectory generation functions.While many of the core components of TMA have been leveraged to provide EDC capability, there are notable differences between arrival TMA and the EDC system.The EDC system serves a different traffic management objective than the arrival TMA system.EDC"s focus is outbound tactical departures leaving from one of the airports within a Center which are destined to a remote Center facility.In contrast the tactical departure scheduling capability in arrival TMA system is only focused on aircraft that are scheduled into its metered airports.EDC is commonly used to assist in the application of miles in trail restrictions between facilities, especially when the airspace being scheduled into is highly constrained or has multiple miles in trail initiatives to satisfy.An additional use of EDC is to assist in regulating departures into sectors which are experiencing high demand.In contrast, arrival TMA use is primarily motivated by the traffic volume in the arrival streams entering the metered airport rather than sector loading considerations.The TMA EDC system is deployed to all 20 Centers within the NAS.Similar to the nationwide deployment of the arrival TMA system, there is significant variability in how EDC is used from one Center to another.As indicated by the blue portion of the bar chart in Fig. 4, the Center with the most frequent EDC usage is Boston Center, followed by Atlanta Center and Indianapolis Center.The combined usage of these three sites alone is greater than total EDC usage at all other Centers.Although Atlanta Center is the second largest user of EDC, the frequency of Atlanta"s EDC usage is significantly less than that of inbound tactical departure scheduling into Atlanta"s arrival TMA system.Figure 4 illustrates inbound and outbound tactical departure scheduling usage.The total departure delays assigned by Arrival TMA versus EDC follow a similar model with inbound tactical departure scheduling assigning a total of 3,563 hours of surface delay to aircraft in the month of January 2011 versus a total of 480 hours of surface delay assigned by the outbound tactical departure scheduling system (13.5% of inbound).
+D. Tactical Versus Strategic Departure SchedulingWhile a significant amount of literature exists on the strategic departure scheduling process within the NAS which utilizes the Traffic Flow Management (TFM) suite of tools, information on the tactical departure scheduling process is quite limited.The two scheduling processes are distinct from one another and are currently not directly integrated.The strategic and tactical schedules have similar, but different objectives and usage characteristics.A significant difference between tactical and strategic departure scheduling is the scope of the initiative.Strategic departure scheduling is focused on correcting large demand/capacity imbalances that exist in the NAS usually due to convective weather or high demand.This often requires significant delays over an extended period of time which may be assigned hours in advance of the affected aircraft"s departure time.In contrast, tactical departure scheduling focuses on a specific air traffic flow that is subject to a local traffic management initiative (like Miles in Trail or Adjacent Center Metering) and generally introduces small delays to specific aircraft on an as-needed basis.Tactical departure scheduling system delays are approximately 4 minutes per aircraft on average with a median of 1 minute, which is significantly lower than TFM delays with approximately 66 minute average and 52 minute median delays.These statistics are derived from January 2011 operational data.The difference in average delays is likely due to the national scope of TFM which must assign departure delay well in advance of departure, in contrast with tactical departure scheduling which applies delay on an as-needed basis to a single aircraft at a time.Tactical departure schedules are able to consider the latest airspace conditions minutes before takeoff.The frequency of use of tactical departure scheduling versus strategic as measured by the number of aircraft affected for January 2011 also varies significantly as illustrated in Fig. 5.The combined number of departures scheduled using the TMA and EDC tactical decision support tools (labeled "inbound" and "outbound" tactical departures in Fig. 5) was approximately 350% greater than aircraft affected by EDCTs (strategic TFM controlled departures).It is worth noting that inbound tactical departure scheduling (i.e. using arrival TMA) occurred significantly more frequently than outbound tactical departure scheduling (i.e. using EDC).For this analysis, an aircraft was counted as being tactically scheduled only if the aircraft was both scheduled and "accepted" or "frozen" into the TMA Arrival or EDC system.A significant number of aircraft (approximately 18,489 during January, 2011) were initially scheduled in the TMA system but the scheduling process was not finalized by "accepting" or "freezing."
+III. NAS-wide Tactical Departure Scheduling Performance AnalysisIn addition to analyzing the January operational data, operational observations of scheduling performance were evaluated at DFW during the month of July 2011.Data from operational observations were used as a point of reference with which to test the data analysis measurement methodologies that were applied NAS-wide.This section discusses the metrics used for tactical departure scheduling performance and the results obtained in this analysis.Potential benefits due to reduced departure time uncertainty from PDRC can be quantified by the improvement in meeting a slot, reduction of manual intervention to mitigate missed or unattainable slots, and increased flight efficiency due to a reduction in airborne vectoring and speed controls.
+A. 'Hit Slot' MetricA key performance measurement in the tactical departure scheduling process is the efficiency with which available airspace in the constrained flow are being utilized by scheduled departure aircraft.Gaining insight into this measurement is important because it allows an objective means to analyze the utilization of tactical departure scheduling into the constrained overhead stream that may be lost due to departure prediction uncertainty.To obtain an assessment of slot utilization, operational data from the TMA and EDC systems were analyzed.A "hit slot" measurement was created for this analysis.The objective of the "hit slot" measurement is to determine whether or not the tactically scheduled departure joined the constrained flow at the sequence in which it was scheduled into prior to departure.This measurement allows an estimation of the effectiveness of the scheduling process based upon detailed scheduling information available in the operational TMA data.This section discusses details on the estimation approach used for this metric as well as results.Figure 6 provides an illustration of the "hit slot" measurement geometry for DFW to IAH tactical departure scheduling.For the "hit slot" measurement, the leading and trailing aircraft identification, TMA and EDC estimated times of arrival to the meter point (known as Meter point ETAs) and scheduled times of arrival to the meter point (known as Meter point STAs) were collected at the time at which the aircraft was scheduled in the operational TMA and EDC systems.Aircraft sequence and scheduling information were also collected at the point at which the aircraft received its first surveillance hit, and then again when it crossed the meter point location.The leading and trailing aircraft identification were examined to determine if they matched at each point in the aircraft"s flight history from scheduling, to first track, to the actual sequence at crossing.An aircraft was said to hit its scheduled slot if its sequence relative to its leading and trailing aircraft remained when it was scheduled and when it crossed the meter point location.The same "hit slot" sequencing analysis was repeated for each aircraft at the point at which surveillance was first acquired.This analysis measured whether or not the sequence provided by TMA and EDC after processing the first track hit matched the sequence at the actual meter fix crossing.This step was added to allow comparison of the difference in predictive accuracy between pre-departure scheduling versus attaining first surveillance.An important consideration of the "hit slot" measurement is determining the inclusion/exclusion criteria for aircraft to be used in the analysis.Aircraft which were excluded from the analysis included: 1) Aircraft which did not cross the meter point they were scheduled to due to lack of receipt of a crossing message, 2) Aircraft which did not have a record of leading and trailing aircraft at the point of scheduling, first track hit and crossing of the meter point based upon information available to the system at the point in time these events occurred 3) International tactical scheduling from Canada to NAS facilities given lack of departure time information available to TMA 4) Atlanta inbound tactical departure aircraft given the "hybrid metering" scenario that Atlanta uses does not allow display of metered sequence, 5) Aircraft for which a Host departure message was not received 6) For arrival TMA only metered aircraft were included, 7) Only aircraft which the TMC scheduled and "accept" or "froze" were used.To determine the sequence of aircraft at the times of interest mentioned above, the native stream class identification used by TMA and EDC was leveraged.For example, all jets scheduled over meter fix RIICE are a part of the RIICE_JETS TMA stream class.This information is made available in the native TMA data utilized for this analysis, as was the scheduled time of arrival to the meter fix (or meter point for EDC) for each stream class.The logic developed to support the "hit slot" measurement ordered all aircraft by STA from lowest to highest, by stream class.This ordering was of all aircraft which were "scheduled" in the operational TMA or EDC system, which included any tactical departure schedules that had been scheduled at that time.Upon each schedule update the leading and trailing aircraft of every flight was identified assuming one existed.If an aircraft did not have a leading or trailing aircraft in the scheduler, these values were subsequently ignored in the analysis as previously mentioned.Upon occurrences of events of interest the sequence was stored along with the other aircraft metadata for later analysis.The results from the "hit slot" analysis were separated into inbound (arrival TMA) versus outbound (EDC) tactical departure scheduling.A number of the results are represented as percentages due to inclusion/exclusion rules and data integrity checks.While certain aircraft had to be excluded to ensure data quality and that the measurements were on the right set of aircraft, the percentages are expected to hold true for the entire population of tactically scheduled departures in January due to the large sample size used for this analysis (over 22,400 aircraft after applying inclusion/exclusion logic).Table 2 shows a high-level summary of the results from running the "hit slot" measurement on all operational TMA and EDC facilities for the month of January.The "Hit Scheduled Slot %" column represents the percentage of all tactically scheduled aircraft in January 2011 that had the same leading and trailing aircraft sequence when scheduled on the surface as when they crossed the meter point being scheduled to.The "Hit First Surveillance Slot %" provides this information but uses updated sequence obtained from TMA or EDC after the first surveillance is made available.The "% Difference" takes the difference between the two hit slot percentages and then applies that percentage to all aircraft that were tactically scheduled to estimate to total number of aircraft that missed their slot due to departure time prediction uncertainty.Given that this difference provides an estimate of what the TMA and EDC algorithms had for their internal sequence prior to versus after first surveillance, this is believed to be a good estimate of slots that were missed due to departure time prediction uncertainty.Figure 7 provides an illustration of the departure events which collectively add to the uncertainty of tactical departure scheduling process.This analysis captures information from TMA and EDC system predictions that occur when the TMC scheduled the aircraft in operations prior to wheels-off, then compares this estimate to the TMA and EDC predictions immediately after wheels OFF when surveillance is first acquired.By capturing the estimates at these two time periods and comparing their difference, the ascent model portion of the prediction which is common between the two estimates, is isolated from the measurement.While a goal of tactically scheduling an aircraft into a constrained flow is to identify and utilize resources ("slots") before the aircraft departs, the impact to the NAS which occurs when a scheduled slot is not met can vary.Observed cases of missed tactically scheduled departure slots indicate that they can often lead directly to lost capacity, most notably delay caused by the case in which an aircraft is scheduled frozen into an arrival TMA slot but does not meet its expected departure time window.Other observed impacts of missing the departure slot are inefficient flight paths due to required vectoring and/or speed controls (which can lead to excess fuel utilization) as well as increased controller and TMC workload (discussed in later section).According to the hit slot metric data obtained, approximately 1 in 4 aircraft hit their arrival slot in TMA, while more than 1 in 3 hit their slot in the EDC system.The primary reason for the difference is believed to be the size of the slot being scheduled into given that the average stream class separation difference in EDC is much larger than that of TMA.Based upon operational data from January 2011, the average stream class separation for arrival TMA is 8.2 nm, while the average stream class separation in EDC is 23.6 nm.The larger separation in EDC is consistent with intuition given that EDC"s purpose is primarily to ensure MIT separations are met and the required separation being enforced is often quite large.The size of the slot being scheduled into is also believed to be the primary difference in percentage of aircraft that hit their scheduled slot in arrival TMA and EDC after the first track hit.As table 2 indicates there is a significant difference with EDC approximately 18% more aircraft hitting the slot at this point in time versus arrival TMA.The percentage  of aircraft that hit their slot after surveillance suggest that there may be room for improvement in the predictive capabilities of the ascent modeling of TMA and EDC.Future analysis may be warranted to analyze predictive accuracy of the ascent modeling due to aircraft weight, wind error, inaccurate routing, etc. While, on average, aircraft hit their TMA-scheduled slots approximately 26.9% of the time, a fairly significant variation exists by site.The results of the hit slot metric were calculated for all TMA and EDC locations nationwide.The highest site percentage of the "hit slot" measurement of all the arrival TMA systems was 32.9%, while the lowest was 18.5% The highest site percentage of all EDC systems was 52.5%, with the lowest being 22.7%.The site specific variance may warrant additional consideration to determine the primary factors which lead to the variance.Given that the "hit slot" percentage differs on a site by site basis, this suggests that the impact to the NAS may vary by facility as well.
+B. Arrival Metering Workload metricIn addition to missed slots from departure time uncertainty, another shortfall to consider in current day tactical departure scheduling is the workload for the TMC and controllers.During the month of January 2011 approximately 153,426 flights had metering information delivered to sector controllers with the expectation that the controller would delay aircraft as necessary to meet the metered times.Of the metered aircraft, approximately 34,360 (22.4%) were scheduled into the arrival stream using arrival TMA arrival scheduling capability.This represents a statistically significant portion of the overall metered aircraft during January.The large sample of metered flights was analyzed to determine if manual intervention by either the sector controller or TMC during metering was higher for tactically scheduled departures than for flights which were not tactically scheduled.Three measures were utilized for this evaluation, which were the frequency controller swaps, controller resequences and individual aircraft reschedules by the TMC.The following gives a brief explanation of what these measures capture.Sector controller tools associated with metering include two capabilities to control the sequence that TMA associates with arrival aircraft.These capabilities are known as swap and re-sequence.The swap capability allows the controller to identify any two aircraft on their display and exchange their meter point crossing times.This capability is used when the sector controller may disagree with the sequence or times that are being presented to him/her by the TMA system.The tactically scheduled departure aircraft and the flights which were not tactically scheduled were analyzed to determine the frequency of required manual activity.The increased percentage of aircraft that required manual controller or TMC activity during metering suggests that tactical departure scheduling is a factor in increased workload for both sector controllers and TMCs.The highest increase of manual activity observed was the percentage increase of aircraft that undergo a single aircraft re-schedule.This measure showed a 6.1% increase for tactically scheduled departures over those aircraft which were not tactically scheduled.A summary of these results can be seen in Table 3.
+C. Effect of not scheduling a tactical departure into a constrained flowObservations of tactical scheduling performance from DFW into IAH during June and July of 2011 indicate that the benefit of increased departure time prediction accuracy may not be limited to the set of tactically scheduled departures previously discussed.Examples of these potential benefits were observed during PDRC engineering shadow evaluations.A typical example of this was for aircraft departing DFW with a destination of IAH which were not scheduled in the TMA system.In these examples the departing aircraft was sequenced ahead of several other aircraft in the stream class that were in close proximity.The addition of the departing aircraft added a 1 minute delay to the immediate trailing aircraft, which in turn added two minutes of delay to its trailing aircraft, and so on for a total of four aircraft which received airborne delay due to the departing aircraft.Vectoring off of nominal routes was visually observed in a number of these cases.During PDRC observations in July, a number of occurrences were noted in which departures that were not tactically scheduled and coordinated between Center and ATCT personnel resulted in the use of speed controls and/or vectoring to accommodate the departing aircraft.During evaluations the "not scheduling" scenario which leads to this situation was discussed with Center personnel.Comments received indicate that while additional work is needed by sector controllers to accommodate uncoordinated departures, this is not viewed as an issue for sector controllers so long as other sector workload does not rise to a level of saturation that makes handling uncoordinated departure scheduling problematic.This information is consistent with previous research into the effect of "not scheduling" an aircraft into an arrival TMA flow. 3,4However, beyond the sector workload implications is the consideration of flight efficiency which effect fuel consumption.A coordinated departure release may have helped to reduce speed controls and vectoring which may in turn help reduce fuel consumption.
+IV. Surface Departure Prediction AnalysisThe objective of PDRC is to leverage trajectory-based OFF time predictions to improve upon the current-day tactical departure scheduling process.Achieving this objective requires that one have accurate OFF time predictions from the surface system at the point in time which this information is required by the en route scheduling system.This section discusses a method to estimate the minimal required look-ahead time for OFF time predictions to satisfy tactical departure scheduling requirements.Also discussed are surface departure prediction accuracy requirements for present-day operations as well as recommendations for future surface analysis.
+A. Estimation of departure prediction look-ahead time requirement for Tactical Departure SchedulingIn an ideal scenario, highly accurate aircraft wheels OFF times would be available to tactical and strategic planners hours ahead of the point at which the aircraft was ready to depart.In this ideal scenario all planners would be working from the same set of accurate information and making decisions that could be used to address local, regional, or national demand/capacity imbalances.However, highly accurate OFF times hours in advance of departure is not a feasible objective given the amount of pre-departure uncertainty which exists today. 3,4,6,7The cumulative effect of uncertainty from pushback prediction, through ramp taxi, spot transition, air movement area taxi, departure queue management, departure release, take off roll, ascent modeling, and forecast wind errors prior to reaching the meter crossing point provide a large amount of unpredictability.This uncertainty makes the departure planning process quite challenging.While accurate wheels OFF estimates hours in advance may be an unrealistic objective in the NAS, providing accurate OFF time estimates minutes in advance of wheels OFF is an achievable objective which may help reduce or eliminate some of the challenges faced by tactical departure scheduling.An important question to consider for departure prediction accuracy is "how far in advance of departure does the downstream scheduling system need to have accurate OFF time predictions?"In order to estimate the minimal look-ahead time at which accurate OFF time predictions are required for aircraft departing into an arrival metering flow, one should consider the relative positions of the departure airport and the arrival metering freeze horizon.The geometry of the DFW-to-IAH metering scenario is illustrated in Fig. 8. DFW airport lies within the IAH arrival metering freeze horizon and the standard tactical departure scheduling procedure is to accept and freeze the aircraft into the arrival IAH flow to prevent the aircraft from receiving both a ground delay and an airborne delay.Due to this scheduling methodology, any surface or airborne prediction error in tactical departure scheduling to IAH during metering directly impacts the airborne arrival stream.For present-day operations this OFF time prediction is entirely manual.For the DFW to IAH metering scenario, the typical airborne aircraft scheduled into IAH over meter fix RIICE freezes at approximately 30.4 minutes prior to meter fix crossing when IAH traffic is in East flow, which is the predominant configuration used during metering at IAH.The typical flight time from DFW airport to the RIICE meter fix crossing is approximately 27.7 minutes.This means that an aircraft on the DFW surface which is ready to depart will be competing for slots with airborne aircraft whose schedules have been frozen on average for 30.4 -27.7 = 2.7 minutes (162 seconds).If the DFW aircraft are to compete with unfrozen aircraft for a slot into the constrained flow then the tactical scheduling process must occur at least 162 seconds prior to departure.The 162 second figure represents a theoretical minimum for the tactical departure scheduling lead time.Additional time is required for the Center TMU to consider the schedule and communicate the release time to ATCT.Some time is also required for the TMA scheduler to find a slot for the aircraft in its schedule and optimize the overall arrival stream schedule based upon the new information.The time needed for scheduling purposes in addition to the theoretical 162 seconds is being called the "coordination time" in Fig. 8.Operational observations of PDRC at DFW during July 2011 have revealed that the typical departure schedule process is initiated approximately 5 minutes prior to departure during Call For Release situations.According to ATCT and Center personnel this amount of time prior to departure allows for sufficient coordination and meets the minimal need for look-ahead time requirements at DFW.That is not to say that both ATCT and Center don"t want the times earlier, but this was an acceptable timeframe for the manually-coordinated tactical departure scheduling process in place today.Considering site feedback and the 2.7 minute flight time difference which would allow these aircraft to compete with non-frozen aircraft in the IAH metered stream, this allows approximately 2.3 minutes of "coordination time" for the tactical departure scheduling process at DFW.It is believed that this look ahead time estimation process can be used for other airports that have a high demand for tactical departure scheduling to identify the look ahead time at which accuracy departure time predictions are needed.Based upon PDRC field test observations as well as data obtained from FAA evaluation of TMA scheduling from air traffic control towers, 8,9 it is estimated that through automation the "coordination time" taken for the tactical departure scheduling process can be reduced to approximately 30 seconds.Thus, the minimal look ahead time requirement for DFW is 162 + 30 = 192 seconds prior to wheels OFF.
+B. Surface prediction accuracy at required look-ahead time for Tactical Departure SchedulingThe look-ahead time need was based upon relative geometry of the departure airport to the arrival metering freeze horizon plus required coordination time.Look-ahead requirement will likely vary based upon different airport geometry relative to arrival metering freeze horizons, or the airspace geometry associated with EDC flows.Beyond the look-ahead requirement, there remains the question of required departure prediction accuracy at the specified look-ahead time.The departure prediction accuracy requirement may be estimated from observed CFR time compliance in today"s tactical scheduling scenario.If surface automation delivers the same level of accuracy provided today by the manual CFR procedure, then it follows that it should provide similar benefit to the existing system.Any increase in the accuracy of the departure prediction times or increased look-ahead time for the prediction would be potentially beneficial to tactical departure scheduling system performance.An additional observation to consider is that workload associated with the manual CFR procedure may lead to relatively infrequent use.Any automation that may help reduce the workload threshold at which this level of accuracy could be obtained would likely be used more frequently, which would potentially lead to increased benefits.Another factor to consider is that of any uncertainty that is the result of manual entry or miscommunications like those reports in a companion paper. 1 Currently, the manual CFR procedure must deliver OFF times that comply with a -2/+1 minute window.Based upon tactical departure scheduling data for the month of January 2011, this time window is being met approximately 62% of the time by ATCT control of flights to meet their CFR coordinated OFF time.Based upon measurements obtained of the Surface Decision Figure 9. SDSS prediction accuracy at DFW -June 2011.Support System (SDSS) accuracy in June of 2011, SDSS can predict aircraft wheels OFF at the same level of controlled CFR flights at approximately 137 seconds prior to OFF time.That is to say that without any CFR manual coordination required (e.g.closed loop system); SDSS can achieve similar levels of predictive accuracy as departure time compliance being achieved today through the CFR process at 137 seconds prior to departure.To meet the tactical departure scheduling requirements for DFW, this level of accuracy must be extended at least to the point of 162 seconds as mentioned previously including any coordination time required for the tactical departure flight.However, it is not necessarily true that SDSS must provide this level of accuracy out to the five minutes which current DFW procedure provides.This is due to the coordination time required when using automation is expected to be reduced from the time it takes in the current procedure.During the initial evaluation of PDRC the focus was on establishing confidence in the surface and en route scheduling components, not on reducing the time period it takes for tactical departure scheduling to occur.Future evaluations should work to increase the amount of look ahead time that accurate OFF time predictions are available while reducing the amount of coordination time required for the tactical departure scheduling process.Work is currently underway to increase the accuracy of the existing surface management system"s predictive capability for those aircraft which have acquired surface surveillance.In addition to the increasing the system"s predictive accuracy, areas of research that are recommended are: stability of the OFF time estimates which are provided to the downstream scheduler, utilization of departure prediction confidence in tactical departure scheduling, evaluation of tactical scheduling methods which require OFF time estimates in excess of 10 minutes prior to departure and expansion of OFF time estimates to include airports without ASDE-X surveillance capability.
+V. NAS facilities likely to have greatest benefit from PDRC TechnologyGiven knowledge of the current tactical departure scheduling demand at each NAS facility, as well as estimated look ahead time requirements for each facility based upon geometry like that illustrated in Fig. 8, a list of the top NAS facilities which would benefit from PDRC technology was constructed.This survey focused on inbound tactical departure scheduling since 86.5% of tactical departure scheduling ground delay incurred in the NAS today is scheduled in this manner.The estimation methodology begins with sites that have a proven demand for tactical departure scheduling like those listed in Table 1.Only the top 10% users of tactical departure scheduling airport pairs (e.g.KDFW into KIAH) excluding international scheduling were considered.This yielded 81 airports scheduling into 7 different metered airports, each of which tactically scheduled over 130 aircraft during the month of January.The next step was to analyze each departure/arrival airport pair to determine the lookahead time need of each airport, like that illustrated in Fig. 8.In order to include look-ahead time needs that are achievable based upon surface surveillance availability, it was necessary to bound the look-ahead time by the average surface taxi out time.The nationwide average of unimpeded taxi out time of 10.7 minutes was obtained from the FAA"s Aviation System Performance Metrics (ASPM) database.Those airports with greater than 10.7 minutes look-ahead time requirement prior to departure were eliminated from the list, which left 55 airports.The remaining candidate airports were further filtered according to current or planned availability of an ASDE-X surface surveillance system which would allow for trajectory based OFF time estimates to be supplied to the tactical departure scheduler.This remaining list consisted of 26 airports, which were ordered by the delay they incurred in January 2011, as listed in Table 4.The "Scheduling From" column in Table 4 indicates the airport from which tactical departure scheduled aircraft are departing, while the "Scheduled into Metered Airport" indicates the destination of the tactical departure scheduled.At the top of the list are two airports that are not only ASDE-X equipped, but also have a current Surface Decision Support System (SDSS) adapted.In addition, the third and fourth airports on the list are currently being adapted for the SDSS system in support of other research.A notable omission from Table 4 is scheduling from Charlotte to Atlanta.While 426 aircraft were tactically scheduled from Charlotte to Atlanta during the month of January, only 35 of these occurred during an Atlanta metering period.The lack of tactical departure scheduling during metering may be due to the "hybrid metering" design that Atlanta uses in which adjacent centers meters outside of Atlanta Center airspace but the metering advisories are not displayed on Atlanta Center glass.Analysis of site geometry relative to the freeze horizon indicates that the look-ahead time at which accurate departure predictions are needed becomes greater as the distance from the departure airport within the freeze horizon increases.Inbound tactical departure scheduling analysis has demonstrated that the majority of scheduling occurs near the arrival freeze horizon boundary (11.3 minute average flight time to freeze horizon with 11.4 minute standard deviation).Some of the airports being scheduled from to an arrival metering facility lie geographically inside of the freeze horizon, while others lie outside of the freeze horizon.Heavier usage of tactical departure scheduling near the freeze horizon is consistent with intuition as flights which are sufficiently far away from the TMA freeze horizon generally have sufficient time and space in the arrival stream in order to secure a slot prior to the freeze horizon location.As departing airports get closer to or are within the TMA freeze horizon, the scheduling process becomes more dependent upon the departure prediction accuracy as there is less time for a departing aircraft to compete for resources in the overhead stream while the demand for overhead resources generally also becomes greater.In this manner the geometry of a departure airport relevant to the freeze horizon of the arrival TMA system being scheduled into is an important factor to consider.Figure 10 illustrates this relationship which is being referred to as the "Goldilocks Zone" in which achievable levels of departure prediction accuracy can be used for tactical departure scheduling.The following example considers if a departure airport requires 15 minutes flying time within the arrival freeze horizon to an arrival metering facility.To actively compete with non-frozen aircraft which are currently airborne in the arrival stream, the look-ahead time predictions must be accurate enough for TMA at least 15 minutes prior to departure.Any error in the departure prediction estimate scheduled at this point will directly impact the arrival stream efficiency as well as controller workload if the sector controller meter list is rippled due to changes.On the other hand if the departure airport is 60 minutes flying time outside of the freeze horizon, then despite the level of departure prediction accuracy, the aircraft will likely have adequate time to be scheduled into the arrival TMA system.
+VI. ConclusionsAnalysis of operational TMA and EDC data from all current deployed facilities covering over 1,082,000 flights during the month of January 2011 indicates that these tactical departure scheduling capabilities are widely used in the NAS today with over 65,000 scheduled aircraft per month using these methods.Increased utilization of tactical departure scheduling decision support tools has been fueled by expansion of adjacent center metering and nation-wide deployment of the EDC capability.Although tactical departure scheduling with TMA and EDC has become a widely used component in NAS operations today and represents a significant improvement over the previous process which lacked trajectory based ascent modeling, analysis of the current system"s performance indicates that significant room for improvement exists by reducing departure time uncertainty.Based upon operational data analysis described in this paper, 6,792 inbound tactically scheduled aircraft and 1,911 outbound tactically scheduled aircraft in January 2011 NAS wide are estimated to have missed the airspace slot they were scheduled into due to departure time prediction uncertainty.The effect to the NAS of a missed scheduled departure slot often leads directly to lost capacity, most notably in the case in which an aircraft is scheduled frozen into an arrival TMA slot but does not meet its expected departure time window.However, measuring the impact to the NAS of a missed departure slot is not always straightforward as some ability to recover the airspace resources exists, often at the cost of additional TMC or controller workload and/or inefficient flight paths.While the shortfalls of the existing tactical departure scheduling system have become more evident and quantifiable, solutions to these shortfalls are in early stages of maturity relative to other NAS systems.Determining the level of predictive accuracy that trajectory based OFF time predictions must attain for tactical departure scheduling delay reduction benefit is complicated by the lack of surface automation available in operations today and the challenges associated with evaluating a passive OFF time estimation process.This paper proposes metrics and methods to estimate the look ahead time requirement of surface predictions, as well as to identify target airports that are likely candidates for NAS deployment of PDRC technology based upon the departure airport"s geometry relative to areas of high airspace demand like those encountered near time based metering freeze horizons.Indications are that departure prediction accuracy requirements for tactical departure scheduling in the NAS are likely not a single value, but rather a range of values that vary in significant part based upon site specific geometry and airspace demand.Figure 1 .1Figure 1.Inbound tactical scheduling geometry which requires a high level of departure prediction accuracy.
+Figure 2 .2Figure 2. Average Nationwide Departure time compliance for January 2011.
+55
+Figure 3 .3Figure 3. Example of TMA Flow Program into Atlanta.
+Figure 4 .4Figure 4. Tactical Scheduling of Arrival TMA and EDC -Jan 2011.
+Figure 5 .5Figure 5. Departures Scheduled with Decision Support Tool -Jan 2011.
+Figure 6 .6Figure 6."Hit Slot" metric geometry for DFW to IAH Scheduling.
+Figure 7 .7Figure 7. Tactical departure scheduling to the meter point incorporates cumulative uncertainty from a number of departure events.
+Figure 8 .8Method to estimate OFF prediction look ahead time need for DFW aircraft departing into Houston arrival metering.
+Figure 10 .10Figure 10.Inbound tactical departure scheduling 'Goldilocks Zone' relationship between departure airport location and freeze horizon.
+
+3,4Metering Freeze HorizonDeparture AirportOuter Meter ARCMeter FixArrival Airport
+Table 11gives examples and frequency of usage of inbound tactically scheduled aircraft across Center boundaries during the month of January 2011.The "Number of Aircraft Scheduled into remote Center" lists the number of times a TMC from a Center other than the destination Center scheduled aircraft using TMA capability.Note that not all scheduling performed is from an adjacent center, for instance Indianapolis Center schedules into New York Center although the two Centers do not share a boundary.Another unique case occurs when aircraft departing Canadian airspace Call For Release into New York Center airspace.
+Table 1 . Departure Scheduling from remote ARTCC Jan 2011.1Number of AircraftScheduled intoFrom Center Into Centerremote CenterJacksonvilleAtlanta6267WashingtonAtlanta6072BostonNew York3955WashingtonNew York3719IndianapolisAtlanta3081ClevelandNew York3012OaklandAtlanta2951Los AngelesAlbuquerque2243MemphisAtlanta1619CanadaNew York1234IndianapolisNew York469ClevelandAtlanta389Albuquerque Los Angeles384Fort WorthHouston382ChicagoCleveland210Kansas CityChicaco
+Table 2 . 'Hit Slot' measurement results for all operational TMA/EDC facilities during January 2011. System Hit Scheduled Slot % Hit First Surveillance Slot % % Difference Estimated Number of Aircraft that missed their slot due to departure time prediction uncertainty2Arrival TMA26.939.312.56792EDC39.457.117.71911
+Table 3 . Percentage of aircraft which required manual intervention-Jan 2011. Workload Category Not Tactical Departure % Tactical Scheduled Departure % % Difference Approximate # Aircraft subject to increased manual activity Controller Swaps34.46.62.3792Controller Re-sequences4.46.01.7572Single Aircraft Re-schedule5.011.16.12125
+Table 4 . Sites which would benefit from PDRC technology -Jan 2011.4ScheduledScheduled IntoNumber ofFromMetered AirportHoursScheduledAirport CodeScheduling From Airport NameCodeScheduling Into Metered Airport NameDelayAircraftKMCOOrlando InternationalKATLHartsfield -Jackson Atlanta International47.9628KMEMMemphis InternationalKATLHartsfield -Jackson Atlanta International38.0381KATLHartsfield -Jackson Atlanta InternationalKCLTCharlotte/Douglas International32.4426KBOSLogan InternationalKPHLPhiladelphia International28.0385KLASMc Carran InternationalKLAXLos Angeles International18.8381KIADWashington Dulles InternationalKCLTCharlotte/Douglas International17.4263KDTWDetroit Metropolitan Wayne CountyKPHLPhiladelphia International16.1278KSDFLouisville InternationalKATLHartsfield -Jackson Atlanta International15.9230KCLECleveland-Hopkins InternationalKPHLPhiladelphia International15.7203KLAXLos Angeles InternationalKLASMc Carran International15.4318KSFOSan Francisco InternationalKLAXLos Angeles International15.0333KDFWDallas/Fort Worth InternationalKIAHGeorge Bush Intercontinental/Houston13.3168KCVGCincinnati/Northern Kentucky InternationalKCLTCharlotte/Douglas International12.9258KDCARonald Reagan Washington NationalKCLTCharlotte/Douglas International12.0246KBWIBaltimore/Washington InternationalKCLTCharlotte/Douglas International11.4271KPHXPhoenix Sky Harbor InternationalKLASMc Carran International11.1196KCVGCincinnati/Northern Kentucky InternationalKATLHartsfield -Jackson Atlanta International10.8199KSANSan Diego InternationalKPHXPhoenix Sky Harbor International7.3189KSJCNorman Y. Mineta San Jose InternationalKLAXLos Angeles International7.2168KLASMc Carran InternationalKPHXPhoenix Sky Harbor International6.6200KMCOOrlando InternationalKCLTCharlotte/Douglas International6.1250KLAXLos Angeles InternationalKPHXPhoenix Sky Harbor International5.7213KSDFLouisville InternationalKCLTCharlotte/Douglas International5.7190KSNAJohn Wayne-Orange CountyKLASMc Carran International5.5140KSNAJohn Wayne-Orange CountyKPHXPhoenix Sky Harbor International3.6173KSFOSan Francisco InternationalKLASMc Carran International2.6154
+		
+		
+			
+
+				
+
+
+	
+		Trajectory-Based Takeoff Time Predictions Applied to Tactical Departure Scheduling: Concept Description, System Design, and Initial Observations
+		
+			ShawnEngelland
+		
+		
+			RichardCapps
+		
+		10.2514/6.2011-6875
+	
+	
+		11th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference
+		Virginia Beach, VA
+		
+			American Institute of Aeronautics and Astronautics
+			September 20-22, 2011
+		
+	
+	submitted to AIAA 11th Aviation Technology
+	Engelland, S., Capps, A, "Trajectory-Based Takeoff Time Predictions Applied to Tactical Departure Scheduling: Concept Description, System Design, and Initial Observations", submitted to AIAA 11th Aviation Technology, Integration, and Operations (ATIO) Conference, Virginia Beach, VA., September 20-22, 2011.
+
+
+
+	
+		Benefit Assessment of Precision Departure Release Capability Concept
+		
+			KeePalopo
+		
+		
+			GanoChatterji
+		
+		
+			Hak-TaeLee
+		
+		10.2514/6.2011-6834
+	
+	
+		11th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference
+		Virginia Beach, VA
+		
+			American Institute of Aeronautics and Astronautics
+			September 20-22, 2011
+		
+	
+	submitted to AIAA 11th Aviation Technology
+	Palopo,K., Lee, H, Chatterji, G., "Benefit Assessment of the Precision Departure Release Capability Concept", submitted to AIAA 11th Aviation Technology, Integration, and Operations (ATIO) Conference, Virginia Beach, VA., September 20-22, 2011.
+
+
+
+	
+		Mitigating the Effect of Demand Uncertainty due to Departures in a National Time-Based Metering System
+		
+			StevenLandry
+		
+		
+			AlvaroVillanueva
+		
+		10.2514/6.2007-7713
+	
+	
+		7th AIAA ATIO Conf, 2nd CEIAT Int'l Conf on Innov and Integr in Aero Sciences,17th LTA Systems Tech Conf; followed by 2nd TEOS Forum
+		Belfast, Northern Ireland
+		
+			American Institute of Aeronautics and Astronautics
+			September 18-20, 2007
+		
+	
+	AIAA's 7th Annual Aviation Technology, Integration, and Operations (ATIO) Technical Forum
+	Landry, S. J., and Villanueva, A., AIAA-2007-7713, "Mitigating the Effect of Demand Uncertainty Due to Departures in a National Time-Based Metering System". AIAA's 7th Annual Aviation Technology, Integration, and Operations (ATIO) Technical Forum, Belfast, Northern Ireland, September 18-20, 2007.
+
+
+
+	
+		Effects of the uncertainty of departures on multi-center traffic management advisor scheduling
+		
+			JThipphavong
+		
+		
+			SJLandry
+		
+		AIAA-2005-7301
+		
+			September 26-28, 2005
+			Arlington, VA
+		
+	
+	AIAA's 5th Annual Aviation Technology, Integration, and Operations (ATIO) Technical Forum
+	Thipphavong, J., and Landry, S. J., "Effects of the uncertainty of departures on multi-center traffic management advisor scheduling.", AIAA-2005-7301, AIAA's 5th Annual Aviation Technology, Integration, and Operations (ATIO) Technical Forum, Arlington, VA., September 26-28, 2005.
+
+
+
+	
+		Traffic Management Advisor Flow Programs: an Atlanta Case Study
+		
+			ShonGrabbe
+		
+		
+			BanavarSridhar
+		
+		
+			AvijitMukherjee
+		
+		
+			AlexMorando
+		
+		10.2514/6.2011-6533
+	
+	
+		AIAA Guidance, Navigation, and Control Conference
+		Portland, Oregon
+		
+			American Institute of Aeronautics and Astronautics
+			August 08-11, 2011
+		
+	
+	Grabbe, S., "Traffic Management Advisor Flow Programs: an Atlanta Case Study", AIAA Guidance, Navigation, and Control Conference, Portland, Oregon, August 08-11, 2011.
+
+
+
+	
+		Multi-Center Traffic Management Advisor: Operational Test Results
+		
+			ToddFarley
+		
+		
+			StevenLandry
+		
+		
+			TyHoang
+		
+		
+			MonicarolNickelson
+		
+		
+			KerryLevin
+		
+		
+			DennisRowe
+		
+		
+			JerryWelch
+		
+		10.2514/6.2005-7300
+		AIAA-2005-7300
+	
+	
+		AIAA 5th ATIO and16th Lighter-Than-Air Sys Tech. and Balloon Systems Conferences
+		Arlington, VA
+		
+			American Institute of Aeronautics and Astronautics
+			September 26-28, 2005
+		
+	
+	Farley, T. C., Landry, S. J., Hoang, T., Nickelson, M., Levin, K. M., Rowe, D., and Welch, J. D., "Multi-Center Traffic Management Advisor: Operational Test Results," AIAA-2005-7300, Proceedings of the 5th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference, Arlington, VA, September 26-28, 2005.
+
+
+
+	
+		Analysis of En Route Sector Demand Error Sources
+		
+			JimmyKrozel
+		
+		
+			DanRosman
+		
+		
+			ShonGrabbe
+		
+		10.2514/6.2002-5016
+		AIAA-2002-5016
+	
+	
+		AIAA Guidance, Navigation, and Control Conference and Exhibit
+		Monterey, California
+		
+			American Institute of Aeronautics and Astronautics
+			August 5-8, 2002
+		
+	
+	Krozel, J., Rosman, D., Grabbe, S., "Analysis Of En Route Sector Demand Error Sources", AIAA-2002-5016, AIAA Guidance, Navigation, and Control Conference and Exhibit, Monterey, California, August 5-8, 2002.
+
+
+
+	
+		Linking Traffic Management to the Airport Surface: Departure Flow Management and Beyond
+		
+			NDoble
+		
+		
+			JTimmerman
+		
+		
+			TCarniol
+		
+		
+			MKlopfenstein
+		
+		
+			MTanino
+		
+		
+			VSud
+		
+	
+	
+		Eighth USA/Europe Air Traffic Management Research and Development Seminar (ATM2009)
+		Napa, CA
+		
+			29 Jun -2 Jul 2009
+		
+	
+	Doble, N., Timmerman, J., Carniol, T., Klopfenstein, M., Tanino, M., and Sud, V., "Linking Traffic Management to the Airport Surface: Departure Flow Management and Beyond," Eighth USA/Europe Air Traffic Management Research and Development Seminar (ATM2009), Napa, CA, 29 Jun -2 Jul 2009.
+
+
+
+	
+		TMA En Route Departure Capability Jacksonville ARTCC and Orlando ATCT Usage Data
+		
+			SFutato
+		
+		
+			KMcmillan
+		
+		
+			SCallon
+		
+		
+			April 8, 2008
+		
+	
+	Futato, S., McMillan, K., Callon, S., "TMA En Route Departure Capability Jacksonville ARTCC and Orlando ATCT Usage Data", April 8, 2008.
+
+
+				
+			
+		
+	

file115.txt

@@ -0,0 +1,475 @@
+
+	
+	
+		
+I. Introductionecent NASA research [1][2][3] has focused on improving tactical departure scheduling in scenarios where wellequipped airport Towers interact directly with Center Traffic Management Units (TMUs) to implement departure management initiatives such as Call For Release (CFR).The research presented in this paper is part of an effort to extend tactical departure scheduling improvements to lesser-equipped airports and to address constraints that exist in the terminal environment.The FAA's Next Generation Air Transportation System (NextGen) plans, 4,5 call for the ability to accurately schedule a flight from its departing gate to its arrival gate in advance of its actual gate departure (i.e.gate-to-gate scheduling).Specifically, gate-to-gate scheduling presumes the planning and control of a flight from its departure gate to the runway, to the terminal departure fix, Center departure metering fix, through En Route airspace to the arrival metering fix, runway and finally to the arrival gate.For gate-to-gate scheduling to be effective in the NextGen environment, surface, terminal, Center, and national constraints must all be simultaneously satisfied by the departure scheduling tool.NextGen gate-to-gate scheduling also requires accurate prediction and execution of, trajectory-based operations in the terminal area.Observations at the Dallas/Fort Worth communication requirements all add substantial workload to air traffic control.Thus, a terminal solution that increases workload during these busy periods is unlikely to be accepted by operational personnel.
+B. Terminal Departure SchedulerThe prototype terminal departure scheduler seeks to resolve many of the unique challenges mentioned in the previous section.The following sections discuss the process employed to sequence and schedule each flight.
+SequencingThe terminal departure scheduler gives greater priority to the flights that are ordered earlier.The process the terminal departure scheduler uses to decide what order is used in scheduling is referred to as the sequencing logic.This section briefly describes the sequencing logic which was derived, in large part, from two existing schedulers, the Traffic Management Advisor (TMA) Dynamic Planner (DP) 18 and the Surface Decision Support System (SDSS) 19 surface scheduler.These two schedulers were chosen as a basis for terminal departure scheduling logic because of their relevance to the problem at hand, demonstrated success in operational enviornments and their ability to handle flights at various stages in the departure process.The terminal departure scheduler runs on a user defined periodic rescheduling interval, hereafter referred to as the scheduling cycle.Currently a five second scheduling cycle is used for processing given it matches the frequency of position data updates.For each scheduling cycle, flights are re-sequenced and rescheduled.The sequencing order ensures that flights which are higher priority (see Table 1) from an operational readiness standpoint are scheduled first and that frozen flight times do not change from one iteration to the next.Table 1 describes the categories that are used to sequence flights.This table is listed in priority order from highest to least.Thus all flights that are in the first category (crossed departure fix) are scheduled prior to the flights in the second category (terminally controlled airborne).While some flights may fall into multiple traffic management initiative (TMI) categories, a flight can only belong to one sequencing category.The highest priority sequencing category a flight qualifies for is assigned to it.Thus, a flight with both a terminally controlled frozen OFF time and a CFR will be assigned the higher priority sequencing category associated with terminally controlled frozen flights.Each sequencing category has its own sorting rules.For flights that have already crossed the departure fix, they are sorted by their crossing time.Airborne flights that have yet to cross the departure fix are sorted by their undelayed estimated departure fix crossing time.Surface flights which have a TMI use the controlled OFF time associated with that constraint, while all other surface flights use their undelayed estimated OFF time to determine sequencing order.
+Scheduling Processing LogicRather than accomplishing this with a single monolithic entity, a collection of smaller schedulers are orchestrated by a master scheduler.This design approach was chosen to model industry best practices of loosely coupled, course grained services 20 and to maximize reuse of existing components from prior research.The individual schedulers can be seen at the top of Fig. 1 and are comprised of a pre-scheduler, departure fix scheduler, airport scheduler and a post-scheduler.The terminal departure scheduling process begins by a call to the master scheduler from the main processing logic of the terminal departure system.The frequency of the call to the master scheduler is configured in system properties files.For this research a frequency of five seconds was utilized.Once invoked, the first step the master scheduler performs is initializing a temporary copy of the flight object for the scheduling process.The primary purpose of this activity is to know the starting time for each flight and allow it to be bound by time ranges in later processing if necessary.Given the scheduler must resolve times at multiple locations (airport and fix) a temporary copy of the flight is initialized at both locations.If the flight has a CFR or EDCT time, this time is used to set the latest time the flight can depart.For CFR and EDCT flights, a single minute-level of granularity is used for departure time as opposed to a departure time window.After initializing the flights, the master scheduler calls the pre-scheduler component.The pre-scheduler's primary role is to address flights that have missed their terminally controlled OFF times and thus need to be rescheduled.The pre-scheduler will evaluate all the flights being scheduled to determine if any have missed their coordinated OFF time by the configured number of seconds.If so, the flight will lose its controlled time which will result in the flight having lower sequencing priority as described in the previous section.Once flights have been initialized, they are sorted according to the sequencing categories listed in Table 1.Each flight then undergoes scheduling from earliest to latest in each sequencing category.For each flight, the terminal departure transit time is calculated.For the simulation, this transit time is supplied by a flight time decision tree and terminal departure transit time error described later in this document.For realtime prototype system processing, the terminal departure transit time prediction is provided by the research Traffic Management Advisor (rTMA) system.The rTMA terminal departure transit time prediction includes the effect of winds at crossing altitude.Once the flight time is calculated, it remains constant for the remainder of the scheduling cycle assuming no changes to departure fix have occurred.If the flight is airborne, the remaining terminal departure transit time will be calculated by subtracting the amount of time already spent in transit.Next, the scheduler resolves the departure fix and airport times for each flight.The following steps are taken until both the departure fix time and runway departure time are fully resolved.Fully resolved means that both locations have a time that has no scheduling conflict with other flights and meets all specified traffic management constraints.To accomplish this objective the master scheduler calls the departure fix scheduler which schedules each flight to the appropriate departure fix based upon the OFF time estimate and the terminal departure transit time.Once a departure fix time is obtained, the scheduler then calls the airport scheduler.To resolve the flight's scheduled OFF time, the airport scheduler uses the later of the initial OFF time calculated in an earlier step or the adjusted OFF time derived from the departure fix time.The reason for this is the derived OFF time can be no earlier than the flight can achieve.The adjusted OFF time is calculated by subtracting the terminal departure transit time from the resolved departure fix crossing time.If a CFR or EDCT time exists for this flight, this constraint is taken into account in the airport schedule time.In the current implementation, the SDSS scheduler is used for Dallas Fort Worth (DFW) flights.A simplified model was used for less-equipped airports, which included all D10 airports except DFW.The simplified model for less-equipped airports utilizes fewer site adaptation files than the SDSS scheduler.If the resulting OFF time from the airport schedule time matches the required OFF time to meet the departure fix, scheduling for this flight is complete.If the times do not match, the scheduling process continues again using the updated OFF time as a starting point for scheduling to the departure fix.Once all flights are scheduled in the manner described above, the master scheduler calls the post-scheduler process.The primary purposes of the post scheduler are to determine if a flights should be frozen, store the data and distribute the data to the appropriate processes.The purpose of freezing a flight is to ensure that the controlled departure time that has been communicated to a surface local controller does not change.A flight is frozen if its time to departure is less than the configured freeze horizon value.The results from the scheduling process for each flight are stored in the terminal database.The purpose of the database is to assist in system processing and post operational analysis.The results from the scheduling process are then distributed to the other system processes.In real-time operations, this makes the scheduling results available to Tower personnel.
+III. Terminal Departure Scheduling Simulation CapabilityThe objective of this work is to develop a departure scheduler that can be assessed by air traffic personnel in an operational terminal departure environment.The term departure scheduler refers to a software program that is used by terminal personnel to schedule flights from multiple departure airports within their control which possess varying levels of OFF time precision.The departure scheduler receives real-time flight planning data, surface OFF time estimates and terminal transit time estimates as input and produces a controlled wheels off (OFF) time for each flight which meets all required air traffic constraints on the runway, departure fix, downstream Center metering points and strategic traffic management initiatives.The OFF time provided by the scheduler ensures that minimal separation is maintained at both the runway threshold and departure fix.The OFF times provided to operational personnel are expected to be treated as a controlled OFF time.That is, air traffic personnel will actively control the flight to meet the departure time similar to the process used for EDCT and CFR controlled times.Fast-time simulation was used to better understand terminal departure scheduler performance when subjected to variances in OFF time error, flight time error and varying traffic constraints.The terminal departure scheduler used in this simulation is also expected to execute as the prototype scheduling software that will be used in terminal departure prototype system processing.The terminal departure prototype is a new decision support tool being developed and evaluated in the D10 terminal environment.To achieve the objective of using the same scheduler for both fast-time simulation and prototype processing, an evaluation harness was developed to allow the terminal departure scheduler to be evaluated in multiple modes including; real-time data mode, playback mode and simulation mode.This paper focuses only on the departure scheduling simulation mode.The simulation analysis is executed by running the prototype terminal departure scheduler within the fast-time simulation evaluation harness, as illustrated in Fig. 2. Key inputs such as surface taxi time and terminal transit time undergo perturbation by applying stochastic uncertainty.Once the scheduled OFF time for an aircraft is calculated, then the aircraft's actual OFF time is adjusted by a random variable.The application of error to the terminal transit time is called terminal transit error, whereas the application of this error to surface events is called surface error.This research varies both terminal transit error and surface error to assess the robustness of scheduler design to these variations.The evaluation harness developed for this work provided required components enabling fast time simulation; namely a component to provide input data to the scheduler, a feedback mechanism to model controller response to scheduler output and an error generation component to inject realistic operational error into the simulation.Figure 2 illustrates the terminal departure scheduler evaluation harness, which is briefly described in the following subsections.
+A. InputsThe inputs to the terminal departure simulation consist of flight data, constraints and decision trees for multiple airports in the simulation.The following subsections will briefly discuss the inputs required for this research.
+Flight Data Input FilesThe terminal departure simulation capability includes the ability to generate input files that match the demand and operational criteria specified.The result of this input generation process is an output file of flights that match the given criteria.Some of the choices available when creating simulation input files are the amount of desired departure fix demand per hour, the percentage of departure fix demand from each airport by departure runway, the percentage of flights subject to other traffic constraints (e.g.EDCT or CFRs) and other variables.This research modeled D10 airspace to evaluate terminal constraints, as depicted in Fig. 3.This diagram includes airports contained within the boundaries and the departure fixes located on the borders.The D10 TRACON is centered on DFW and extends outward approximately forty miles in all directions.It contains two major scheduled passenger service airports, DFW and Dallas Love Field (DAL), which are separated by approximately ten miles.Several busy general aviation airports, a regional cargo hub, and a Naval Air Station Joint Reserve Base contribute to the complexity of this TRACON environment.The sixteen departure fixes are arranged in groups of four called departure gates (not to be confused with airport parking gates), which depict their general location relative to the TRACON boundaries.For example, the north gate includes departure fixes LOWGN, BLECO, GRABE, and AKUNA.It is common for restrictions to be imposed on entire gates, without mention of the fixes, so it is important to understand which fixes belong to which gates.A year of operational data from 2013 was analyzed to determine average flight times and variation to each departure fix from each departure airport.
+Input Files Control VariablesPrototype  While simultaneous departure fix constraints are often applied in terminal departure operations, this research found it useful to focus primarily on the effect to system performance with a single departure fix constraint.The scenario used most frequently in this paper was a 10 miles in trail (MIT) constraint over departure fix SOLDO on April 10, 2013.This day was selected primarily because of availability of firsthand observations of operations from D10 TRACON and detailed output data to further analyze the traffic scenario.
+ConstraintsTerminal departure constraint inputs possess substantial flexibility.This flexibility is demonstrated in the use of routing constraints, flow control constraint (e.g.MIT) and creative combinations of both.This section discusses terminal departure constraints, as listed in Table 2, and their handling by the simulation framework.The terminal departure constraint is distinct from Center tactical departure metering constraints (i.e.CFRs).However, the terminal constraint and CFR share many properties.For example, they are both local tactical constraints which require a precise departure time and approval prior to releasing the departure.A key distinction between terminal constraints and CFRs is the domain that implements the restriction.While an underlying reason for a terminal constraint may originate from the Center environment, the entire process is implemented in terminal airspace by terminal personnel.Another distinction between terminal constraints and CFRs is the process used to regulate the departure.Unlike CFRs, terminal constraints typically do not come with a specific departure time window but rather only the expected sequence of departing flights.For this reason, the process is often called departure sequencing by terminal personnel.The departure sequencing process is used instead of specific departure times due to extremely high levels of uncertainty that are present in the terminal departure environment.Table 2 lists commonly used terminal departure constraints.The restrictions modeled in this research are complete departure fix combine, MIT and a speed constraint.The gates referred to in these restrictions are groups of departure fixes.For example using the illustration in Fig. 3, departure fixes NOBLY, TRISS, SOLDO and CLARE all belong to the East departure gate.The terminal departure simulation input files control the type of constraint the simulation provides to the scheduler, the time at which it is injected into the system and the duration of the constraint.
+Decision TreesSimulation input files are used to control the size and distribution of surface and terminal transit error.These input files are called decision trees because they provide branch like options that allow methodical selection of a property based upon one or more decision variables.Surface taxi times and terminal departure transit times are supplied by decision trees.The primary purpose of surface taxi times in simulation is to allow analysis of delay distribution on the airport surface.Terminal transit times are used to simulate realistic flight times from each departure airport to each departure fix.These decision tree distributions are specified by a mean value, to which a Gaussian error distribution is applied.The distributions used in simulation were determined by analysis of operational data, information learned in first-hand observations of operational events and prior research. 1he OFF time and flight time error associated with the terminal departures in the simulation is also governed by decision trees.Error is expressed as a stochastic value with the specific mean and standard deviation.Error values used in this research were obtained through a combination of operational data analysis, direct operational observations and results from prior research. 1Using the decision trees, error can be applied to flights at several locations in the departure process, including the pushback time, surface taxi, departure queue and terminal transit.For this research error was applied to surface taxi, departure queue (as controlled OFF time error) and terminal transit.The distribution of error is assumed to be Gaussian, which is consistent with prior research on tactical departures. 1
+Simulation VariablesThe terminal departure simulation framework uses an input file to control simulation variables.This section discusses frequently modified parameters.The simulation requires a scenario start and stop date and time.While the simulation will be the duration the user specifies, it is typically best to ensure that all flights in the input file have sufficient time to cross the departure fix.While it may be desirable to evaluate a fixed traffic demand period, terminal departure pushes tend to have the highest error near the end.Thus, if comparing two scenarios to one another, the average departure delay values may be misleading if the entire flight demand has not been resolved.Error can be applied on the surface in several areas rather than all in the departure queue.The purpose of this capability is to provide a more realistic model of where delay would occur on the airport surface when terminal delays are encountered.Booleans exist in simulation input to allow error to be applied at pushback, taxi and to the controlled OFF time.If these Boolean values are set to true, the decision tree associated with the surface event is used for the distribution.Error can also be applied to the airborne flight time.If this Boolean is set to true then the flight time error decision tree is used to apply the specified error distribution to departure transit time of flights.A simulation input variable specifies the freeze type.The options are either sequence freeze or schedule freeze.A variable exists to specify the freeze horizon, which is the number of seconds prior to OFF that a flight is frozen by the scheduler.Sequence freeze ensures that this departure scheduler maintains the order of departing flights once the flight reaches its freeze horizon.Schedule freeze requires a flight to meet its departure time within the specified parameters in addition to sequence freeze requirements of maintaining departure order.For more information on sequence and schedule freeze capability, see the freeze section of the results.For schedule freeze a variable must be set that specifies the number of seconds past the controlled OFF time a flight is automatically rescheduled.Lastly, an airport switch penalty variable is used to model the current day behavior when departure control alternates from one airport to another (i.e.DFW, Love Field, Addison, etc.).The switch penalty is only used in current day (baseline) modeling as this delay is believed to be eliminated with reduced coordination uncertainty provided by automation and graphical displays in the Towers.
+B. External Feedback Provided to the Scheduler by the Simulation FrameworkThe simulation framework provides feedback to the terminal departure scheduler in response to its guidance.This response seeks to model the response expected in the operational environment.
+Feedback mechanismThe terminal departure simulation capability ensures minimal separation is enabled at the departure runway and departure fix.The purpose of this logic is to create a realistic environment in which to evaluate the terminal departure scheduler.Minimal departure runway separations for large, well-equipped airports rely upon the separation logic from the SDSS.Each smaller airport surface scheduler has adapted separation that is used for minimal separation.For this research the runway separations for all airports was the same as used by the SDSS system that runs at DFW airport.For departure fix separation, the routing or miles in trail constraint is enforced at the departure fix.For flights that would otherwise have insufficient separation at the departure fix boundary, the simulation places the flight in a controller intervention status.When a flight is placed in controller intervention status it is allowed to achieve the required amount of separation at the departure fix boundary.The amount of time that a flight is in controller intervention status is recorded.The purpose of recording the amount of time a flight is in controller intervention status in the simulation is to allow a method for evaluating controller workload associated with terminal departure scheduling.
+Error generationThe error generation component applies a stochastic error to the time component in question as specified in the appropriate decision tree.Error can be applied to flight pushback time, taxi time, controlled OFF time assignment and terminal transit time.The size and distribution of error is controlled by the decision tree files as previously discussed.
+C. OutputsThe terminal departure simulation environment produces output to both a flat file and a database.Database output is especially useful when executing a large number of Monte Carlo simulations like that performed in this research.The simulation output include details on each flight, simulation run, the error exerted on the flight, expected transit times, actual transit times, delay incurred on surface and airborne and the amount of controller intervention.
+IV. ResultsThe results outlined in this section give insights into the effectiveness of the scheduling algorithm when exposed to a range of air traffic constraints, departure time uncertainty and flight time uncertainty.The primary metrics evaluated are size of departure delay, throughput and controller intervention.
+A. Scheduler Performance under Varying Levels of OFF Time CompliancePrior research on tactical departures 1 indicates that substantial improvements to OFF time compliance can be achieved with surface automation and reduced communication uncertainty.Improved OFF time compliance is expected for terminal departures for the same reasons.In addition, OFF time improvement is expected in the terminal environment due to greater situational awareness of upcoming flight departure times which is not possible in operations today due to the opaqueness of the schedule.This section analyzes terminal departure scheduler performance when exposed to a range of OFF time error.
+SetupTo analyze the effect of OFF time compliance error on system performance, all experimental variables other than the OFF time error were held constant.The constraint used in the system was a 10 MIT restriction over departure fix SOLDO with an expected crossing speed of 350 knots.The demand was 30 flights per hour for a duration of 80 minutes.This created a total of 40 flights that were evaluated in 500 Monte Carlo runs.For each Monte Carlo run, the OFF time error was varied according to the specified Gaussian distribution.The OFF time error was varied from levels expected when terminal departure scheduling automation is available, to estimated levels with no automation, to one standard deviation greater than no automation levels.The OFF time compliance used for terminal automation simulation was a mean of -9 seconds and a standard deviation of 60 seconds for DFW flights.This compliance matches that seen in prior research for DFW when using surface automation. 1 The OFF time compliance used for all other D10 airports was slightly higher given the lack of surface automation available at these airports.For these airports, a mean of 0 seconds and standard deviation of 90 seconds was used.The highest OFF time compliance error was obtained by adjusting the standard deviation of the baseline estimate by a factor of two.
+ResultsAs expected, the results indicate that better OFF time compliance leads to better terminal departure performance.Average ground delay per flight was 11.6 minutes, 16.3 minutes and 22.3 minutes for the automation, no automation and high error cases respectively.The 4.7 minute change in average delay between the automation and no automation case suggest that OFF time compliance is a significant factor in achieving reduced delay.The change in average delay between the no automation and high error case suggests that average delay will continue to increase as OFF time error increases.Figure 4 plots the distribution of ground delay associated with each OFF time error level as a function of time.In this diagram the delay for each error scenario was grouped in 10minute increments and plotted as a function of minutes into the departure push.The distribution is plotted as a line instead of histogram to aid in comparing the distribution amongst the error cases.The distribution of delay over time between the error cases is similar, however, there are two key differences.First, as the OFF time error increases the average amount of delay assigned to flights increase.This is visible in the separation between the lines which builds over time due to a slightly higher slope on higher error cases.Secondly, as the OFF time error increases the duration of the departure push is extended.All plots end with zero delay when the complete demand of 40 flights has been resolved by crossing the departure fix.In this case the duration of the departure push was 37 minutes longer in the high error case than with the automation.Differences in the duration of the departure push for each OFF time compliance error case indicate a difference in departure throughput.To analyze the effect on throughput when varying OFF time error, the maximum departure rate metric was used.The maximum departure rate measures the highest number of flights that crossed the departure fix in a 10-minute window.This throughput measure is robust to changes in demand that can occur throughout the push, as well as push startup and shutdown variations.Figure 5 illustrates the departure rate of each error scenario over time.During the first 20 minutes all three scenarios show increasing departure rate as additional flights are injected into the simulation over time.Once the available capacity is saturated, the throughput difference between the automation levels of OFF time error and other cases becomes more apparent.The highest difference in throughout is 5.4 flights per hour, which is seen in the 80-89 minute window between the automation and high error cases.The automation error case ends first, followed by the no automation error case and last is the highest error case.This suggests that OFF time error has a direct effect on departure throughput.These findings underscore the benefits to terminal departure delay reduction and increased throughput that can be provided by greater OFF time precision from surface automation.
+B. Scheduler Performance under Varying Levels of Flight Time ErrorScheduling a departure in the terminal environment in the NAS today requires two mental calculations by controllers, an OFF time estimate and a flight time estimate.In some cases the controller may not attempt to estimate the flight time but rather wait for the flight to clear a pre-determined airborne location prior to departing a trailing flight.Observation of the terminal scheduling process indicates that different methods may be employed by different personnel.Flight time estimates are important for future automation as well.The terminal scheduler de-conflicts a departure with other flights on the runway and the departure fix.Thus, if the flight time is inaccurate the model upon which flights are being assigned delay can be incorrect.This experiment analyzes the sensitivity of departure scheduler performance to flight time error.
+SetupAll experimental variables other than the flight time error were held constant.The April 10 th , 2013 scenario mentioned in the previous section was utilized, however, for all scenarios the OFF time error remained at expected levels with future automation.The flight time error was varied from a mean error of 0 seconds to a mean error of 25 seconds.The standard deviation of flight time was varied from 15, 30, 60 and 240 seconds.The flight time error level used to estimate
+Auto No AutoNo Auto high error future automation was a mean of 25 seconds with a standard deviation of 30 seconds for DFW.This flight time error was chosen because it matches that seen in prior research. 1In the automation scenario, flight time error from smaller airports was slightly higher due to variance from less frequent demand from departure airports to departure fixes in non-standard terminal constraint situations.Small airport flight time error for future automation is expected to be a mean of 35 seconds and standard deviation of 40 seconds.
+ResultsConsistent with intuition, the simulation results indicate that lower flight time error leads to less controller intervention.As indicated in Table 4, in the lowest flight time error case the percentage of flights that are estimated to require controller intervention of one minute or greater are 22%.As previously discussed a flight is considered in controller intervention status when the simulation evaluation harness determines that inadequate separation will exists at the departure fix boundary.The flight remains in controller intervention status until enough simulation time transpired to achieve the required amount of separation.Controller intervention percentage grows a modest 1% in the automation scenario but increases substantially to 37% of flights in the largest error scenario.In addition to increased need for controller intervention, the duration of the intervention is also longer.In the low flight time error case average controller intervention is 115 seconds, while in the largest flight time error scenario average controller intervention is estimated to be 135 seconds.Perhaps a less obvious effect of increased flight time error is the transitive effect which may lead to ground delay.Given short flight times in the terminal area, unexpected airborne delay can ripple back to departing airports that are scheduling into this environment.As indicated in Table 3, the average ground delay changes by 3.6 minutes from the lowest flight time error scenario to the highest.The maximum effective throughput listed in Table 3 is defined as the highest percentage of the hourly rate achieved during a 10 minute window.This throughput metric is useful to analyze max throughput despite demand variations that can occur due to clumping of demand at the beginning or end of the push.The highest throughput was demonstrated by the low flight time error case in which 88% of its given demand was resolved.The lowest throughput was demonstrated by the highest flight time error case, with a maximum effective throughput of 79%.Given 500 Monte Carlo simulation runs were performed for each scenario, it was possible to compare the shortest and longest length of a departure push.This metric can be used to estimate the best and worst case scenarios from a system performance standpoint.The difference between the longest and shortest departure push was 135 minutes versus 115 minutes for the low and highest flight time error cases.While the variations to departure performance are not as significant as those demonstrated by OFF time compliance, they do suggest a strong correlation between improved flight time prediction and better system performance.It is worth noting that the flight time results discuss in this section all use the east gate.Given D10's predominant use of south flow configuration, terminal transit times departing the north gate are generally longer.Longer terminal transit times often increase flight time variance.Additional analysis would be required to determine the effect airspace geometry has on the metrics measured in this section.
+C. Analyzing the Switching Penalty in Baseline OperationsA less obvious benefit from terminal departure scheduling automation is that associated with loss of throughput due to coordination timing between departing facilities.This time is referred to in this research as switching time, which has an associated switching penalty.Based upon observations of current day terminal departure scheduling, the primary reason that a switching penalty exists today is the inherent opaqueness in the schedule.During terminal constraints, key personnel are often so busy with tasks that they are not able to coordinate with all the required parties in a manner that allows adequate lead time to prepare the next flight in sequence for departure.Inadequate lead time can result in unutilized departure demand.This phenomenon is known as a switching penalty.However, with automation available to all required parties and an indication of the forthcoming flight's departure time, the switching penalty is expected to be removed.This experiment assesses the impact of a switching penalty in baseline operations.Later in this paper, the switching penalty is combined with expected improvements to OFF time and flight time error to estimate benefit of automation over the current day baseline.
+SetupTo analyze the effect of switching time penalty on system performance, all experimental variables other than the switching time penalty variable were held constant.The switching time penalty variable is specified in the simulation as the number of seconds of delay incurred when switching from one departure airport to another.In terminal departure operations today this delay is generally experienced over time while waiting for the tower that just received authorization to depart flights communicate with the pilots and prepares them for departure.While this waiting is occurring, delay at other airports continues to build.The switching penalty is only imposed if the flight is ready to depart.For instance, if a 30 second switching penalty is enforced but a flight is 60 seconds late due to OFF time error compliance, then no penalty is enforced.However if the flight was ready to depart, a 30 second delay would be added to that flight and any other flights that were immediately trailing the flight.When switching from one departure airport to another, the departure controller must first recognize the readiness for this activity by observing the departing flight from the preceding airport.Then the terminal departure controller communicates with the airport departure controller to allow the flight to depart.Finally, the airport departure controller then communicates with the pilot to clear the flight for departure.Based upon estimates from prior research 1 which analyzed response times to controller commands, the entire switching process is estimated to take at a minimum 30 seconds.Thus, the switching penalty values analyzed were 0 seconds, 30 seconds, 60 seconds and 120 seconds.The OFF time error was held constant at future automation levels to minimize the effect of multiple error sources on the switch penalty analysis.
+ResultsAs expected, the results indicate that as the switch penalty increases, so does the average delay and push duration.No switch penalty resulted in a 13 minute delay average, while a 120 second switch penalty resulted in a 15.9 minute delay average.Increased switch penalty also has an effect on throughput.The longest duration departure push occurred in the highest switch penalty case, which was 14 minutes longer than no switch penalty scenario.These results indicate the switching time period encountered in current day terminal operations has a substantial impact on flight delay and throughput.Increased visibility into the departure schedule from automation is expected to reduce or eliminate this shortfall.
+D. Scheduler Performance with Varying Miles in Trail ConstraintsTerminal departure simulation was used to investigate the effect of increasing MIT restrictions on terminal departure performance.This section gives insight into how terminal departure delay grows as MIT increases and what the expected benefit of terminal departure automation is as a function of MIT constraint.
+SetupA MIT constraint over a single departure fix was used for this experiment.The size of the MIT constraint was varied from 10 to 30 miles in trail in 5 mile increments.The demand for all scenarios was 30 flights per hour for a duration of 80 minutes.Two scenarios were analyzed at varying MIT, one representing a current day without automation and the other terminal automation.The no automation scenario used 30 seconds switching penalty, a mean OFF time error of 15 seconds and a standard deviation of 115 seconds.The automation scenario had no switching penalty and used a mean OFF time error of -9 seconds with standard deviation of 60 seconds for DFW, and a mean of 0 seconds with standard deviation of 90 seconds for other airports.
+ResultsResults indicate that MIT has a strong relationship to average ground delay and throughput.As illustrated in Fig. 6, as MIT increases, so does the average ground delay assigned.In all cases evaluated the automation scenario outperformed the no automation case.In terms of percentage improvement over no automation, the greatest benefit is seen at lower MIT values.Specifically, at 10 MIT the automation scenario demonstrates a 35.2% reduction in average departure delay over the no automation case.There are believed to be two reasons for the decreased percentage of delay as a function of increased MIT.First, the portion of delay that is saved due to removal of the airport switching penalty stays the same across all MIT values.As a percentage of the total ground delay this portion is higher in the 10 MIT case than in larger MIT cases.Secondly, as MIT values continue to grow, the demand reaches a point of saturation such that flights with high OFF time compliance error that would not have made their departure time now do.Thus, the benefit of improved OFF time compliance error is proportionally smaller as the delays grow larger.To analyze the effect on throughput when varying MIT, the maximum departure rate metric was used.The maximum departure rate measures the highest number of flights that crossed the departure fix in a 10 minute window.This throughput measure is robust to changes in demand that can occur throughout the push, as well as push startup and shutdown variations.Figure 7 illustrates the departure throughput by MIT.As expected, the departure rate decreases as MIT increases.The largest gains in throughput between the no automation and automation case are seen at the lowest MIT.At 10 MIT a 3.7 flights per hour departure rate improvement is observed by the automation scenario over the no automation scenario.For the traffic levels analyzed in this experiment, that is approximately 16.6% improvement in departure throughput.
+E. Analysis of Schedule versus Sequence Freeze PerformanceA shortfall of the current day manual terminal departure scheduling process is the lack of visibility into the scheduled departure time.Given the short notice flights may receive prior to departure, efficiency is lost due to time required for staging the departure for takeoff.Future terminal departure automation is expected to provide the departure information with sufficient lead time for controllers to prepare the flight for departure.However, when communicating this information it is important that as few changes to the communicated departure time occur as possible.Changes to controlled departure time can increase controller workload and decrease efficiency.Given the expectation that the time will remain unchanged, the process of communicating a controlled time to operational personnel is often referred to as 'freezing' the flight. 18This section discusses the evaluation of two distinct terminal departure freeze methodologies.Observations of D10 terminal departure scheduling indicate that operational personnel adhere to a specified departure sequence, however, they are not required to adhere to a specified departure time.The process of specifying and communicating the order of departing flights that will not change barring a re-plan is called a sequence freeze. 18The process of specifying both the order and departure time for departing flights is referred to as a schedule freeze.In this prototype terminal departure scheduler the user must specify either a sequence or schedule freeze capability.If schedule freeze in enabled and a flight misses its assigned departure time by a specified number of seconds, the terminal scheduler will reschedule the flight.When a flight is rescheduled it will be assigned the next available time which does not impact other frozen flight times.If sequence freeze is enabled, the specified flight order is maintained without regard to departure compliance.This prototype terminal departure scheduler requires the user to specify a freeze horizon window for use in the scheduling process.The freeze horizon window is the time prior to departing the airport that a flight becomes frozen.The freeze horizon must be large enough to allow adequate time for controllers to prepare a flight for departure.However, the challenge with extending this time period too far prior to departure is that uncertainty can be frozen into the schedule which might otherwise be resolved more efficiently later using more accurate departure time information.For the analysis discussed in this section, the freeze horizon value was held constant at 180 seconds.A goal of this analysis is to analyze the tradeoffs of sequence freeze versus schedule freeze capability in terminal departure scheduling.Given substantial uncertainty that exists in terminal departure operations, the overall stability and efficiency of the departure schedule may be compromised if a large percentage of flights have to be re-planned due to missing their scheduled departure time.The balance between greater visibility into the schedule and departure planning stability is analyzed with simulation using realistic assumptions for expected departure time compliance.
+SetupThe variables modified in this experiment were OFF time error and rescheduling time window.OFF time error was varied between the three levels described in earlier sections.The resecheduling time window specifies the number of seconds past the expected OFF time that a flight must be rescheduled.The rescheduling time window value was varied from 30 seconds through 360 seconds.
+ResultsFigure 8 illustrates the effect of OFF time error on the schedule freeze capability.These results were obtained by using a 60 second reschedule time window for the schedule freeze capability.The OFF time error was varied from levels expected with future terminal departure automation, to estimated levels in today's operations without automation, to twice the standard deviation of the no automation estimate.As seen with previous results, average delay incurred by flights generally grows as the departure push continues.Varying the OFF time error has a visible effect on the distribution of delay over time.Consistent with previous results, lower OFF time error yields lower delay.A difference from prior results is that in this case a number of flights were required to be rescheduled due to missing their assigned OFF time by greater than 60 seconds.Out of a 40 flight scenario, the number of flights that required rescheduling were 5, 14 and 16 for the automation, no automation and high error cases respectively.Thus, as OFF time compliance error grows, the number of flights that missed the required departure window also grew.To compare the performance of the sequence freeze capability against the schedule freeze capability in a fair manner, OFF time error was held constant.Since this capability is targeted at future automation, the expected OFF time error associated with that environment were used.Figure 9 illustrates the delay distribution over time of sequence freeze capability against schedule freeze capability at varying rescheduling time windows.The performance of the schedule freeze capability improves as the rescheduling window is raised.The worst performance of all freeze scenarios is the 30 second rescheduling time window.The reason for this is the number of flights that require rescheduling are higher given the low threshold for OFF time compliance performance.The process of rescheduling a flight creates more demand that must be resolved.This in turn takes more time which leads to higher delay.As the rescheduling time window gets larger fewer flights are required to be rescheduled, resulting in improved performance of the schedule freeze capability.As such, the best performance of the schedule freeze capability is seen when the rescheduled window is at 360 seconds.Even at 360 seconds the schedule freeze time window's average delay is slightly larger than that of the sequence freeze average delay.Thus, at the error levels expected in future terminal departure automation, the sequence freeze capability performs better than the schedule freeze capability.The results shown in Fig. 9 are based upon the expected levels of OFF time error with future automation.As Fig. 8 demonstrates, the size of the OFF time error has an effect on the delay distribution of scheduled freeze capability.The assumption on OFF time error made in this research is lesser-equipped airports will have a 50% larger standard deviation of OFF time error than well-equipped airports.If the OFF time error is more disproportionate than assumed given higher OFF time error from lesser-equipped airports, then the schedule freeze may become a more attractive option to ensure lesser-equipped airport delay is not propagated to well-equipped airports.Additional research is needed to study the sensitivity of schedule freeze parameters to varying and disproportionate OFF time error levels.
+V. DiscussionThe simulation results described in previous sections illustrate the cumulative nature of terminal departure delay.Terminal departure delay builds upon itself until the demand is resolved or the constraint is removed.This finding underscores the importance of reducing the duration of the terminal departure restriction to the greatest degree possible.To support this objective the terminal departure solution should aim for simplicity to reduce the amount of time required to set up the constraint and communicate it to all required parties.Equally, if not more important, is the need to ensure that a terminal departure restriction does not remain in place unnecessarily.This suggests the tactical departure scheduling capability would benefit from close integration with future automation geared toward automatic detection of local flow imbalances like the Integrated Departure Route Planner (IDRP). 11esults indicate a direct relationship between OFF time compliance and departure scheduler performance.Improved compliance demonstrates a notable improvement to delay and throughput.High OFF time compliance error may also lead to increased controller workload and airborne fuel utilization.This underscores the importance of leveraging newer technologies like that demonstrated in prior tactical departure scheduling research 1 as well as focused efforts to improve departure compliance at lesser equipped airports.Results also indicate a direct relationship between terminal transit time prediction and departure scheduler performance.In addition to creating higher controller workload and greater fuel utilization, flight time error can result in delay being propagated back to the airport surface.The terminal departure scheduling solution should seek to build upon improvements to predictive accuracy of terminal transit time made in prior work. 1 The departure scheduler used in this evaluation demonstrated robustness to terminal transit prediction error up to twice the levels expected in today's operations.However, at prediction error levels 4 times the variation expected, substantial controller workload and additional ground delay is experienced.These error levels may occur during inclement weather scenarios in which the nominal departure route is blocked.More research is needed to further assess the sensitivity of performance to flight time error and identify approaches to resolve this challenge.Schedule freeze capability allows greater transparency into flight's departure plan than does sequence freeze capability.Additionally, schedule freeze can help ensure that uncertainty at one airport does not impact departing flights at a separate airport.However, implementing a schedule freeze requires a rescheduling methodology for those flights which do not make their controlled departure time.The results of this analysis indicate that, at the OFF time error levels expected with terminal departure automation, the additional demand caused by schedule freeze rescheduling creates higher average delays and longer pushes than sequence freeze.Additional research may be warranted to more fully evaluate freeze options in the terminal departure environment.
+VI. ConclusionsFast time simulation modeling of terminal departure traffic was used to assess performance of a new terminal departure scheduler.The prototype terminal departure scheduler was exposed to a range of air traffic constraints, departure time uncertainty and flight time uncertainty to better understand its sensitivity to these variables.Simulation was used to assess the tradeoffs of sequence and schedule freeze methodologies in the terminal departure environment.Both freeze capabilities were evaluated under a range of possible OFF time errors.Sequence freeze capability demonstrated lower average delay than schedule freeze capability for expected levels of OFF time compliance in future automation.Simulation results of D10 airspace indicate delay reductions of 35 percent over current-day scheduling practices are possible for commonly used terminal departures constraints, as well as an increased departure throughput of 17 percent.This benefit is derived via a combination of improved OFF time compliance, reduced flight time error and removal of the airport switching penalty associated with lack of automation in terminal departure operations today.Results indicate modest decreases to controller workload and airborne fuel utilization are possible.The results of this study were used to establish design considerations for a terminal departure scheduler which will undergo evaluation at NASA's North Texas Research station.The results are also used to inform the concept of operations (ConOps) document being developed on the future terminal departure scheduling process.Figure 1 .1Figure 1.Terminal Departure Scheduler Processing Flow.
+Figure 2 .2Figure 2.An evaluation harness was developed to assess a prototype terminal departure scheduler.
+Figure 3 .3Figure 3. D10 departure airspace was modeled to evaluate terminal constraints.
+Figure 4 .4Figure 4. Lower OFF time error leads to lower average delay.
+Figure 5 .5Figure 5. Departure throughput varied substantially between the lower and higher error scenarios.
+Figure 6 .Figure 7 .67Figure 6.Average Ground Delay Varies by MIT.
+Figure 8 .8Figure 8. Delay distribution of 60 second reschedule time window at various OFF time error levels.
+Figure 9 .9Figure 9.Comparison of delay distribution of sequence freeze and scheduled freeze capability at varying rescheduling time windows.
+Table 1 . Sequencing categories determine the order a flight is scheduled.1Downloaded by NASA AMES RESEARCH CENTER on June 20, 2014 | http://arc.aiaa.org| DOI: 10.2514/6.2014-2020Sequencing Category(in priority order fromAircraftgreatest to least)DescriptionLocationSorted ByThe flight has crossed the departureCenterActual departureCrossed Departure Fixfix.Airbornefix crossing timeUndelayedTerminally ControlledFlights that have a terminal constraintTerminalestimatedAirborneand are airborne.Airbornedeparture fix timeUndelayedFlights that have no terminal constraintTerminalestimatedUncontrolled Airborneand are airborne.Airbornedeparture fix timeTerminally controlled flights have aTerminallyTerminally Controlledfrozen OFF time. This category is theSurfacecontrolled frozenFrozenfocus of this research.ActiveOFF timeFlights that are surface active and haveSurfaceCall For ReleaseCall For Releasea Call for release TMI.ActivetimeFlights that are surface active and haveSurfaceStrategic TMIan EDCT.ActiveEDCT timeTerminally ControlledTerminally controlled flights that areSurfaceEstimatedUnfrozen Surface Activesurface active but not yet frozen.Activeundelayed OFFFlights that are surface active with noSurfaceEstimatedSurface ActiveTMI constaint.Activeundelayed OFFTerminally ControlledTerminally controlled flights that areSurfaceEstimatedSurface Inactivenot yet surface active.Inactiveundelayed OFFFlights that have no terminal constraintSurfaceEstimatedSurface Inactiveand are not surface active.Inactiveundelayed OFF
+Table 2 . Common terminal departure constraints.2Constraint NameConstraintDescriptionTypeComplete DepartureRouteAll traffic that was assigned to the original fix isFix Combinemoved to one or more alternate departure fixes.Limited DepartureRouteA select set of flights bound to a departure fix isFix Combinemoved to one or more alternate departure fixes.Stream basedRouteA departure fix becomes the only location terminalDeparture fixflights bound to the destination in question mayCombinedepart the terminal area.Gate SwapRouteChanges the gate a departure fix traffic is bound toone or more alternate gate(s). This adds therequirement for departing flights to file a new flightplan.Miles in TrailFlowRequires MIT separation enforced at the departurefix. This flow constraint is often enforced with arouting constraint.Speed ConstraintFlowRequires departures to meet a speed restriction,typically until reaching the departure fix.Downloaded by NASA AMES RESEARCH CENTER on June 20, 2014 | http://arc.aiaa.org| DOI: 10.2514/6.2014-2020
+Table 4 . Results of Flight Time Error Variation on Departure Scheduler Performance.4Standard% RequiredControllerAverageMaximum EffectiveDurationMeanDeviationControllerInterventionGroundThroughput per hourLongestScenario NameError (s)Error (s)InterventionDuration (s)Delay (m)(% total demand)Push (m)Low Flight Time Error0152211512.288115Automation (expected)25302311713.884117Automation w/2sigma25602911913.984119Automation w/4sigma252403713515.879135
+			Downloaded by NASA AMES RESEARCH CENTER on June 20, 2014 | http://arc.aiaa.org| DOI: 10.2514/6.2014-2020
+		
+		
+
+			
+AcknowledgementsThe authors would like to acknowledge the essential support provided by FAA personnel at Dallas/Fort Worth terminal Radar Approach Control (TRACON) facility, DFW ATCT and DAL ATCT.Finally, we wish to thank our colleagues at NASA/FAA North Texas Research Station (NTX) and NASA Ames whose support was critical to the success of terminal departure research objectives.
+			
+
+			
+
+				
+
+
+	
+		
+			SAEngelland
+		
+		
+			ACapps
+		
+		
+			KDay
+		
+		
+			MKistler
+		
+		
+			FGaither
+		
+		
+			GJuro
+		
+		NASA/TM-2013-216533
+		Precision Departure Release Capability (PDRC) Final Report
+		
+			June 2013
+		
+	
+	Engelland, S.A., Capps, A., Day, K., Kistler, M., Gaither, F., and Juro, G., "Precision Departure Release Capability (PDRC) Final Report," NASA/TM-2013-216533, June 2013.
+
+
+
+	
+		Precision Departure Release Capability (PDRC) Concept of Operations
+		
+			SAEngelland
+		
+		
+			ACapps
+		
+		
+			KDay
+		
+		NASA/TM-2013-216534
+		
+			June 2013
+		
+	
+	Engelland, S.A., Capps, A., and Day, K., "Precision Departure Release Capability (PDRC) Concept of Operations," NASA/TM-2013-216534, June 2013.
+
+
+
+	
+		Trajectory-Based Takeoff Time Predictions Applied to Tactical Departure Scheduling: Concept Description, System Design, and Initial Observations
+		
+			ShawnEngelland
+		
+		
+			RichardCapps
+		
+		10.2514/6.2011-6875
+		NASA/TM-2013-216531
+	
+	
+		11th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference
+		
+			American Institute of Aeronautics and Astronautics
+			June 2013
+		
+	
+	Engelland, S.A., Capps, A., Day, K., Robinson, C., and Null, J.R., "Precision Departure Release Capability (PDRC) Technology Description," NASA/TM-2013-216531, June 2013.
+
+
+
+	
+		NextGen Mid-Term Concept of Operations for the National Airspace System
+		
+			Faa
+		
+		
+			September 2010
+		
+	
+	FAA, "NextGen Mid-Term Concept of Operations for the National Airspace System," version 2.1, September 2010.
+
+
+
+	
+		A functional analysis of integrated arrival, departure and surface (IADS) operations in NextGen
+		
+			MSimons
+		
+		10.1109/dasc.2012.6382982
+		MITRE CAASD MTR110240R1
+	
+	
+		2012 IEEE/AIAA 31st Digital Avionics Systems Conference (DASC)
+		
+			IEEE
+			January 2012
+		
+	
+	MITRE CAASD MTR110240R1, "A Concept for Integrated Arrival, Departure, and Surface (IADS) Operations for the Mid-Term", January 2012.
+
+
+
+	
+		Design and Evaluation of the terminal Area Precision Scheduling and Spacing System
+		
+			HNSwenson
+		
+		
+			JThipphavong
+		
+		
+			ASadovsky
+		
+		
+			LChen
+		
+		
+			CSullivan
+		
+		
+			LMartin
+		
+		
+			2011
+			Berlin, Germany
+		
+		
+			Ninth USA/Europe Air Traffic Management Research and Development Seminar
+		
+	
+	Swenson, H. N., Thipphavong, J., Sadovsky, A., Chen, L., Sullivan, C., & Martin, L. (2011). "Design and Evaluation of the terminal Area Precision Scheduling and Spacing System." Ninth USA/Europe Air Traffic Management Research and Development Seminar. Berlin, Germany.
+
+
+
+	
+		Effects of scheduling and spacing tools on controllers' performance and perceptions of their workload
+		
+			LynneMartin
+		
+		
+			HarrySwenson
+		
+		
+			AlexanderSadovsky
+		
+		
+			JaneThipphavong
+		
+		
+			LiangChen
+		
+		
+			AnthonyYSeo
+		
+		10.1109/dasc.2011.6096106
+	
+	
+		2011 IEEE/AIAA 30th Digital Avionics Systems Conference
+		Seattle
+		
+			IEEE
+			2011
+		
+	
+	30 th DASC
+	Martin, L., Swenson, H., Thipphavong, J.,Sadovsky, A., Chen, L., & Seo, Y. (2011). "Effects of Scheduling and Spacing Tools on Controllers' Perceptions of the Their Load and Performance." Digital Avionics Systems Conference (30 th DASC). Seattle.
+
+
+
+	
+		Development of Conflict Free, Unrestricted Climbs for a terminal Area Departure Tool
+		
+			JThipphavong
+		
+		
+			HSwenson
+		
+		
+			PLin
+		
+		
+			ASeo
+		
+		
+			LBagasol
+		
+		
+			YCJung
+		
+		
+			DIsaacson
+		
+		
+			YCIsaacson
+		
+		
+			DR
+		
+	
+	
+		Proc. of the AIAA's Aircraft Technology, Integration, and Operations (ATIO) Forum
+		of the AIAA's Aircraft Technology, Integration, and Operations (ATIO) ForumDenver, Colorado; Los Angeles, CA
+		
+			2011. 17-19 November 2003 10 Jung,. Oct 1-3, 2002
+		
+	
+	AIAA's 3 rd Annual Aviation Technology, Integration and Operations Conference
+	Thipphavong, J., Swenson, H., Lin, P., Seo, A., & Bagasol, L. (2011). "Efficiency Benefits Using the terminal Precision Scheduling and Spacing System." AIAA Aviation Technology Integration and Operations (ATIO). Virginia Beach 9 Jung, Y.C., Isaacson, D. "Development of Conflict Free, Unrestricted Climbs for a terminal Area Departure Tool", AIAA's 3 rd Annual Aviation Technology, Integration and Operations Conference, Denver, Colorado, 17-19 November 2003 10 Jung, Y.C., Isaacson, D.R., "Design Concept and Development Plan of the Expedite Departure Path," Proc. of the AIAA's Aircraft Technology, Integration, and Operations (ATIO) Forum, Oct 1-3, 2002, Los Angeles, CA.
+
+
+
+	
+		12 Federal Aviation Administration Air Traffic Organization Operations Planning Research & Technology Development Office Air Traffic System Concept Development, AJP-66
+		
+			JJWang
+		
+		
+			PaulChang
+		
+		
+			KDatta
+		
+	
+	
+		Prepared for NASA Ames Research Center, Contract No. NAS2-98074
+		
+			August 2003. September 2007
+		
+	
+	Life-Cycle Cost/Benefit Assessment of Expedite Departure Path (EDP). Integrated Arrival/Departure Control Service (Big Airspace) Concept Validation
+	Wang, J.J., Chang, Paul, and Datta, K, "Life-Cycle Cost/Benefit Assessment of Expedite Departure Path (EDP)," Prepared for NASA Ames Research Center, Contract No. NAS2-98074, August 2003. 12 Federal Aviation Administration Air Traffic Organization Operations Planning Research & Technology Development Office Air Traffic System Concept Development, AJP-66, "Integrated Arrival/Departure Control Service (Big Airspace) Concept Validation", September 2007.
+
+
+
+	
+		Benefits Analysis of a Departure Management Prototype for the New York Area
+		
+			JDearmon
+		
+		
+			NTaber
+		
+		
+			HBateman
+		
+		
+			LSong
+		
+		
+			TMasek
+		
+		
+			DGilani
+		
+		
+			2013
+			ATM
+		
+		
+			Tenth USA/Europe Air Traffic Management Research and Development Seminar
+		
+	
+	DeArmon, J., Taber, N., Bateman, H., Song, L., Masek, T., Gilani, D., "Benefits Analysis of a Departure Management Prototype for the New York Area", Tenth USA/Europe Air Traffic Management Research and Development Seminar (ATM2013).
+
+
+
+	
+		Generic Operational Concept for Pre-departure Runway Sequence Planning and Accurate Take-Off Performance
+		
+			ETuinstra
+		
+		
+			KHaschke
+		
+		
+			09 July 2009
+		
+	
+	Tuinstra, E., Haschke, K., "Generic Operational Concept for Pre-departure Runway Sequence Planning and Accurate Take- Off Performance", 09 July 2009.
+
+
+
+	
+		Characterization of Nationwide TRACON Departure Operations
+		
+			MatthewSKistler
+		
+		
+			AlanCapps
+		
+		
+			ShawnAEngelland
+		
+		10.2514/6.2014-2019
+	
+	
+		14th AIAA Aviation Technology, Integration, and Operations Conference
+		Atlanta, GA
+		
+			American Institute of Aeronautics and Astronautics
+			June 16-20, 2014
+		
+	
+	Kistler, M., Capps, A., Engelland, S., "Characterization of Nationwide TRACON Departure Operations", 14 th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference, Atlanta, GA, June 16-20, 2014.
+
+
+
+	
+		Measuring Uncertainty in Airspace Demand Predictions for Traffic Flow Management Applications
+		
+			CraigWanke
+		
+		
+			MichaelCallaham
+		
+		
+			DanielGreenbaum
+		
+		
+			AnthonyMasalonis
+		
+		10.2514/6.2003-5708
+	
+	
+		AIAA Guidance, Navigation, and Control Conference and Exhibit
+		Austin, Texas
+		
+			American Institute of Aeronautics and Astronautics
+			August 2003
+		
+	
+	Wanke, C. R., Callaham, M. B., Greenbaum, D. P., and Masalonis, A. J., "Measuring Uncer-tainty in Airspace Demand Predictions for Traffic Flow Management Applications," Proceedings of the AIAA Guidance, Navigation, and Control Con- ference, AIAA, Austin, Texas, August 2003.
+
+
+
+	
+		Analysis of En Route Sector Demand Error Sources
+		
+			JimmyKrozel
+		
+		
+			DanRosman
+		
+		
+			ShonGrabbe
+		
+		10.2514/6.2002-5016
+		AIAA-2002-5016
+	
+	
+		AIAA Guidance, Navigation, and Control Conference and Exhibit
+		Monterey, California
+		
+			American Institute of Aeronautics and Astronautics
+			August 5-8, 2002
+		
+	
+	Krozel, J., Rosman, D., Grabbe, S., "Analysis Of En Route Sector Demand Error Sources", AIAA-2002-5016, AIAA Guidance, Navigation, and Control Conference and Exhibit, Monterey, California, August 5-8, 2002.
+
+
+
+	
+		The dynamic planner: the sequencer, scheduler, and runway allocator for air traffic control automation
+		
+			GWong
+		
+		NASA/TM-2000-209586
+		
+			2000
+		
+	
+	Wong, G., "The dynamic planner: the sequencer, scheduler, and runway allocator for air traffic control automation", NASA/TM-2000-209586 (2000)
+
+
+
+	
+		A probabilistic modeling foundation for airport surface decision support tools
+		
+			ChrisBrinton
+		
+		
+			SteveAtkins
+		
+		10.1109/icnsurv.2009.5172837
+	
+	
+		2009 Integrated Communications, Navigation and Surveillance Conference
+		
+			IEEE
+			May 2009
+			
+		
+	
+	Brinton, C. and S. Atkins, "A Probabilistic Modeling Foundation for Airport Surface Decision Support Tools," 2009 ICNS Conference, 13-15 May 2009.
+
+
+
+	
+		Best practices for designing and implementing decision services, Part 1: An SOA approach to creating reusable decision services
+		
+			JBoyer
+		
+		
+			June 13, 2012
+		
+	
+	Boyer, J.,"Best practices for designing and implementing decision services, Part 1: An SOA approach to creating reusable decision services", June 13, 2012
+
+
+				
+			
+		
+	

file116.txt

@@ -0,0 +1,217 @@
+
+	
+	
+		
+INTRODUCTIONAs the number of operations at the nation's largest hub airports continues to grow, and as future operational concepts (e.g., NASA's Small Airport Transportation System or SATS concept) allow more aircraft to use smaller, non-hub airports, the need for environmental consideration in shaping the terminal * AIAA Member, Systems Analyst, Aviation R&D † Software Analyst, Aviation R&D ‡ Vice President, Aviation R&D § AIAA Member, Aerospace Engineer area operations is essential to minimize adverse impact to residents in the immediate vicinity of the airport (traditionally taken as those people inside the 65dB DNL contour).Reduction of noise exposure levels around the airport requires intelligent modification of the arrival and departure trajectories.Promising concepts being investigated in this area include low-noise approaches such as the Continuous Descent Approach (in which idle thrust levels are utilized over the majority of the descent) [1], Curved Approaches (which can potentially minimize over-flight of noise-sensitive areas) [2], and Precision Navigation Instrument Departures [2].Similar studies in the U.S. are being conducted at MIT [3].One downside of these approaches at this stage of their development is their tendency to require significantly higher levels of separation as well as improved navigational precision.Similar efficiencyrelated concerns have arisen in the context of regional airspace re-design efforts in the Chicago, Washington, and New York areas.Terminal area air traffic control (ATC) procedures contribute to noise issues (e.g., holding one flow below another flow), but their contributions have been difficult to quantify and even more difficult to remedy.This paper addresses both of these issues.First, we quantify the impact of postulated changes (such as consolidation of base-leg extensions and strict 24-hour adherence to noise abatement procedures) in terminal procedures at four U.S. airports [4].Second, we present an architecture for a Noise Avoidance Planner (NAP) that integrates with the Center-TRACON Automation System (CTAS) for both departures and arrivals to show how noise-awareness can be combined with efficient scheduling and sequencing.
+AIRSPACE DESIGN AND NOISE METRICSNoise exposure is inherently tied to the nature of the flight paths flown over a given population distribution.Thus, assessing the impact of flow American Institute of Aeronautics and Astronautics changes (geometry, traffic mix, runway use, etc.) requires a set of tools for first realizing such changes and then computing the impact of the changes on noise exposure.
+Airspace DesignThe primary tool used to emulate the effects of noise-aware decision support tools (DSTs) on traffic patterns was Metron Aviation's Airspace Design Tool (ADT).This tool provides the capability to import tracks from various sources (ETMS, ARTS, INM, NIRS, etc.), to display them in two and three dimensions, and to manipulate them graphically in 3space.Supporting tools allow manipulation of nonspatial characteristics, such as aircraft types, event times, and runway assignments.
+Noise Impact ModelingThe FAA's Noise Impact Routing System (NIRS) was used to compute noise exposure and to quantify the impact of postulated changes in operational procedures (assumed to be enabled by the availability of noiseaware DSTs).NIRS was chosen because it provides the same noise-calculation capability as the FAA's Integrated Noise Model (INM).It also includes several additional key features that are useful for the comparison of noise effects when aircraft have their flight characteristics influenced by the different DSTs.The two most important of these additional NIRS features are the ability to follow a specified flight profile and the ability to compare noise impacts in graphical and tabular formats for alternative cases (e.g., with or without DSTs).The input data for the noise calculations consists of detailed track and event data.Each track is represented by a sequence of points (latitude, longitude, altitude) defining a flight path into or out of a given airport on a given runway.Associated with each track is a set of events that represent the specific flights that are to operate on this track.Each event contains a description of the aircraft type, the engine/airframe type, the time of the event, and the number of events of this particular type.Noise computation involves tracing aircraft states along the spatial track and calculating the noise impact at each population location ("centroid") due to each flight.For this purpose, a flight path is generated for each track-event combination which models the thrust required for the aircraft to follow the prescribed trajectory.This process utilizes the altitude control codes specified in the track definition in order to "fly" the prescribed track in a fashion consistent with the modeled flight dynamics of each aircraft type.The principal noise metrics are the Sound Energy Level (SEL) and the Day-night Noise Level (DNL).SEL quantifies perceived acoustic energy across events of different intensities and durations, while DNL quantifies total noise exposure due to multiple events at different times of day.In this work, we quantify the effectiveness of postulated modifications to operational procedures in terms of the change in the size of the total population receiving 50dB DNL or greater exposure.This is consistent with increasing pressure on the aviation community to substantially reduce noise exposures below the traditional 65 dB DNL.
+OPERATIONAL IMPACT OF NOISE-AWAREDECISION SUPPORT TOOLS To quantify the potential benefits of adding noiseawareness to terminal area DSTs, we developed a number of techniques for manipulating data describing arrival and departure operations at four U.S. airports -Chicago O'Hare (ORD) and Midway (MDW), Boston Logan (BOS), and San Francisco (SFO) [4].These airports were chosen as a cross-section of the 20 airports originally investigated in Reference [5] on the basis of data availability and the existence of potential noise-mitigation opportunities.Each of these airports provided important insights into the operational and noise-mitigation issues that will be faced by development and deployment of noise-aware DSTs.The source of the operational data used for each airport in this study is given in Table 1.Data representing an annualized average day of traffic was used in each case.The approach used in developing changes to operational procedures was to visually study the arrival and departure traffic flows and identify potential changes in terminal area ATC procedures that might yield noise benefits (see Figure 1).This process included searching for interactions between arriving and departing flows and defining ways to mix these flows apart from the traditional method of procedural separation.The primary source for the noise mitigation opportunities identified in this study was a survey of 20 airports [5] based upon contacts with local noise offices (including Fly Quiet Programs where applicable) and controllers.The result of this assessment was an enumeration of (a) the current operational features which could be improved through design of suitable noise-abatement procedures, and (b) limitations in executing current noise abatement procedures.These limitations include navigational errors, controller instruction errors (e.g., speed/heading directives), and the fact that the majority of such procedures are generally only carried out under low-demand situations (due to increase separation requirements, etc.).Population data for the regions surrounding each airport was extracted from U.S. Census Bureau data for the year 2000.This population data extended well beyond the areas in which noise exposures would change due to effects of DSTs, so all such changes were captured in the noise calculations described below.
+Types of Opportunities SimulatedThe impact of new procedures and improved navigational capabilities was approximated by modifying the original operational data in different ways.Table 2 summarizes the general types of noise mitigation opportunities explored during this study, along with the relevant data characteristic modified in order to simulate the opportunity.The noise mitigation opportunities that were evaluated included "Avoid Dive and Drive", "Direct Climb to Cruise", the construction of additional runways, and the movement of "noisy" aircraft to noise-preferred runways.As can be seen in Table 2, a large number of the simulations involved the modification of track (lat/lon) location.However, mitigation strategies requiring altitude profile modifications, such as "Avoid Dive and Drive" and "Direct Climb-to-Cruise", were also addressed.Each opportunity's evaluation consisted of: assessing the current traffic situation (as described by operational data, aeronautical charts, etc.) and proposing new procedures or capabilities (e.g., greater adherence to noise-preferred trajectories via RNAV type capability).Since nighttime operations are weighted so heavily by the Day-Night Sound Level (DNL) metric, a substantial effort was also made to reduce the spillage of evening-night shoulder events into the night hours and to increase the usage of noise-preferred runways for night operations.Simulation of mitigation opportunities involving wind, speed profiles, and noisepower distribution (NPD) curves were not addressed.Figure 3 shows a similar situation at BOS in which departing aircraft off runway 04R are postulated to have precise enough navigation to enable them to be tightly funneled over a low population area rather than being dispersed over more highly populated regions.In the case of Figure 4, the departure events from runways 19L/R were reassigned to existing tracks departing over the water on runway 10L.
+Results ObtainedAfter modifying the track and event data to simulate the effects of one or more noise-aware DSTsconsistent with the specific elements of the traffic patterns at each airport and the population distributions encountered in the vicinity of these patterns -a noise impact analysis was carried out using NIRS.Noise exposure was computed for the baseline (without DST) and alternative (with DST) cases, and differences between the cases were calculated.NIRS provides the capability to generate noise impact tables, graphs, and maps.The impact table and impact graph provide categorization of population centroids in noise bins of interest, and quantify the "before DST" and "after DST" effects on population centroids within each bin.The impact map provides a graphical depiction of the population centroids whose noise exposure categories are different between the baseline and alternative cases.
+Avoid EarlyTurns American Institute of Aeronautics and Astronautics (1) applying traffic-scaling factors that raised the 2001 traffic levels to those future levels estimated for ORD/MDW, BOS, and SFO and (2) re-calculating noise impacts at all population locations based on the new traffic levels.The chosen net measure of this noise mitigation benefit was the total number of people receiving annual DNL at 50 dB or greater.For our initial analysis, we assumed that the proposed mitigation strategies would be implemented for 100% of the affected flights.Thus, the noise mitigation results represent the impact of 100% compliance to the noise-mitigation traffic patterns.Later, we also assessed the sensitivity of these results to partial implementation of the proposed traffic modifications where possible.This assessment was done in recognition that operational limitations related to safety and capacity might constrain the application of the suggested noise-sensitive procedures.In other words, the postulated mitigation strategies may not be 100% effective.Thus, for each noise mitigation strategy, we enumerated the potential operational limitations that might inhibit their use.We then performed a simple analysis to provide a rough estimate of the impact of partial implementation on noise exposure benefits.As a first approximation, we have chosen to define a measure of DST effectiveness, called the DST Effectiveness Factor, which is meant to capture an estimate of the realizability of a given DST.We define this factor independently for each mitigation opportunity, since operational constraints vary across airports depending on the nature of their traffic flows.This DST Effectiveness Factor is used to scale the noise impact (measured in terms of the net change in the population experiencing noise of 50 dB or higher) to produce a more realistic estimate of the potential noise benefits.In order to compare the effectiveness of different mitigation strategies, we define a measure called the Expected Noise Benefit (ENB), as the product of the percentage decrease in the total population above 50dB DNL and the DST Effectiveness Factor.The ENB gives the total expected noise benefit associated with each mitigation opportunity.We provide an overall rating for each mitigation opportunity on the basis of the ENB to identify those with the highest potential value should the identified strategy be implemented.In this portion of our work, we manipulated tracks solely to quantify noise benefits measured in terms of reduced average exposure via the DNL metric.Estimating the impact that these manipulations would have on operations was beyond the scope of this effort, but every effort was made to perform the manipulations in a manner that would not have significant operational impact.Approaches, and Noisy Aircraft on Preferred Runways) at the estimated levels of DST effectiveness represents a 13% decrease in the population above 50 dB, or over 31,000 people These quantitative results lead to several general conclusions: (1) DSTs can provide substantial benefits at airports that have noise-mitigation opportunities similar to those analyzed in detail here for ORD, MDW, BOS, and SFO; (2) the benefits will probably lie in the range of 10% to 50% of the population exposed at 50 dB DNL or above; and (3) such benefits are likely to be extremely attractive to airports that desire to improve public acceptance of aircraft noise, especially in light of conflicting pressures for decreases in noise impact and increases in capacity.American Institute of Aeronautics and Astronautics
+NOISE AVOIDANCE PLANNERThe aforementioned study of mitigation opportunites consisted of essentially "static" changes to operational procedures.Although these changes showed potential for benefits from a noise perspective, there was no means of judging the impact of such changes on the operational efficiency of the airport.As a means of overcoming this limitation, we describe the initial development of the Noise Avoidance Planner (NAP).NAP, being developed under a Phase II SBIR with NASA Ames Research Center, is a noise-aware version of the CTAS Final Approach Spacing Tool (FAST) and Expedite Departure Path (EDP) DSTs.NAP is intended to operate dynamically in real-time to enable the FAST/EDP scheduling logic ( [6]) to utilize a noise figure-of-merit (FOM) in determining path stretching and speed modifications for resolution of spacing constraints.Currently, FAST and EDP operate on the basis of analysis categories -a set of unique states into which aircraft arriving or departing the terminal area are partitioned for the purpose of route generation, sequencing and conflict resolution.The DFW site adaptation database currently has approximately 520 unique analysis categories.For example, the category DFW_18R_BAMBE_ JET_BEFORE_FEEDER_GATE applies to a jet assigned to DFW Runway 18R while the aircraft is outside of the Bambe arrival metering fix.These analysis categories define the initial route for the aircraft as well as its confliction resolution sets.Associated with each analysis category is also a set of degrees-of-freedom (DOFs) that FAST and EDP utilize to realize the required path stretching and speed control for achieving the desired inter-aircraft spacing.A DOF is defined to have both FAST and SLOW limit values (e.g., the minimum and maximum extent of fanning from a particular waypoint).These FAST and SLOW limit values define the lower and upper bounds for the time required to fly along a portion of the trajectory.These bounds form the basis for spatial constraint resolution.This resolution is time-based in nature.Specifically, combinations of DOFs are sought which provide the necessary amount of delay to properly space leading and trailing aircraft at various segments American Institute of Aeronautics and Astronautics along the trajectory.At present, the FAST/EDP constraint resolution process terminates once a single satisficing solution has been found.The search through the possible space of DOF combinations is currently deterministic in nature and follows a pre-defined recipe (including a fixed set of mixing ratios for the various DOFs).From the perspective of generating noise-preferred advisories with FAST/EDP, however, this search methodology is unacceptable.What is needed is a set of satisficing solutions (those that satisfy all constraints) from which the best noise solution can be selected.Figure 5 illustrates this idea, showing a search through a sequence of points in DOF combination space (A i ) where multiple satisficing solutions (e.g., the green X's) are collected and compared relative to one another using a noise figureof-merit.to enable noise metrics to influence decisions, need to enable CTAS to discover multiple solutions -not just a single satisficing solution.A 1 A 2 A N A 3 A 4
+SET OF EQUALLY GOOD NON-NOISE SOLUTIONS noise scale best (noise + CTAS) solution returned
+Figure 5. Change in FAST/EDP Logic Needed for Noise Avoidance Planning
+ArchitectureAs an initial step toward introducing noise awareness into the FAST/EDP scheduling logic, we focus on the noise sensitivity of the vector (e.g., spatial) DOFs of each analysis category.As such, we define the trajectory space for a given category to be bounded spatially by the FAST and SLOW limits of its vector DOFs, assuming all other DOFs (e.g.speed DOFs) are set to their FAST limits (see Figure 6).This trajectory space defines a time band ranging from ∆ min (which represents no delay from the vector DOF) to ∆ max (which represents the most delay achievable with the vector DOF).In general, each vector DOF will have a different range of possible delay absorption values.We can define the noise exposure for the bounds on the trajectory space by computing the SEL experienced by the population underlying the FAST and SLOW trajectories, respectively.To determine the noise exposure for trajectories between these bounds we have two options.We could choose to linearly interpolate in the noise metric (SEL) space -assuming that the noise exposure level is monotonic between the two end points.Alternatively, we can choose to interpolate in the trajectory space first, and then subsequently compute the noise exposure values for each of the interpolated trajectories.This latter approach does a better job of capturing local fluctuations in noise exposure level due to the distribution of the underlying population.Space for Noise Avoidance Planning At present, the Noise Avoidance Planner consists of two distinct off-line processing steps combined with real-time data handling.The first step stimulates the FAST/EDP scheduling logic to generate the interpolated set of trajectories for each analysis category.The second step then processes these trajectory sets with NIRS to compute a noise sensitivity for each analysis category (and its associated vector DOF).These sensitivities take the form of noise exposure curves as a function of DOF value (i.e., as a function of delay absorption).Note that since the offline steps are solely a function of the site adaptation and population distribution, they only need to be performed once in their entirety.Results of this offline processing are stored in a database for access during run-time.Localized changes to the site adaptation can be accommodated simply by reprocessing any new or revised analysis categories.Changes in population distribution due to new census data will require reprocessing of all analysis categories.The run-time usage of the noise sensitivity data is anticipated to consist of a simple table lookup into the F: FAST limit S: SLOW limit ∆ min ∆ max American Institute of Aeronautics and Astronautics database to return the tradeoff between noise exposure and delay.The noise-aware version of the FAST/EDP scheduling logic will be modified to incorporate this sensitivity data to search for a combination of DOFs which minimizes the noise exposure level whenever possible.For this purpose, we assume noise preferred values of the vector DOFs will be selected and values of the speed DOFs will be chosen to achieve the remaining amount of delay needed for traffic separation.Figure 7 illustrates the basic components of the FAST/EDP architecture used to develop the initial NAP functionality.
+Off-Line Noise ComputationAs shown in Figure 7, the noise sensitivity generation process is driven by a Simulation File containing a set of aircraft radar hits which correspond one-to-one with the set of analysis categories for a particular site adaptation database.Our process for creating the Noise Sensitivity Database leverages off of the existing FAST/EDP flight processing logic.In the current architecture, the Communications Manager (CM) is used to distribute the flight plans to both the Route Analyzer (RA) and Profile Selector (PFS) processes.The CM then distributes the radar hits to the RA which classifies the aircraft into a particular analysis category.The RA uses the category's binary analysis tree (specifying the order in which its N DOFs are to be used in absorbing delay) to define the set of 2 N unique combinations of DOF limits.These combinations correspond, for example, to each of the paths to the right-most leaves of the tree in Figure 6.For each of these combinations of FAST and SLOW DOF values, the Trajectory Synthesizer returns the corresponding 4-dimensional trajectory.From this trajectory, the aircraft's time of arrival to the meter fix and/or runway can be computed.The set of arrival times is then passed to the Profile Selector (PFS) which uses its own TS to define the initial trajectories for all flights.Note that, unlike the RA which is event-driven in nature, the PFS scheduling process is initiated every six seconds.As such, the scheduling process is applied to all flights which accumulate between updates.Since we have established (in the Simulation File) only a single aircraft in each analysis category at any given time, there are nominally no conflicts for the PFS to resolve.Thus, no iteration through the DOF analysis tree is initiated.Since our approach hinges on the exploration of the vector DOFs, we have modified the PFS logic to initiate a pseudo resolution cycle in which we set the amount of delay to be absorbed incrementally between the FAST (zero seconds delay) and SLOW (∆ max ) limits.Each time through this cycle, the TS returns a resolution trajectory which is then recorded to a file.The set of stored trajectories is then post-processed using NIRS to develop the corresponding noise exposure values.In this fashion, we are able to span the trajectory space for each analysis category and define the noise sensitivity curve as a function of the category's vector DOFs.Examples of the exercising of the vector DOFs are given in Figure 8 and Figure 9 for DFW arrivals on 18R over BAMBE (FAN FROM WAYPOINT) and FEVER (BASELEG EXTENSION) respectively.
+ResultsAs an initial demonstration of the value of providing a noise figure-of-merit to FAST/EDP, we present the variation in Sound Energy Level (SEL) exposure over the space of trajectories for the two analysis categories described in the previous section.The variation (for population centroids experiencing greater than 55dB of exposure) is shown in Figure 10.We present the results in terms of the percentage of the total population (within a 30 nautical mile radius of DFW, or 4710027 people) experiencing SEL values greater than 55dB.This figure shows that interpolating linearly (in noise space or SEL value) between the FAST and SLOW limits can provide a rather poor estimate of the actual noise impact for the intermediate trajectories.Instead, by interpolating in DOF space (e.g., geometric trajectory space), we are able to capture the finer detail of the noise exposure/delay tradeoff surface.For example, simple linear interpolation in SEL space for the 18R BAMBE FAN FROM WAYPOINT category would predict that 2 percent of the population would experience SEL greater than 55dB for 100 seconds of delay absorption.Interpolating in trajectory space, however, provides a better estimate of 3.5 percent.Currently we use a simple uniform sampling (in DOF/time) between the two limits (e.g.∆ i = ∆ max /10).In general, one could define more sophisticated sampling schemes (for example an iterative bisecting scheme with a difference threshold) to maximize the capture of the details of the noise exposure surface between the DOF's FAST and SLOW limits.One can also assess the variation in noise exposure for different delay values graphically by examining the noise footprint created by displaying the color-coded SEL values for each population centroid.Figure 11(a)-(c) show the noise footprint created by arrivals into DFW 18R over BAMBE using the FAST route (zero delay), 69 seconds of delay, and 212 seconds of delay, respectively.A considerable shift of the noise footprint can be observed in these figures.In particular, as the aircraft fans closer to the SLOW limit, it is actually at a higher altitude prior to initiating its turn onto the final approach course.Therefore, the noise footprint is reduced for populations away from the final approach course.
+A MERGING OF PHILOSOPHIESThe previous section demonstrated the variability in noise exposure for a single aircraft flying each of the trajectories contained within the space bounded by the vector DOF's FAST and SLOW limits.These sensitivity curves are a first step towards the integration of noise-awareness into the decision support capabilities of ATM tools such as FAST/EDP.Such a capability can allow noise to influence localized routing decisions (e.g., trading path stretching for speed adjustment).Will aggregation of noise-preferred trajectories (on the basis of SEL) over the course of a 24-hour period result in a net reduction in noise exposure (in terms of DNL)?Shaping the noise impact of routings on a given population distribution is a terminal area-wide airspace utilization problem.FAST/EDP, however, are designed as tactical, time-based sequencing tools.Therefore, it seems plausible that if one consistently modifies traffic flow on a given segment in a similar fashion (e.g., locally noise-preferred), one could inadvertently create a new noise problem under that modified flow.This leads one to the consideration of adding a "rolling window" type of noise exposure Base extension Arrivals over FEVER FAST limit SLOW limit American Institute of Aeronautics and Astronautics measure to the run-time FAST/EDP processing.In this manner, statistics regarding the distribution of noise exposure over the affected communities as a whole could be collected and used to either positively or negatively reinforce certain routing decisions in a timevarying manner.The one-to-one correspondence between analysis category and DOF sensitivity we take advantage of at present is enabled by the fact that only a single vector DOF is currently defined (and used for resolution) for each analysis category.If one relaxes that restriction such that multiple vector DOFs are chained together, the resulting bounds of the reachable trajectory space increase in complexity.One then must consider various combinations of limit trajectories and take into account the spatial coupling of multiple vector DOFs.This reiterates the point that the noise exposure for a trajectory is defined by the combination of all segments, not necessarily a single segment.One possible approach to this problem, derived through analogy to graph search algorithms such as A* ( [7]), is to develop a single heuristic estimate of the noise sensitivity downstream of a given decision point for each possible "branch" of the decision tree.This would allow a tool such as NAP to condition current routing decisions based on an estimate of future noise impact.The idea would be to avoid situations in which a locally optimal sequencing decision on one segment leads to a excessively high noise impact a later segment.The current NAP architecture leverages the existing FAST/EDP infrastructure to explore the trajectory space.This was a natural choice given the manner in which FAST/EDP uses this space to resolve spatial conflicts.Another option, however, would be to essentially ignore the existing scheduling logic and instead, simply search for an "optimal" (e.g., in a "global" sense -with minimal procedural constraints) noise trajectory for a given aircraft/engine combination from each meter fix to the runway (or vice versa for departures).This trajectory could be represented as a "cloud" defining the relative sensitivity of noise values in its immediate neighborhood.In other words, any trajectory contained in the cloud would be essentially equal from a noise perspective.The burden would then fall on FAST/EDP to stay within this noise preferred region as much as possible given its sequencing and scheduling constraints.An obvious issue with this approach is if the noise-preferred cloud region does not overlap with the FAST/EDP trajectory space.In this case, the current technique of spanning the FAST/EDP trajectory space in some manner would be applicable.
+CONCLUSIONS AND FUTURE WORKWe have described results obtained from a quantitative study which demonstrated the potential benefits of noise-aware DSTs by simulating modifications to terminal area air traffic control procedures.What this initial benefits study lacked was a means of addressing the throughput and efficiency impacts related to such procedures.We then presented some initial results obtained during the development of a noise-aware version of the FAST/EDP DSTs, called the Noise Avoidance Planner.We focused on vector degrees of freedom to develop their noise sensitivity with respect to delay absorption.These results seem to indicate that there is potential for tools such as FAST (arrivals) and EDP (departures) to factor noise into their sequencing and scheduling decisions.It was pointed out, however, that decisions that provide a noise benefit for certain population centroids can have a corresponding negative impact (e.g., increase in noise exposure) for other locations.Aggregate measures of benefit, potentially over extended time horizons, are thus generally preferred.The definition and evaluation of such metrics are the next steps in the development of the Noise Avoidance Planner.Future research will involve examination of the impact of the vertical degrees of freedom (thrust, altitude profile) and increased emphasis on departure scenarios.Finally, it should be pointed out that noise exposure is a time-varying phenomenon that is a strong function of the terminal area weather conditions, including cloud coverage and winds.For this reason, a spatial trajectory that is noise-preferred on a calm, clear day may not be the same as that required on a windy day with low ceilings.Future research is needed to incorporate actual and forecast winds into the noise prediction process -with a need for real-time evaluation of the trajectory space given winds.Figure 1 .1Figure 1.Identification of potentially noise-sensitive ATC procedural separation at ORD and MDW
+Figure 2 through2Figure 2 through Figure 4 illustrate several of the mitigation opportunities simulated in this manner.
+Figure 2 .2Figure 2. Improved Flight Corridor Adherence for ORD Runway 22L Departures
+Figure 3 .3Figure 3. Increased Flight Corridor Adherence for BOS Runway 04R Departures
+Figure 4 .4Figure 4. SFO Noisy Aircraft Departures (19L/19R) Reassigned to 10L (red = original; cyan = modified)In each of these figures, the original tracks are shown in red, with the corresponding modified tracks indicated in cyan or blue.Figure2highlights a mitigation strategy for ORD which involves tighter adherence to departure corridors specified by the Chicago Fly Quiet Program -in this case, keeping departure tracks over highways for an extended period during the climb out prior to initiating their turns.Figure3shows a similar situation at BOS in which departing aircraft off runway 04R are postulated to have precise enough navigation to enable them to be tightly funneled over a low population area rather than being dispersed over more highly populated regions.In the case of Figure4, the departure events from runways 19L/R were reassigned to existing tracks departing over the water on runway 10L.Results ObtainedAfter modifying the track and event data to simulate the effects of one or more noise-aware DSTsconsistent with the specific elements of the traffic patterns at each airport and the population distributions encountered in the vicinity of these patterns -a noise impact analysis was carried out using NIRS.
+Impacts for future years (2006 and 2011) were obtained from the baseline and modified 2001 data by baseline tracks modified tracks
+Figure 6 .6Figure 6.Defining the Endpoints of the TrajectorySpace for Noise Avoidance Planning At present, the Noise Avoidance Planner consists of two distinct off-line processing steps combined with real-time data handling.The first step stimulates the FAST/EDP scheduling logic to generate the interpolated set of trajectories for each analysis category.The second step then processes these trajectory sets with NIRS to compute a noise sensitivity for each analysis category (and its associated vector DOF).These sensitivities take the form of noise exposure curves as a function of DOF value (i.e., as a function of delay absorption).Note that since the offline steps are solely a function of the site adaptation and population distribution, they only need to be performed once in their entirety.Results of this offline processing are stored in a database for access during run-time.Localized changes to the site adaptation can be accommodated simply by reprocessing any new or revised analysis categories.Changes in population distribution due to new census data will require reprocessing of all analysis categories.
+Figure 7 .7Figure 7. Basic Noise Avoidance Planning Architecture (NAP-specific components yellow)
+Figure 8 .Figure 9 .89Figure 8.The trajectory space spanned by the FAN FROM WAYPOINT DOF for arrivals over BAMBE
+Figure 10 .10Figure 10.The variation of noise SEL with respect to delay absorption for several DOFs
+(a) No delay (FAST) route (b) 69 Seconds of Delay (c) 212 Seconds of Delay Figure 11.Frames (a)-(c) show noise footprints for arrivals into DFW 18R for different delay values.DFW 18R BAMBE ARRIVALS DFW 18R BAMBE ARRIVALS DFW 18R BAMBE ARRIVALS American Institute of Aeronautics and Astronautics
+Table 1 . Summary of Track and Event Data Sources for Quantification Study Study Airport Original Data Source Original Data Format1ORD&Chicago TRACONNIRS tracks and eventsMDWAnalysisProjectforthefivedata for 2000configurationsmostoften used on an annualbasisBOSMassPort 2001 dataPre-INM tracks andand Metron 1997INM tracks/events fordataaverage annual daySFOSFO Noise OfficeINM tracks/events for2001 dataaverage annual day
+Table 2 .2Summary of mitigation opportunitiesNoise MitigationData CharacteristicsOpportunityModifiedNoise-sensitive ATM approachproceduresAvoid dive and driveAltitude profileAvoid base leg extensionTrack locationinto noise sensitive areasSide-step approachesTrack location
+Route tracking (stay in precise route corridor)Follow routes over lowTrack locationpopulation areasAvoid shortcuttingTrack locationRunway/route selectionFanning across regionTrack locationRoute older aircraft to lessTrack location,noise-sensitive runwaysequipment typeGreater usage of noise-Runwaypreferred runwaysassignmentAirport interactions within aTRACONModify existing proceduresTrack location,to consider noisealtitude profilesNighttime operationsExtend procedures to higherEvent timetraffic levelsImprove efficiency so thatEvent timenight time operations can beinitiated on timeNoise-sensitiveATMdeparturesAltitude/speedDirect climb-to-cruiseprofiles American Institute of Aeronautics and Astronautics
+Table 3 .3Categorizing Expected Noise BenefitsCategory Percent Improvement RequiredHigh>=10% reduction in population above 50 dBModeratefrom 2% to 10% decreaseLowless than 2% decreaseThe summary of quantitative results in Table 4indicates that Expected Noise Benefits vary across abroad range, from less than 0.1% to over 20%. This isdue to the enormously varied traffic patterns,mitigation-opportunity characteristics, and populationdistributions across the airports studied. In particular:• At ORD and MDW, 2 of the 7 mitigationopportunities studied were rated high or moderateimpact. Achievement of the 2 high and moderatemitigation objectives at ORD and MDW (PreferredFlight Track Conformance and Direct Climb toCruise) at the estimated levels of DSTeffectiveness represents a 15% decrease in thepopulation above 50 dB, or over 460,000 people.• At BOS, all 4 of the mitigation opportunitiesstudied rated high or moderate impact.Achievement of the 4 high and moderatemitigation objectives at BOS (Arrival andDeparture Corridor Adherence, Noisy Aircraft onPreferred Runways, and Night Operations onPreferred Runways) at the estimated levels of DSTeffectiveness represents a 47% decrease in thepopulation above 50 dB, or over 105,000 people• At SFO, 4 of the 7 mitigation opportunities studiedrate high or moderate impact. Achievement of the4 moderate mitigation objectives at SFO (Shorelineand Dumbarton Departures, Quiet Bridge
+Table 4 .4Summary of Noise Mitigation Effectiveness Including DST EfficiencyNet Population Noise Impact(100% effective)
+		
+		
+
+			
+ACKNOWLEDGEMENTSThis research was funded by the Advanced Air Transportation Technologies (AATT) Project at NASA Ames Research Center under Research Task Order 61 of the Air Traffic Management System Development and Integration (ATMSDI) contract and a Phase II SBIR through NASA Ames Research Center.
+			
+
+			
+
+				
+
+
+	
+		Optimization of noise abatement arrival trajectories
+		
+			HGVisser
+		
+		
+			RA AWijnen
+		
+		10.2514/6.2001-4222
+	
+	
+		AIAA Guidance, Navigation, and Control Conference and Exhibit
+		Montreal, Canada
+		
+			American Institute of Aeronautics and Astronautics
+			August 2001
+		
+	
+	Visser, H.G., Wijnen, R.A.A., "Optimization of Noise Abatement Arrival Trajectories," Proc. of the AIAA GNC Conference, Montreal, Canada, August 2001.
+
+
+
+	
+		Research into new noise abatement procedures for the 21st century
+		
+			LouisJErkelens
+		
+		10.2514/6.2000-4474
+	
+	
+		AIAA Guidance, Navigation, and Control Conference and Exhibit
+		Denver, Colorado
+		
+			American Institute of Aeronautics and Astronautics
+			August 2000
+		
+	
+	Erkelens, L.J., "Research into New Noise Abatement Procedures for the 21 st Century", Proc. of the AIAA GNC Conference, Denver, Colorado, August 2000.
+
+
+
+	
+		Systems analysis of noise abatement procedures enables by advanced flight guidance technology
+		
+			John-PaulClarke
+		
+		
+			RHansman, Jr.
+		
+		
+			John-PaulClarke
+		
+		
+			RHansman, Jr.
+		
+		10.2514/6.1997-490
+	
+	
+		35th Aerospace Sciences Meeting and Exhibit
+		Reno
+		
+			American Institute of Aeronautics and Astronautics
+			Jan 1997
+		
+	
+	Proc. of the 35 th Aerospace Sciences Meeting
+	Clarke, J.P., and R.J. Hansman, "Systems Analysis of Noise Abatement Procedures Enabled by Advanced Flight Guidance Technology", Proc. of the 35 th Aerospace Sciences Meeting, Reno, Jan 1997.
+
+
+
+	
+		Quantification of Noise Benefits for ATM Decision Support Tools, Task 2: Quantification of Benefits Results for
+		
+			SAugustine
+		
+		
+			BCapozzi
+		
+		
+			JDifelici
+		
+		
+			TThompson
+		
+		
+			MdwOrd
+		
+		
+			BosSfo
+		
+	
+	
+		Final Report for Research Task Order
+		
+			61
+			February 2002
+		
+	
+	Contract NAS2-980005
+	Augustine, S., Capozzi, B., DiFelici, J., and T. Thompson, "Quantification of Noise Benefits for ATM Decision Support Tools, Task 2: Quantification of Benefits Results for ORD, MDW, BOS, and SFO", Final Report for Research Task Order 61 (Contract NAS2-980005), February 2002.
+
+
+
+	
+		Data-Driven Analysis of Departure Procedures for Aviation Noise Mitigation
+		
+			JiratBhanpato
+			0000-0001-8246-2425
+		
+		
+			TejasGPuranik
+			0000-0002-4701-0674
+		
+		
+			DimitriNMavris
+		
+		10.3390/engproc2021013002
+	
+	
+		The 9th OpenSky Symposium
+		
+			MDPI
+			July 2001
+		
+	
+	Landrum and Brown, Inc., and Metron Aviation, "Benefits Analysis for Noise Mitigation, Task 1: Survey of Noise Issues Potentially Related to ATM Operational Procedures", July 2001.
+
+
+
+	
+		A concurrent sequencing and deconfliction algorithm for terminal area air traffic control
+		
+			JohnRobinso
+		
+		
+			DouglasIsaacson
+		
+		10.2514/6.2000-4473
+	
+	
+		AIAA Guidance, Navigation, and Control Conference and Exhibit
+		Denver, Colorado
+		
+			American Institute of Aeronautics and Astronautics
+			August 2000
+			
+		
+	
+	Robinson, J.E. III, and D.R. Isaacson, "A Concurrent Sequencing and Deconfliction Algorithm for Terminal Area Air Traffic Control", Proc. of the AIAA GNC Conference, Denver, Colorado, 14-17 August 2000.
+
+
+
+	
+		Principles of Artificial Intelligence
+		
+			NJNilsson
+		
+		
+			1980
+			Tioga Pub. Co
+			Palo Alto, CA
+		
+	
+	Nilsson, N.J., Principles of Artificial Intelligence, Tioga Pub. Co., Palo Alto, CA, 1980.
+
+
+				
+			
+		
+	

file117.txt

@@ -0,0 +1,246 @@
+
+	
+	
+		
+IntroductionThe continued growth of air traffic within the United States, combined with the use of Òhub and spokeÓ operations by air carriers, has led t o increased congestion and delays in the terminal airspace surrounding the nationÕs busier airports.The problem of congestion is exacerbated at hub airports, where air carriers schedule large numbers of flights to arrive and depart within a short time period.These arriving and departing groups of aircraft are commonly referred to as banks, and the simultaneous arrival of several banks of aircraft can easily strain the capacity of an airport.In order to ensure that the safe capacity of the terminal area is not exceeded, ATM often places restrictions on arriving flights transitioning from en route airspace to terminal airspace.The constraint of arrival traffic is commonly referred to as arrival flow management, and includes techniques such as metering, vectoring, and the imposition of miles-in-trail restrictions.These constraints are enacted without regard for the relative priority which airlines may be placing on individual flights, based on factors such as crew criticality, passenger connectivity, critical turnaround times, gate availability, on-time performance, fuel status, or runway preference [1].To air carriers, ÒhubbingÓ makes good economic and competitive sense [2].At the same time, however, hubbing operations often lead to overcapacity periods and precipitate delays which can directly impact the economic efficiency of an air carrierÕs flight operations.Air traffic control automation tools are used in arrival flow management to assist controllers in efficiently matching traffic demand and airport capacity while minimizing delays.These tools use sequencing and scheduling algorithms t o automatically plan the most efficient landing order and landing times for arriving aircraft [3].NASA and the Federal Aviation Administration (FAA) have designed and developed a suite of software decision support tools (DSTs) to improve the efficiency of high-density airspace [4].Collectively known as the Center-TRACON Automation System (CTAS), operational evaluation of these DSTs has shown them to be effective in improving airport throughput and reducing delays while maintaining controller workload at a reasonable level [4].One of these tools, the Traffic Management Advisor (TMA), is currently being used at the Fort Worth Air Route Traffic Control Center (Center) to perform arrival flow management of traffic into the Dallas/Fort Worth airport (DFW).The TMA is a time-based planning tool that assists Traffic Management Coordinators (TMCs) and Center controllers in efficiently balancing arrival demand with airport capacity [5].The primary algorithm in the TMA is a real-time scheduler which generates efficient landing sequences and landing times for arrivals within about 200 n.mi.from touchdown [6].Aircraft are scheduled so that they arrive in a firstcome-first-served (FCFS) order based on an estimated time of arrival (ETA) at the runway.While FCFS scheduling establishes a fair order based on estimated times of arrival, it does not take into account individual airline priorities among incoming flights.The development of new arrival flow management techniques which consider priorities expressed by air carriers will likely reduce the economic impact of ATM restrictions on the airlines.This will in the future lead to increased airline economic efficiency by allowing airlines to have greater control over their individual arrival banks of aircraft.As part of its Collaborative Arrival Planning (CAP) research and development program, NASA-Ames is exploring the possibility of allowing airlines to express relative arrival priorities to ATM through the development of new CTAS sequencing and scheduling algorithms which take into account airline arrival preferences.An earlier study focused on the feasibility of scheduling Òdelay exchangesÓ among pairs of individual arrival aircraft as a means of accommodating an airline request for an earlier arrival [7].
+Priority SchedulingSuccessful airline operations today require increasingly complex airline schedules.The interconnection of the schedules of major airlines with their subsidiary carriers and code-sharing partners adds to this complexity.As a result of increasing scheduling complexities and interdependencies, achieving a specific order within a bank of arrival aircraft has become of greater importance to the smooth and efficient operation of many airlines.Even a small group of aircraft belonging to a single airline may be interconnected in a fairly complex manner, with passengers and cargo from multiple flights feeding one large connecting flight or vice-versa.For example, an arriving bank of aircraft may include a large jet which is primarily delivering passengers to a number of smaller turboprop aircraft arriving in the same bank.This same large jet in turn, may be taking on passengers from other jet aircraft in the bank to deliver passengers/cargo to their final destinations.Passenger connectivity is only one of many factors which influence an airlineÕs schedule.Consideration must also be given to factors such as the availability of gates, and ground equipment and personnel to service aircraft and transfer passengers and cargo between flights.Even in the simple example just cited, the efficient operation of these flights will depend strongly on maintaining the integrity of the airline schedule by meeting the planned times of arrival and hence the desired order of arrival.For most airlines, the schedule which is determined internally by the airline to satisfy its business and economic objectives is an ÒidealÓ schedule.This schedule is ideal in the sense that the everyday realities of operating an airline and interacting with the various elements of the National Airspace System (NAS) largely preclude this ideal schedule from ever being achieved.Because of the uncertainties throughout both the airline (equipment breakdowns, maintenance problems, personnel shortages) and the NAS (weather, ground delays, ATM restrictions), aircraft often arrive in the terminal airspace in an order which does not match the ideal order of the airline schedule.Current arrival flow management using FCFS sequencing and scheduling algorithms will likely result in aircraft arriving at the runways in an order which does not match the preferred arrival order.The ability to specify the preferred arrival order within the userÕs own arrival bank is useful for maximizing bank integrity and minimizing bank time (i.e., exchange of passengers/cargo, and aircraft servicing) [8].Earlier studies have shown that scheduling aircraft according to an FCFS sequence based on estimated time of arrival at the runway produces a schedule which is considered to be both fair to air carriers and efficient in terms of minimizing delays which must be absorbed [3].These studies also have shown that the resulting scheduled arrival sequence at the runway will, for the most part, match the FCFS sequence which is input to the scheduling algorithm.Because the scheduling algorithm attempts to preserve the input sequence, specifying a preferred sequence will result in a schedule which closely approximates the preferred arrival order.The concept of Òpriority schedulingÓ is then defined as the scheduling of a bank of arrival traffic according to a preferred order of arrival.The focus of the present study is t o determine the feasibility of scheduling a bank of arrival aircraft using a preferred sequence instead of an FCFS sequence based on ETA at the runway.
+It is importantto distinguish between ÒschedulingÓ or ÒscheduleÓ in the context of airline operations, and ÒschedulingÓ or ÒscheduleÓ in the context of air traffic control automation.The former refers to the daily scheduled times of departure and arrival which an airline determines for all of its flights, while the latter refers to the process of automatically choosing (a) the order or sequence in which the aircraft should land or cross a particular fix, and (b) the time that each aircraft in the sequence should pass over a specified fix [6].
+Fast-time SimulationA fast-time simulation originally developed for statistical evaluation of CTAS sequencing and scheduling algorithms has been modified for use in this investigation [9].In contrast to real-time simulation or field tests, which would require on the order of ninety minutes to examine a single traffic rush period, the fast-time simulation allows examination of large numbers of statistically similar rush periods in a matter of minutes.For each simulated traffic situation, the deviation of a designated bankÕs scheduled arrival order from the preferred arrival order can be determined.The impact of priority scheduling on delays is also determined by comparing delays for priority scheduling and FCFS scheduling.Because this simulation does not provide any information about the controller workload required to meet the calculated schedule, the output of the simulation is used only to determine the effectiveness of priority scheduling and its impact on scheduled delays.The fast-time simulation is comprised of three major components: an airport model, a statistical model of the arrival traffic flow, and the scheduler.
+Airport ModelThe arrival airspace at DFW is divided into Center and Terminal Radar Approach Control (TRACON) regions, with the TRACON encompassing the airspace within approximately 40 n.mi. of the airport.Arrival traffic is merged at four waypoints on the Center-TRACON boundary which correspond to the four primary arrival directions.These waypoints are referred to as feeder gates because during heavy traffic periods traffic is funnelled through these gates as a means of controlling or metering the flow rate into the terminal area [6].Traffic flowing to each gate is separated into two independent streams which are vertically separated by 2,000 feet at the feeder gate.This allows jet and turboprop aircraft, which have significantly different airspeed ranges, t o cross the feeder gates independently and avoid conflicts due to overtakes near the gates.The airport is modeled according to the landing practices at DFW with four feeder gates and three runways available for landing.The runways are considered to be independent so that no stagger requirements are necessary for scheduling.The airport model is comprised of the minimum flight times from each feeder gate to all landing runways for each independent stream.These TRACON transition times were obtained from an analysis using the minimum flight times measured for several traffic samples [10].The TRACON transition times vary with feeder gate, aircraft type, runway assignment, and airport configuration.The airport model contains transition times for both airport configurations at DFW: Ònorth flowÓ with arrival traffic arriving/departing in a northerly direction, and Òsouth flowÓ with traffic arriving/departing in a southerly direction.It should be noted that since the data used in this simulation were collected, a fourth arrival runway has been added at DFW.However, the three-runway model and traffic data are sufficient for purposes of this investigation.
+Traffic ModelThe traffic model is based on actual traffic data recorded during six rush periods at DFW.Although the traffic data were recorded over a span of several months, the mix of aircraft type remained nearly constant for each traffic sample.The data were recorded during the Ònoon balloon,Ó a daily arrival rush lasting approximately ninety minutes.The noon balloon was chosen as the basis for the traffic model because during this arrival rush demand exceeds airport capacity and air traffic managers impose time-based metering restrictions through CTAS sequencing and scheduling algorithms.Data recorded during the six rush periods include the aircraft type, aircraft identification, arrival stream, and the estimated time of arrival at the feeder gate (ETA FG ).The average of these estimated times of arrival for the six rushes is taken as the nominal ETA FG .Errors in aircraft time of arrival in Center airspace are modeled by adding an approximately Gaussian distribution to the nominal estimated time of arrival at the feeder gate.The maximum range of the variation in the ETA FG is specified as an input to the simulation and is referred to as the Center arrival error.
+Bank DefinitionAlthough an actual arrival bank of aircraft for an airline may consist of between 30 and 50 aircraft, in this study it is assumed that a bank is comprised of a single group of up to 20 aircraft belonging t o one airline and its subsidiary carrier.With a majority of the flights in the traffic model belonging to American Airlines (AAL) and American Eagle (EGF), these flights are used t o form arrival banks.The bank is not a contiguous set of aircraft because aircraft belonging to other airlines are interspersed among the bank aircraft, as would be the case in a real traffic situation.The bank of aircraft is defined by specifying the first member of the bank, and the number of aircraft belonging to the bank.For the purposes of this simulation, we assume that the preferred order of arrival at the runway equals the order of arrival based on the minimum ETA at the runway with no Center arrival error.Each of the bank aircraft is assigned a priority ranking which is simply equal t o the preferred order of arrival for the aircraft within the bank.The minimum estimated time of arrival at the runway (ETA RWY ) is calculated by adding the TRACON transition times for each of the three runways to the nominal ETA FG , and selecting the minimum of the three resulting values.This ETA RWY represents the earliest possible time of arrival for an aircraft provided that the aircraft could fly to the runway with no delay.For example, consider the list of aircraft shown in Table 1, which represents a portion of a single arrival rush where AAL1150 has been designated as the lead aircraft in the bank, and the number of aircraft in the bank has been specified as five.The number in the first column represents the sequence number or position of the aircraft within the arrival rush when the aircraft are time-ordered according to increasing ETA RWY .Each arrival rush or traffic sample consists of 108 aircraft, and in the example in Table 1 the aircraft belonging t o the defined bank range from the 57th aircraft t o the 65th aircraft in the arrival rush (AAL1554).The resulting bank aircraft are denoted by bold text for purposes of illustration.This example shows that aircraft belonging to other airlines are interspersed among the arrival aircraft which comprise the bank.The second column is the aircraft identifier and the third column is each aircraftÕs corresponding minimum ETA RWY .The fourth column shows the priority ranking which is assigned to each of the aircraft belonging to the bank based on this preferred order of arrival.
+Table 1 Bank definition and preferred arrival orderThe actual order of arrival for aircraft in a traffic rush period is generated by adding the Center arrival error to the nominal ETA FG .The Center arrival error represents the uncertainties in the NAS which cause the same flight to arrive in Center airspace at different times on different days.Because the minimum ETA RWY is calculated by adding a TRACON transition time to the ETA FG , the minimum ETA RWY will also vary.As a result, when the aircraft are ordered according t o increasing ETA RWY , the actual order for the bank aircraft will differ from the preferred arrival order.In addition, the number of aircraft interspersed among the arrival bank may vary because the variation in arrival time is modeled for all aircraft in the traffic rush, not only those belonging to the specified bank.Table 2 shows the resulting estimated arrival order for the specified bank when a Center arrival error having a range of up to +/-5 minutes is added to the traffic sample.
+FCFS SchedulingThe FCFS scheduler is intended to approximate the sequencing and scheduling algorithms presently used in CTAS at the Ft.Worth Center.A detailed description of the actual scheduling algorithm can be found in [6].Aircraft are sequenced and scheduled to be first-come-first-served at both the feeder gates and runways while meeting feeder gate and runway threshold separation constraints.Because scheduling is done in time rather than distance, the prescribed minimum separation criteria are translated into minimum time separations at both the feeder gates and the runway threshold.For aircraft crossing the feeder gate, the minimum in-trail separation requirement for aircraft is 5 n.mi., which is translated to a 60second time separation for purposes of this simulation.The separation criteria at the runway threshold are a function of both aircraft weight class and landing order as determined by the FAAÕs wake vortex safety rules.Airport acceptance rate (AAR) is taken into consideration by limiting the number of aircraft which are allowed to enter the TRACON in sliding ten minute intervals, and the scheduler balances flights between runways t o minimize overall delay.The FCFS sequence is established by time-ordering arrival aircraft according to increasing ETA RWY .Beginning with the first aircraft in the sequence, each aircraft is tentatively scheduled to each of the three runways, while ensuring that the prescribed minimum time separation between aircraft at the runway thresholds is met for each subsequent aircraft.The runway which results in the earliest scheduled time of arrival for the aircraft at the runway (STA RWY ) is then chosen as the landing runway .Scheduling to the runway automatically provides the correct amount of traffic to load the runways equally when traffic is heavy (runway balancing), and directs aircraft t o the closest available runway.The scheduled time of arrival at the feeder gate (STA FG ) is determined by subtracting the sum of the TRACON transition time and any TRACON delay from the previously calculated STA RWY .Finally, if STA FG Õs for two flights are less than the required 60 seconds apart, the scheduled times will be altered to meet the required separation at the feeder gate.Table 3 shows the resulting order of arrival when the aircraft are scheduled according to an FCFS sequence.The priority ranking of each bank aircraft is shown in parenthesis following the aircraft identifier.The second and third columns in the table show the FCFS sequence which is input t o the scheduler, with the aircraft time-ordered according to increasing ETA RWY .The fourth and fifth columns are the resulting schedule, with aircraft time-ordered according to increasing STA RWY .Note that the resulting scheduled order of arrival at the runway does not precisely match the FCFS sequence based on ETA RWY which is input t o the scheduler.Because the schedule must meet intrail separation criteria at both the feeder gate and the runway threshold, and the separation criteria at the runway threshold are a function of aircraft weight class and landing order, the FCFS sequence may not be preserved at the runway.Among the aircraft belonging to the designated bank, flights AAL1934 and AAL1428 have shifted positions from the sequence which is input to the scheduler (as have aircraft DAL431 and AAL410, which do not belong to the designated bank).In this case, the position shift has resulted in a scheduled sequence which does more closely match the ideal or desired order of arrival than does the input FCFS sequence based on ETA RWY .However, it is purely fortuitous that the resulting schedule more closely matches the preferred order, and depending on the magnitude of the Center arrival error, the scheduled order may actually deviate further from the preferred order.
+Priority SchedulingThe priority scheduling algorithm is identical t o the FCFS algorithm with one exception: instead of time-ordering the aircraft according to increasing ETA RWY prior to scheduling, the arrival aircraft belonging to the designated bank are ordered according to their priority ranking, which establishes the bank aircraft in the preferred arrival order.It is important to note that only the aircraft belonging to the bank are reordered according to their priority ranking, and that other aircraft in the traffic sample are still sequenced in an FCFS order based on ETA RWY .By reordering only the bank aircraft and scheduling the remaining aircraft according to an FCFS sequence, the impact of the reordering on scheduling efficiency is minimized.Table 4 shows the resulting order of arrival when the bank aircraft are scheduled according to the preferred sequence of arrival.The second and third columns show the priority sequence which is input to the scheduler, with the bank aircraft ordered according to their priority ranking, and the remaining aircraft timeordered according to increasing ETA RWY .The fourth and fifth columns show the resulting schedule time-ordered according to STA RWY .As was the case with FCFS scheduling, the resulting order of arrival does not match the sequence which was input to the scheduler because the schedule must meet separation criteria at the runway threshold which are a function of aircraft weight class and landing order.Although the resulting scheduled bank order does not precisely match the preferred order, it does indeed match more closely the preferred bank order than does the FCFS schedule shown in Table 3.
+Order DeviationTo quantify the effectiveness of the priority scheduling method we need a measure of how closely the scheduled order of arrival for a designated bank matches the preferred arrival order.We first define a position shift (PS) for an aircraft as the difference between the aircraft position in the preferred bank order and the sequence number in the scheduled bank order.
+PS N N PREFERRED SCHEDULED
+= -where N is the sequence number of the aircraft within the bank Table 5 illustrates the calculation of the PS for each of the aircraft in the bank defined in Table 1.The position shift of each aircraft is calculated for both FCFS scheduling (Table 3) and priority scheduling (Table 4).Note that a positive PS indicates that an aircraft is scheduled ahead of its preferred position in the bank, and a negative position shift indicates that an aircraft is scheduled behind its preferred position in the bank.For example, the sequence number of flight EGF628 in the preferred order of arrival is 2 while its sequence number in the FCFS schedule is 5 and its sequence number in the priority schedule is 3.This results in a PS of -3 for the FCFS schedule and -1 for the priority schedule and reflects the fact that EGF628 is scheduled 3 slots behind its preferred position in the bank using FCFS scheduling, and 1 slot behind the preferred position using priority scheduling.Because we are interested in how closely the overall bank order matches the preferred order, we want a single measure which will indicate the deviation from the preferred order for a bank of any length.We then define the order deviation (OD) for a bank as the algebraic sum of the absolute value of the PS for each aircraft in the bank divided by the number of aircraft in the bank.
+OD PS= ∑ | | # of bank aircraft
+# of bank aircraftIt can be seen from this definition that if the OD for a bank of aircraft equals zero, then the scheduled bank order is the same as the preferred bank order.More importantly, the larger the value of the OD, the further the scheduled bank order deviates from the preferred order.This will allow us to easily compare the relative effectiveness of the FCFS and priority scheduling methods in producing the preferred order of arrival.The order deviations for each scheduling method using the example in Table 5 are calculated below.Because the priority scheduling scheme results in the designated bank arriving in an order which more closely matches the preferred arrival order, the OD for the priority scheduled bank is smaller than that for the FCFS scheduled bank.ODFCFS = + -+ + + = | | | | | | | | | | . 0 3 1 1 1 5 1 2 ODPRIORITY = + -+ + + = | | | | | | | | | | . 0 1 1 0 0 5 0 4In order to investigate the statistical performance of the two scheduling methods, a large number of traffic samples are generated for a specified bank.To compare the effectiveness of FCFS scheduling and priority scheduling for a large number of traffic samples, we define the average OD as the sum of the ODÕs for each traffic sample divided by the number of traffic samples.
+OD
+Simulation Inputs/OutputsInputs to the fast-time simulation include the aircraft identifier of the lead aircraft in the bank, the size of the bank, the number of traffic samples, the range in Center arrival error, the airport configuration, and airport acceptance rate.In order to determine the statistical performance of the FCFS algorithm and the priority algorithm, 500 traffic samples are generated for each designated bank.Each traffic sample is comprised of 108 jet and turboprop aircraft, 72 of which are AAL or EGF flights.In this simulation the modeled airport configuration is south flow for DFW.Because the traffic model is limited to a single arrival rush period, and because of the manner in which a bank is defined, banks cannot be formed at or near the end of the arrival rush period.For example, if the bank length is specified as 20, and the designated lead aircraft is the 100th aircraft in the arrival rush, no bank will be formed because there are not enough aircraft following the lead aircraft to form a bank.Although we attempt to form banks across the entire range of the traffic rush period, this cannot be done for the reasons just outlined.The output of the fast-time simulation includes the average OD as well as histograms of the position shifts for each bank of aircraft.Total delays and histograms of individual delays for all aircraft in the traffic rush are generated as well.Results can then be compared for the FCFS scheduling algorithm and the priority scheduling algorithm.
+Results and DiscussionThe primary measure of success of the priority scheduling algorithm is the closeness of the match between the scheduled order of arrival and the preferred order of arrival.Figure 1 is a plot of the average order deviation for a bank size of 20, a range in Center arrival errors of +/-5 minutes, and an AAR of 96 aircraft/hour.For a designated bank whose lead aircraft has a nominal ETA FG given on the x-axis, a corresponding pair of ordinates shows the average OD for the bank using FCFS scheduling and priority scheduling.Figure 1 confirms that the priority scheduling algorithm significantly reduces the average OD from that of the FCFS scheduling algorithm.Note however, that while the OD for each bank is less using the priority scheduling algorithm, the OD is still non-zero for each bank.In other words, while the resulting bank order using priority scheduling matches much more closely the preferred order than does the FCFS order, the scheduled bank order does not precisely match the preferred order.Because the schedule must meet in-trail separation criteria at the runway threshold, and the separation criteria are a function of both weight class and landing order, the preferred order of arrival may not be preserved at the runway.Figure 1 shows the resulting OD for banks of aircraft beginning at different points in the arrival rush.The average order deviation for the FCFS algorithm first increases and then decreases as the ETA FG of the lead aircraft in the bank increases.The change in average OD for the FCFS schedule is due to changing traffic density and mixture in the arrival rush.As the traffic density increases (estimated times of arrival are more closely spaced), a given arrival error will cause larger position shifts within a bank and thus larger order deviations.By the same token, the traffic mix impacts the order deviation because if non-AAL/EGF flights are interspersed among the bank aircraft, the aircraft comprising the bank will be spaced farther apart.Then, for a given arrival error, the OD for the bank will be smaller because the aircraft are not as closely spaced.The average OD for the priority scheduling algorithm also varies with traffic density and mixture and is most effective in a region where some non-AAL/EGF aircraft are interspersed among the bank aircraft.The effects of AAR, bank size, and Center arrival error on the success of the priority scheduling algorithm are also examined.For the sake of brevity, no plots are shown but important results are summarized here.Results show that for a given Center arrival error and bank size, the priority OD tends to decrease with decreasing AAR, meaning that the priority scheduling algorithm is more effective for a more restrictive AAR.This is actually a characteristic of both the priority scheduler and the FCFS scheduler, and it can be shown that for a lower AAR, either scheduler is better able to preserve the order in which the aircraft are scheduled.Lowering the AAR effectively reduces the airport capacity (because demand remains constant), requiring that the scheduled times of arrival (STAÕs) be spaced farther apart.Because the STAÕs must be spaced farther apart, differences in crossing times or separation criteria are less likely to cause the resulting order to deviate from the order in which the aircraft are scheduled.Therefore the resulting schedule for either algorithm will more closely match the sequence in which the aircraft are scheduled.Results also show that increasing the size of the bank of aircraft does not significantly impact the effectiveness of the scheduling algorithm.However, increasing the magnitude of the Center arrival error for a given bank size and AAR does lead to a decrease in the effectiveness of the priority scheduling algorithm.For purposes of illustration, a histogram of the position shifts for a bank of aircraft led by AAL535 is shown in Figure 2.This histogram, along with the OD values labeled in Figure 1, demonstrate the relationship between average OD and the closeness of the match between the scheduled bank order and the preferred arrival order.Priority scheduling reduces the spread of the position shifts for the designated bank of aircraft.In this case, aircraft belonging to the designated bank are scheduled in the preferred position (position shift = 0) approximately 60% of the time using priority scheduling.Using FCFS scheduling, bank aircraft are scheduled in the preferred position only about 25% of the time.The increase in the number of aircraft scheduled in the preferred position leads to a decrease in average OD for the bank.Because this simulation does not provide any information about the controller workload required to meet the priority schedule, the output of the simulation is used only to investigate the feasibility of the priority scheduling method in terms of scheduling efficiency.However, it can be reasonably assumed that an increase in scheduled delays greater than a certain amount would be unacceptable to air traffic controllers because of the likely adverse effect on controller workload.Similarly, airlines would likely find an increase in scheduled delays which exceeds a certain threshold to be unacceptable from the standpoint of increased costs.While the amount of delay increase acceptable controllers and airlines would have to be determined before a priority scheduling method could be considered practicable, the present simulation provides initial insight into the impact of priority scheduling on scheduling efficiency.This can be measured as the change in average delay per aircraft when priority scheduling is used instead of FCFS scheduling.q q q q q q q q q q q q q q q q q q q q q q q q q For each designated arrival bank whose order deviation is shown in Figure 1, a corresponding pair of points in Figure 3 shows the change in average delay for the AAL/EGF aircraft in the arrival rush, and for the non-AAL/EGF (ÒOthersÓ) aircraft in the arrival rush.Figure 3 shows that the change in delays due to priority scheduling varies with the position of the bank in the arrival rush, and that the greatest delay increase occurs for a bank which starts near the beginning of the arrival rush.This is attributable to the changing traffic density and traffic mixture in the arrival rush, and to the fact that all aircraft following the bank lead aircraft may be impacted by the reordering of the bank aircraft before scheduling.Because a larger number of aircraft may be impacted by the reordering, the aggregate increase in delays will be greater for a bank which begins earlier in the arrival rush.The average delay increase then diminishes as the ETA FG of the lead bank aircraft increases, and priority scheduling in some instances results in a slight decrease in average delay per aircraft.In these instances the priority schedule is actually more efficient than the FCFS schedule.The priority scheduling algorithm has the smallest impact on scheduling efficiency in regions where arrivals are not closely spaced and banks have non-AAL/EGF flights interspersed among the bank aircraft.Although a scheduling method which takes into account user preferences would ideally have no impact on scheduling efficiency when compared with FCFS scheduling, Figure 3 shows that for certain traffic conditions, the priority scheduling method results in little or no decrease in scheduling efficiency.v v v v v v v v v v v v v v v v v v vv v v vv v 0 0.5Any type of scheme which allows the introduction of user preferences into the arrival flow management process must ultimately be fair to all air carriers.In light of this, we are particularly interested in determining whether the priority scheduling of flights belonging to one airline disproportionately impacts the scheduled delays of aircraft belonging to other airlines.Examination of the delay increases for AAL/EGF flights in Figure 3 shows that for most of the banks, the delay increase for AAL/EGF flights in the arrival rush is greater than the delay increase for the non-AAL/EGF aircraft.By reordering only the aircraft belonging to the designated bank and scheduling all other aircraft according to an FCFS sequence, the impact of reordering on aircraft belonging to other airlines is minimized.This strategy also minimizes the impact of the reordering on scheduling efficiency, and in some instances results in improved efficiency by decreasing scheduled delays.The effects of AAR, bank size, and Center arrival error on the change in scheduled delays are also examined.For a given bank size and Center arrival error, when priority scheduling is used instead of FCFS scheduling, the change in average delay per aircraft tends to increase as AAR is increased.Results are similar to those seen in Figure 3 with the greatest change in delay occurring for banks which begin early in the arrival rush, and the change in delays decreasing for banks which are positioned later in the arrival rush.Increasing the magnitude of the Center arrival error for a given bank size and AAR substantially increases the change in delays for banks of aircraft arriving early in the rush period, while not significantly impacting the change in delay for banks arriving later in the traffic period.Finally, results show that the change in delays due to priority scheduling is largely unaffected by an increase or decrease in the size of the arrival bank.q q q q q q q q q q q q q q q q q q q q q q q q q v vv v v v v v v v v v v v v v v v v v v v v v v
+Concluding RemarksThis paper introduces the concept of priority scheduling as a means of taking into consideration airline arrival preferences in sequencing and scheduling algorithms for air traffic control automation.Priority scheduling is defined as a method of scheduling a bank of arrival aircraft according to a preferred arrival order instead of according to an FCFS order based on estimated time of arrival at the runway.A fast-time simulation originally developed for statistical evaluation of CTAS sequencing and scheduling algorithms has been modified for use in this investigation.Because this simulation does not provide any information about the controller workload required to meet the priority schedule, the output of the simulation is used only t o investigate the feasibility of the priority scheduling method in terms of scheduling efficiency and how closely the bankÕs scheduled arrival order matches the preferred arrival order.Results show that for the simulated traffic conditions, the priority scheduling algorithm results in a scheduled bank order which closely matches the preferred order.Results also show that when compared with FCFS scheduling, priority scheduling will, for certain traffic conditions, substantially reduce deviations from the preferred bank order while causing little or no decrease in scheduling efficiency.Figure 2 Figure 323Figure 1 Average order deviation
+Figure 33Figure 3 Change in average delays per aircraft when priority scheduling is used instead of FCFS scheduling
+Table 2 Actual arrival order2
+Table 4 Priority Sequence and resulting schedule4SequencePreferredFCFSPositionPriorityPositionNumberOrderScheduleShift forScheduleShift forinFCFSPriorityBankScheduleSchedule1AAL1150AAL11500AAL115002EGF628AAL1934-3AAL1934-13AAL1934AAL14281EGF62814AAL1428AAL15541AAL142805AAL1554EGF6281AAL15540
+Table 5 Calculation of position shift for a bank of aircraft5
+		
+		
+			
+FCFS SequenceResulting			
+			
+
+				
+
+
+	
+		ÒAirline Arrival Prioritization,Ó NAS Status Information Subgroup Memo
+		
+			ALacher
+		
+		
+			DBenfield
+		
+		
+		
+			May 19, 1997
+		
+	
+	Lacher, A., and Benfield D., ÒAirline Arrival Prioritization,Ó NAS Status Information Subgroup Memo, www.metsci.com/faa/cdm/nassi.html, May 19, 1997.
+
+
+
+	
+		Quarterly Update October - December 2015
+		
+			LBond
+		
+		10.1163/2210-7975_hrd-9806-2016016
+	
+	
+		Ó Journal of ATC
+		
+			
+			October -December 1997
+			Brill
+		
+	
+	Bond, L., ÒGlobal Positioning Sense II: An Update,Ó Journal of ATC, October - December 1997, pp. 51 -55.
+
+
+
+	
+		Initial Characterization of the 30 kW Miniature Arc Jet (mARC II) at NASA Ames Research Center
+		
+			FNeuman
+		
+		
+			HErzberger
+		
+		10.2514/6.2020-3108.vid
+	
+	
+		ÒAnalysis of Delay Reducing and Fuel Saving Sequencing and Spacing Algorithms for Arrival Traffic,Ó NASA TM 103880
+		
+			American Institute of Aeronautics and Astronautics (AIAA)
+			October 1991
+		
+	
+	Neuman, F. and Erzberger H., ÒAnalysis of Delay Reducing and Fuel Saving Sequencing and Spacing Algorithms for Arrival Traffic,Ó NASA TM 103880, October 1991, NASA Ames Research Center.
+
+
+
+	
+		
+			HErzberger
+		
+		
+			TJDavis
+		
+		
+			SMGreen
+		
+		ÒDesign of Center-TRACON Automation System,Ó Proceedings of the AGARD Guidance and Control Panel 56th Symposium on Machine Intelligence in Air Traffic Management
+		Berlin, Germany
+		
+			1993
+			
+		
+	
+	Erzberger, H., Davis, T. J., and Green, S. M., ÒDesign of Center-TRACON Automation System,Ó Proceedings of the AGARD Guidance and Control Panel 56th Symposium on Machine Intelligence in Air Traffic Management, Berlin, Germany, 1993, pp. 52-1 -52-14.
+
+
+
+	
+		ÒDesign and Operational Evaluation of the Traffic Management Advisor at the Fort Worth Air Route Traffic Control Center,Ó 1st USA
+		
+			HNSwenson
+		
+	
+	
+		Europe Air Traffic Management Research and Development Seminar
+		
+			June 17-19, 1997
+			Saclay, France
+		
+	
+	Swenson, H. N., et al., ÒDesign and Operational Evaluation of the Traffic Management Advisor at the Fort Worth Air Route Traffic Control Center,Ó 1st USA/Europe Air Traffic Management Research and Development Seminar, Saclay, France, June 17-19, 1997.
+
+
+
+	
+		ÒDesign Principles and Algorithms for Automated Air Traffic Managment,Ó AGARD Lecture Series No. 200, Knowledge-Based Functions in Aerospace Systems
+		
+			HErzberger
+		
+		
+			November 1995
+			San Francisco
+		
+	
+	Erzberger, H., ÒDesign Principles and Algorithms for Automated Air Traffic Managment,Ó AGARD Lecture Series No. 200, Knowledge-Based Functions in Aerospace Systems, San Francisco, November 1995.
+
+
+
+	
+		Delay exchanges in arriving sequencing and scheduling
+		
+			GregoryCarr
+		
+		
+			HeinzErzberger
+		
+		
+			FrankNeuman
+		
+		10.2514/6.1998-4478
+	
+	
+		Guidance, Navigation, and Control Conference and Exhibit
+		Boston, MA
+		
+			American Institute of Aeronautics and Astronautics
+			August 10-12, 1998
+		
+	
+	Carr, G. C., Erzberger, H., Neuman, F., ÒDelay Exchanges in Arrival Sequencing and Scheduling,Ó AIAA Guidance, Navigation, and Control Conference, Boston, MA, August 10-12, 1998.
+
+
+
+	
+		Enabling user preferences through data exchange
+		
+			StevenGreen
+		
+		
+			TsuyoshiGoka
+		
+		
+			DavidWilliams
+		
+		
+			StevenGreen
+		
+		
+			TsuyoshiGoka
+		
+		
+			DavidWilliams
+		
+		10.2514/6.1997-3682
+	
+	
+		Guidance, Navigation, and Control Conference
+		New Orleans, LA
+		
+			American Institute of Aeronautics and Astronautics
+			August 10-12, 1997
+		
+	
+	Green, S. M., Goka, T., Williams, D. H., ÒEnabling User Preferences Through Data Exchange,Ó AIAA Guidance, Navigation, and Control Conference, New Orleans, LA, August 10-12, 1997.
+
+
+
+	
+		ÒFast-Time Statistical Evaluation of Sequencing and Scheduling Algorithms for Multiple Runways
+		
+			FNeuman
+		
+		
+			HErzberger
+		
+		
+			MSchueller
+		
+		
+	
+	Ó to be published as a NASA technical memorandum
+	Neuman, F, Erzberger, H., Schueller, M., ÒFast-Time Statistical Evaluation of Sequencing and Scheduling Algorithms for Multiple Runways,Ó to be published as a NASA technical memorandum.
+
+
+
+	
+		Benefits analysis of terminal-area air traffic automation at the Dallas/Fort Worth International Airport
+		
+			MarkBallin
+		
+		
+			HeinzErzberger
+		
+		10.2514/6.1996-3723
+	
+	
+		Guidance, Navigation, and Control Conference
+		
+			American Institute of Aeronautics and Astronautics
+			July 1996
+		
+	
+	Ballin, M, and Erzberger, H., ÒAn Analysis of Landing Rates and Separations at the Dallas/Fort Worth International Airport,Ó NASA TM-110397, July 1996.
+
+
+				
+			
+		
+	

file118.txt

@@ -0,0 +1,310 @@
+
+	
+	
+		
+Nomenclature
+I. IntroductionNE of the largest challenges faced by the Air Traffic Management (ATM) community is the integration of new tools and concepts into the existing airspace system.These new concepts often need to be integrated with legacy systems.In the past few years much of the global ATM research community has proposed advanced systems based on Trajectory-Based Operations (TBO) 1 .The concept of TBO uses four-dimensional aircraft trajectories as the base information for managing safety and capacity.Both the US and European advanced ATM programs call for the sharing of the trajectory data between decision support tools for successful operations.However, the actual implementation of sharing trajectory information presents many challenges.Many advanced tools and concepts define functional and accuracy requirements for the trajectory predictor to meet their specific needs.These requirements can often be inconsistent or even conflicting across different systems.Two integration case studies, involving three systems, are discussed.These cases will focus specifically on the trajectory prediction functionality, a common feature for all three systems but very different in requirements and implementation.The first case study examines the issues with integration of the Efficient Descent Advisor (EDA), a tool which provides advisory conflict-free trajectories to meet a scheduled time, with the conflict detection/resolution functions of ERAM.These systems were developed completely independently.The integration of these tools is complicated by the fact that they perform similar functions but are driven by different requirements.The difference in the resulting trajectories can lead to conflicting advisories.The second case describes the issues with integrating the existing operational scheduler, the Traffic Management Advisor (TMA) with the EDA, which generates speed and altitude advisories to meet the TMA scheduled times.Both tools were originally developed by NASA from the same code baseline.It was anticipated that the integration of the tools would be simplified by their common source.However, recent efforts at NASA to integrate new concepts into the Operational TMA baseline, which has diverged significantly from the NASA research baseline, have shown that this is not the case.Difficulties have arisen due to the divergence of code and requirements development of the two tools.
+II. HistoryThere have been many studies about the synchronization of trajectory predictions between airborne avionics systems and ground-based systems. 2,3These studies show that performance of the ground-based trajectory predictors can be improved by the reception of aircraft state and intent data.Of less consideration has been the integration of different ground-based decision support tools that perform overlapping but not identical functions.This paper will examine issues with the operational integration of three such tools, ERAM, Efficient Descent Advisor (EDA) and Traffic Management Advisor.ERAM is the system that replaces the En-Route HOST computer.In this paper, the primary ERAM function of concern is the detection of potential conflicts and maintenance separation between aircraft in the en-route airspace 4 .TMA is a legacy system deployed and maintained by the Federal Aviation Administration (FAA) 5.6 .The purpose of TMA is to reduce congestion in the terminal airspace during high traffic periods.Predictions are made of when the aircraft will transition from the en-route to the terminal airspace.These predictions are used to create a sequence for the aircraft to enter the terminal area using a first-come, first-serve algorithm.Controllers use time of crossing into the terminal airspace to ensure appropriate spacing between arrivals.If necessary, aircraft are assigned some amount of delay in order to cross at their assigned time, which the controller may absorb by issuing speed changes, temporary altitudes, or flight path modifications.EDA was designed by NASA as an enhancement to the TMA system 7,8 .EDA provides the controller with advisories on how to meet the crossing time assigned by TMA.These advisories specify trajectories that meet the assigned time while minimizing fuel usage.The integration of the TMA and EDA systems would seem to be fairly straightforward as both systems originated from NASA and were built on a common software baseline.The TMA software was delivered to the FAA over 15 years ago.Initially, an effort was undertaken to synchronize the two code baselines so enhancements and code corrections could be shared.This became burdensome as some of the fixes and requirements made to support the operational system were incompatible with changes or functionalities for the research baseline and vice versa.The decision was made to abandon joint development resulting in the divergence of the operational and research baselines.During the period of divergence many of the prototypical algorithms were removed from the code baseline in order to reduce code complexity and minimize risk of unintended behavior.The research prototype meanwhile underwent several efforts of refactoring in order to better support EDA and other topics under investigation.At this point, these must really be considered to be two separate systems.Although the two code bases started with the same aircraft models and equations of motion, the outcome of these years of divergence is that the two code bases have very significant differences in the parsing of constraints and intent.Consequently, the trajectory requests still share a common data structure, but the resulting trajectories may look significantly different.This becomes problematic for the integration of EDA and TMA if the TMA system provides a scheduled time that is unattainable by EDA's methods of calculation.Lockheed Martin is the contractor tasked by the FAA to enable EDA in the ERAM architecture.As the current custodians of the software, they have been enhancing the functionality of the ERAM predictor to improve performance in the transition airspace and to introduce the EDA concepts into their legacy system.However, there are significant differences in aircraft modeling and the equations of motion used, as the two systems were developed independently.The complexity of integration of these two systems in not unexpected.
+III. IssuesThe problems faced in integrating EDA into the ERAM/TMA environment are fairly common in the development of large software systems.Many of the issues that arise when running these systems in an integrated environment is that while the tools have overlapping functions, several underlying requirements are incompatible.ERAM detects conflicts based on its best estimation of the trajectory the airplane will fly, using all available intent information and historical profiles modified by observed flight performance in the adapted airspace.In contrast, EDA and TMA base their advisories on fuel-optimal profiles that the airline would prefer to fly and more detailed models of aircraft performance.These tools use airline or aircraft manufacturer data to select speeds to descend at the latest possible point and still make the crossing restriction at the transition point into the terminal airspace.This can lead to very different shapes and flight times between ERAM and EDA/TMA trajectories.Another significant issue in the maintenance of large-scale legacy software systems is ensuring that the validation process adequately tests that the performance meets the specified requirements for a project, particularly when those requirements may evolve over time.The primary metric for validation of TMA was a time accuracy metric.However, analysis of software change requests shows correction of errors for large jumps in the estimation of the time of arrival as a flight progresses toward the meter fix as opposed to the accuracy of the original ETA.These jumps would cause issues in the scheduling algorithm and also reduce controller confidence in the predicted times.The testing procedures were not updated to reflect the addition of this "stability requirement."This stability requirement could complicate integration of new EDA functionality as some of these modifications modify or suppress changes to the trajectory shape or default speeds, which are degrees of freedom used by EDA to build advisory trajectories.The software architecture can be a hindrance to integrating new concepts into legacy systems.For ERAM, altitude restrictions are critical intent information that must be met as part of the generated profiles.Altitude restrictions are locations where aircraft must comply with a specific altitude for procedural separation and workload balance.These altitude restrictions have been implemented in the software in such a way that they cannot be ignored during a trajectory prediction.Conversely, EDA seeks to avoid the fuel penalty of the aircraft leveling off at the altitude restriction by providing advisories for an uninterrupted descent profile that meets the scheduled time.The trajectories generated do not account for the altitude restrictions in any condition.The EDA-generated trajectory may cause ERAM to detect a conflict, as the aircraft is no longer flying to the intent that ERAM expects.EDA advisories are vetted for conflicts but due to the different models used by the two systems, may not find the same results.At best, these inconsistencies could lead to unnecessary iteration and less than optimal descent profiles.These differences in software architecture will complicate merging the two algorithms.Ideally, the capabilities should all be functions of compatible requirements if not the same trajectory predictor, but how to integrate and validate the different algorithms so that the functionality and accuracy are sufficient for all applications is difficult.One of the difficulties in integration is that the boundary between the functionality of the trajectory predictor and that of the "client" application is drawn in different places. 10The Action Plan 16 group, founded by an initiative of the joint FAA/Eurocontrol Cooperative R&D Committee, developed a Common Trajectory Predictor (TP) structure to capture the essential components of a trajectory predictor 9 .A major point of contention in the development of this structure was the determination of which components were intrinsic to the trajectory predictor and which were functions of the decision support tool.Often, for convenience or improved performance, decision support tool functions are embedded in the trajectory prediction software.This can make integration into a single baseline difficult.The TP functions may expect specific data to be processed prior to being called, or the decision support tool or system may rely on the TP to perform certain functions specific and unique to its own need.One example of this would be in the parsing of the flight plan route provided by the aircraft.This route is comprised of a series of navigation fixes and airways, which must be decomposed into physical locations.For ERAM, this function is grouped into the trajectory prediction capability.For TMA and EDA, the decision support tool parses the route, as it may modify the fixes as part of its process in analyzing or advising the controller on the best ATC instructions to issue.EDA uses the lateral path as a degree of freedom for meeting a TMA secheduled arrival time, while TMA in some areas may modify the proscribed jet route in order to support requirements for getting traffic to the meter fix earlier (essentially to mimimize back-up during congestion).Vivona et al. 10 proposes separating these functionalities using the TP boundary rule in development of the software architecture.The TP Boundary Rule is "…if the capability directly supports a key function of the client application, then the capability is considered a client application capability and outside the scope of the TP."However, refactoring two large legacy systems, while necessary, would be time consuming and expensive.In 2010, a comparison of the functional requirements that drove the development of the trajectory predictors for ERAM Release 1 and EDA/TMA was conducted by NASA and Lockheed Martin 11 .This study compared the tools for the aircraft behavior modeled and the mathematical assumptions that were used to calculate the trajectory based on those behaviors.It was critical to break down the comparison into these two factors as the former determines which profiles could be processed and the latter accounts for the differences in profiles.As illustrated in the paper and Table 1, both EDA and ERAM handle similar types of altitude constraints.
+IV. Case StudiesEDA • • • • ERAM r1 • • •It is in the handling of the constraints where the difference emerges (Table 2).ERAMr1 uses an empirically based model of aircraft speed and vertical rate as the best approximation of speed in an environment where the speeds are unknown.The empirical model is based on analysis of thousands of trajectory histories at a particular site, where speed and altitude profiles are averaged over many operational conditions.EDA, conversely, uses a speed schedule of mach and CAS as these are the values that would be modified as advisories to meet a time, along with a high fidelity model of the aircraft performance for that specific type of aircraft and a nominal model of expected pilot behavior, assuming no intermediate restrictions from controllers prior to the TMA meter fix.ERAM handles additional functionality of waypoint-defined "AT or BELOW" constraints, where the constraint is a limit rather than an exact target state.These could be of great use to EDA but would require additional integration efforts to minimize possible negative effects such as ETA jumping when iterating to meet a TMA advised time.
+B. Integration of EDA and TMAThe change in the operational TMA system that would have a great impact on the EDA system is the modification of nominal speed profile used for the calculation of descents.The nominal speed value was originally selected based on typical airline and aircraft manufacturers' preferences.However, in response to an FAA requirement to increase the number of aircraft at the beginning of each arrival rush (known as "front loading" to increase pressure on the runways), the operational TMA system was modified to increase the indicated nominal descent speed.Doing so produced the desired earlier scheduled times of arrival, but had several potential unintended consequences.First, these speed changes can degrade the accuracy of the trajectory prediction made by TMA.For example, the nominal speed for A320 aircraft flying to LAX used by the TMA system is given as 310 kts.Fig. 1 shows estimated calibrated airspeeds flown by aircraft in descent into LAX.TMA and EDA model the descent of a jet such as the A320 using a speed profile of a constant mach speed segment (used for acceleratingto a descent speed higher than the cruise speed, if necessary) followed by a constant CAS segment to the arrival fix.This constant CAS segment can be viewed in Fig. 1 as the approximately vertical values from 25,000 to 10,000 ft.As can be seen, for this day the tendency for aircraft of this type was to fly in the constant CAS portion in the range of 250-300 kts.Using the higher values as a nominal descent speed can also increase the risk that the software will be unable to calculate the trajectory that can meet all the aircraft constraints.In both these cases, the FAA has validated that these risks do not have a significant effect on the performance of TMA.These effects would have more impact on the EDA systsm.Both the trajectory accuracy and failure issues could cause the iteration algorithm used by EDA to perform suboptimally.Most significantly, the increase of speeds removes a degree of freedom from EDA as there would be less ability to use faster speed to meet scheduled times.
+V. ConclusionIntegration of new concepts into legacy systems will always be challenging.Even systems that were created from a common software baseline may not be trivial to recombine after a significant period of isolated development.The trajectory prediction functions in EDA, ERAM and TMA systems have all been developed separately to meet the needs of their specific system.The trajectory predictors are composed of many optimizations to meet their project's requirements that should not be lost in the process of integration.The issues with integration of the trajectory predictors could be addressed by longer-term consideration of future requirements and how they would be implemented.These do add complexity to the code baseline.Uncertainty in the acceptance of the new capability also discourages retention of the additional functionality.Since many of the tools are deployed and maintained by different organizations and contractors during different timeframes, it is challenging to find and maintain commonality between the tool's trajectory predictors.Still, many of these advanced capabilities can be maintained with low risk to the performance of the initially implemented systems and save considerable money and development time in the future.Both case studies illustrate the need for clear, consistant and cross-comparible TP requirements, developed at as early a stage as possible.It would be difficult to build a single trajectory predictor which could meet the requirements of all systems.If the requirements were written in a consistant format, this could drive the code implementation to find common structures and minimize architectural differences.Similarly, consistant application of the TP boundary rule would enable more common TP structure between systems, allowing the TPs to function interchangeably.A.Integration of ERAM and EDA Downloaded by NASA AMES RESEARCH CENTER on August 29, 2013 | http://arc.aiaa.org| DOI: 10.2514/6.2013-443
+Figure 1 .1Figure 1.Calibrated airspeeds for Airbus 320 aircraft flying into LAX over a 24-hour period, 2011.
+Table 1 : Vertical Constraints Handled TP Cruise Altitude1Departure/Arrival SpeedInterimLimitTransition Alt (altimeterAltitudeAltitudesetting)
+Table 2 : Vertical Speed Models TP Descending Segments2Level Segments
+			Downloaded by NASA AMES RESEARCH CENTER on August 29, 2013 | http://arc.aiaa.org| DOI: 10.2514/6.2013-443
+		
+		
+
+			
+AcknowledgmentsThe author would like to thank Robert Vivona and Gabriele Enea of the Engility Corporation for the analyses they have conducted on the Operational TMA baseline.
+			
+
+			
+
+				
+
+
+	
+		NextGen next generation air transportation system: NextGen policy issues
+		10.1109/icnsurv.2011.5935406
+		
+	
+	
+		2011 Integrated Communications, Navigation, and Surveillance Conference Proceedings
+		
+			IEEE
+			June 2007
+		
+	
+	Joint Planning and Development Office: Concept of Operations for the Next Generation Air Transportation System, Version 2.0, June 2007 Available for public download from http://www.jpdo.gov/library/NextGen_v2.0.pdf.
+
+
+
+	
+		Air-ground trajectory synchronization — Metrics and simulation results
+		
+			DavidS KChan
+		
+		
+			GlenWBrooksby
+		
+		
+			JoachimHochwarth
+		
+		
+			JoelKKlooster
+		
+		
+			SergioTorres
+		
+		10.1109/dasc.2011.6095977
+	
+	
+		2011 IEEE/AIAA 30th Digital Avionics Systems Conference
+		Seattle, Washington
+		
+			IEEE
+			October 2011
+			
+		
+	
+	Chan, D.S.K., Brooksby, G.W., Hochwarth, J., Klooster, J., and Torres, S., "Air-Ground Trajectory Synchronization --Case Studies and Metrics", 30th Digital Avionics Systems Conference, Seattle, Washington, 16-20 October 2011
+
+
+
+	
+		Trajectory Synchronization between air and ground trajectory predictors
+		
+			SergioTorres
+		
+		
+			JoelKKlooster
+		
+		
+			LilingRen
+		
+		
+			MauricioCastillo-Effen
+		
+		10.1109/dasc.2011.6095978
+	
+	
+		2011 IEEE/AIAA 30th Digital Avionics Systems Conference
+		Seattle, Washington
+		
+			IEEE
+			October
+			
+		
+	
+	Torres, S., Klooster K. J., Ren, L., and Castillo-Effen, M., "Trajectory Synchronization between Air and Ground Trajectory Predictors", 30th Digital Avionics Systems Conference, Seattle, Washington, 16-20 October
+
+
+
+	
+		Evaluation of Prototype Enhancements to the En Route Automation Modernization's Conflict Probe
+		
+			ACrowell
+		
+		
+			AFabian
+		
+		
+			CYoung
+		
+		
+			BMusialek
+		
+		
+			MPaglione
+		
+		
+		
+			2011
+		
+	
+	Crowell, A., Fabian, A., Young, C., Musialek, B. and Paglione, M., "Evaluation of Prototype Enhancements to the En Route Automation Modernization's Conflict Probe," DOT/FAA TC-TN12/3, 2011
+
+
+
+	
+		En route Descent Advisor concept for arrival metering
+		
+			StevenGreen
+		
+		
+			RobertVivona
+		
+		10.2514/6.2001-4114
+		AIAA-2001-4114
+	
+	
+		AIAA Guidance, Navigation, and Control Conference and Exhibit
+		Montreal, Canada
+		
+			American Institute of Aeronautics and Astronautics
+			Aug. 2001
+		
+	
+	Green, S. M., and Vivona, R. A., "En route Descent Advisor Concept for Arrival Metering," AIAA-2001-4114, AIAA Guidance, Navigation, and Control Conference, Montreal, Canada, Aug. 2001.
+
+
+
+	
+		Design and Development of the En Route Descent Advisor (EDA) for Conflict-Free Arrival Metering
+		
+			RichardCoppenbarger
+		
+		
+			RichardLanier
+		
+		
+			DougSweet
+		
+		
+			SusanDorsky
+		
+		10.2514/6.2004-4875
+		AIAA-2004-4875
+	
+	
+		AIAA Guidance, Navigation, and Control Conference and Exhibit
+		Providence, RI
+		
+			American Institute of Aeronautics and Astronautics
+			Aug. 2004
+			
+		
+	
+	Coppenbarger, R. A., Lanier, R., Sweet, D., and Dorsky, S., "Design and Development of the En Route Descent Advisor (EDA) for Conflict-Free Arrival Metering," AIAA-2004-4875, AIAA Guidance, Navigation, and Control Conference, Providence, RI, 16-19 Aug. 2004.
+
+
+
+	
+		The Traffic Management Advisor
+		
+			WilliamNedell
+		
+		
+			HeinzErzberger
+		
+		
+			FrankNeuman
+		
+		10.23919/acc.1990.4790788
+	
+	
+		1990 American Control Conference
+		San Diego, CA
+		
+			IEEE
+			1990. May 1990
+		
+	
+	Nedell, W., and Erzberger, H., "The Traffic Management Advisor," 1990 American Control Conference, San Diego, CA, May 1990.
+
+
+
+	
+		Design and Operational Evaluation of the Traffic Management Advisor at the Fort Worth Air Route Traffic Control Center
+		
+			HNSwenson
+		
+		
+			THoang
+		
+		
+			SEngelland
+		
+		
+			DVincent
+		
+		
+			TSanders
+		
+		
+			BSanford
+		
+		
+			KHeere
+		
+	
+	
+		1st USA/Europe Air Traffic Management R&D Seminar
+		
+			June 1997
+			Saclay, France
+		
+	
+	Swenson, H. N., Hoang, T., Engelland, S., Vincent, D., Sanders, T., Sanford, B., and Heere, K., "Design and Operational Evaluation of the Traffic Management Advisor at the Fort Worth Air Route Traffic Control Center," 1st USA/Europe Air Traffic Management R&D Seminar, Saclay, France, June 1997.
+
+
+
+	
+		Action Plan 16 Common Trajectory Prediction Capability: Generic Trajectory Predictor Structure Available
+		
+			Faa/Eurocontrol
+		
+		
+		
+	
+	FAA/Eurocontrol Action Plan 16 Common Trajectory Prediction Capability: Generic Trajectory Predictor Structure Available for public download from http://acy.tc.faa.gov/cpat/tjm//.
+
+
+
+	
+		Abstraction Techniques for Capturing and Comparing Trajectory Predictor Capabilities and Requirements
+		
+			RobertVivona
+		
+		
+			StevenGreen
+		
+		
+			KarenCate
+		
+		10.2514/6.2008-7408
+	
+	
+		AIAA Guidance, Navigation and Control Conference and Exhibit
+		Honolulu, HI
+		
+			American Institute of Aeronautics and Astronautics
+			18-21 Aug. 2008. 11
+		
+	
+	Vivona, R. A., Cate, K. T., and Green, S. M., "Abstraction Techniques for Capturing and Comparing Trajectory Predictor Capabilities and Requirements," AIAA Guidance, Navigation, and Control Conference and Exhibit, Honolulu, HI, 18-21 Aug. 2008. 11
+
+
+
+	
+		Comparison of Aircraft Trajectory Predictor Capabilities and Impacts on Automation Interoperability
+		
+			RobertVivona
+		
+		
+			KarenCate
+		
+		
+			StevenGreen
+		
+		10.2514/6.2011-6856
+	
+	
+		11th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference
+		Virginia Beach, VA
+		
+			American Institute of Aeronautics and Astronautics
+			Sep. 2011
+			
+		
+	
+	AIAA-2011-6856, 11th AIAA Aviation Technology, Integration, and Operations Conference
+	Vivona, R. A., Cate, K., and Green, S., "Comparison of Aircraft Trajectory Predictor Capabilities and Their Impacts on Air Traffic Management Automation Interoperability," AIAA-2011-6856, 11th AIAA Aviation Technology, Integration, and Operations Conference, Virginia Beach, VA, 20-22 Sep. 2011.
+
+
+				
+			
+		
+	

file120.txt

@@ -0,0 +1,261 @@
+
+	
+	
+		
+IntroductionThere is a need to integrate the proposed increase of new entrants and their diverse missions into the National Airspace System (NAS), manage the corresponding expected increase in traditional operations, and enable enhanced collaboration between users and the Federal Aviation Administration (FAA) 1 .Some new entrant operations (e.g., large Unmanned Aerial Systems (UAS) and commercial space) have been active in the airspace for several years but more recently there is a significant movement to enable the large scale use of small electric Vertical Take-Off and Landing aircraft (eVTOL) for rapid passenger movement in an urban environment 2 .The US airspace is already managing, on average, approximately 26,000 scheduled traditional commercial operations per day 3 and their delays cost the US approximately $26.6B in 2017. 4The inclusion of new entrant operations will likely increase costs due to delays as more users access the airspace.To address these issues, NASA's Aeronautics Research Mission Directorate (ARMD) has defined a "pivot" to address these transportation challenges and recently approved the Air Traffic Management -eXploration (ATM-X) project to achieve the goals of equitable access to the airspace for all users, vehicles, and missions while also improving current operations.
+II.Project Description ATM-X is one of four projects in NASA's Airspace Operations and Safety Program (AOSP) within ARMD.The ATM-X technical goals, to be realized over two phases, are to explore alternate methods to incorporate new technologies into the NAS using a software "service-oriented paradigm" based on Unmanned Aerial System (UAS) Traffic Management (UTM) principles. 5UTM principles allow for system and management flexibility whenever possible and imposing structure when required and includes seamless airspace access for all users, scalability for new demand and users, collaboration through digital information exchange, resilience to disruptions and uncertainty, and increased availability and use of services.Individual software services allow greater flexibility for airspace management support and potential for faster modernization, as the new algorithms are not forced to be intertwined with larger, centralized software systems.Breaking out algorithms, such as airborne weather routing, into a modular software service allows the algorithms to be developed and updated by third-parties with less modification to the larger system using a software service-oriented paradigm.This service-oriented paradigm is currently being developed and used by the UTM project to manage small UAS operations; an example of a field-tested capability for ATM-X research and a model for managing flight vehicles in field demonstrations.With a large set of challenges to enable access by the new entrant market, as well as to continue to improve traditional operations, Phase 1 of ATM-X (FY18-FY20) will focus on two use-cases supported by four sub-projects.The "Urban Air Mobility (UAM) Operations" use-case will address passenger-carrying eVTOL urban operations by performing research to understand how to develop verifiably-safe and secure airspace and vertiport management technologies to enable missions at a user-specified tempo in low altitude, controlled airspace.The second use-case, "Northeast Region Operations", will focus on improved collaboration, dynamic airspace access and planning, and integrated scheduling in the US Northeast Region for traditional NAS users.The following lists the objectives of the four ATM-X sub-projects that will address those two use-cases; the subprojects are described in more detail later in the document:
+ATM-X Subproject DescriptionsDuring Phase 1 research, the project will develop concepts and prototype technologies for improved data exchange and digital negotiation to enable traditional operators' desire for improved collaboration, throughput, flexibility and predictability and be extensible to allow new entrants' desire to access and operate in controlled airspace.The guiding principles of Phase 1 are "Explore, Build and Learn" to address these activities.There will be research in both UAM and traditional operations with appropriate prototype development and evaluations to learn what research and prototypes will be carried forward into Phase 2. To leverage the research of other projects, ATM-X will collaborate with NASA's AOSP System Wide Safety (SWS) Project to ensure the development of safe, secure, and efficient operations in both use-cases.The second Phase for ATM-X (approximately from FY21-FY26) will build towards longer-term goals, and address challenges and activities informed by Phase 1 results and evaluations.Phase 2 will mature the technologies and the service-oriented ATM system for a possible field demonstration of capabilities for routine access for a wide range of vehicles and missions.The relationship between the ATM-X sub-projects is illustrated in Error!Reference source not found.. Passenger carrying UAM operations are addressed in the Initial Urban Air Mobility Operations Integration sub-project and traditional operations in the Northeast Region are addressed by the Increasing Diverse Operations and Integrated Demand Management sub-projects.The arrows in this figure show representative work areas of these sub-projects that will be integrated and tested utilizing the Testbed.The Testbed will provide a realistic and flexible environment, allowing advanced concepts and the study of interaction across components, and will be critical in the development of a service-oriented ATM architecture.
+III. Development ApproachThe ATM-X project will conduct research towards the service-oriented paradigm by developing prototypes and focus on research capabilities that will significantly improve the safety, security, efficiency, and reliability of the future NAS.To address the safety aspects of the technologies, ATM-X will leverage the work in other AOSP activities including verification and validation of software systems in SWS and NASA's cybersecurity efforts.In addition to the research focus of Phase 1, ATM-X will develop prototype services and a reference implementation of a service-oriented architecture design for technical evaluations that leverages recent work by the UTM project.Additionally, there will be opportunities to identify technologies for more mature development in Phase 2 and evaluate technologies collaboratively in this phase.To guide the development, the key principles of ATM-X are presented below, derived from the UTM principles described earlier.These ATM-X principles are traceable in the research and resultant technologies.• Seamless access to the airspace for both on-demand and scheduled operations;• Scalability to match future air traffic demand where and when it occurs;• System and management flexibility whenever possible and imposing structure when required;• Collaboration between all participants through secure, integrated information sharing; • Resilience to uncertainty, degradation and disruptions; and • Availability of user and third-party services.The two use-cases are:• UAM Operations: An extension to current passenger carrying helicopter operations, but with a higher tempo and a denser network, using radically different vehicles being proposed by a wide range of new and existing aviation companies.This use-case is part of the larger UAM concept.NASA defines UAM as a safe and efficient system for air passenger and cargo transportation within an urban area.It is inclusive of small package delivery and other urban UAS services and supports a mix of onboard/ground-piloted and increasingly autonomous operations 6 .These passenger carrying vehicles would conduct high-frequency, short-distance flights between fixed locations, through dense urban centers and can also provide shortdistance movement of goods.Most vehicles will be electrically powered and carry two to six passengers, or equivalent cargo, on flights of 10-70 miles.A mature vision of these operations will generally be on-demand with limited planning and may not necessarily follow pre-approved routes.However, near-term implementations of these operations may employ schedules using fixed routes and some routine short-cuts.• Northeast Region Operations: Research focused initially on enhancing the efficiency and robustness of traditional airspace operations in the complex, highly-controlled airspace in the northeast of the U.S. For years, the FAA and NASA have focused on developing, testing, and deploying new capabilities in less complex operational airspaces to enhance efficiency with reduced risk.However, these new capabilities have not migrated to this more complex area due to a mix of technological and procedural barriers.This research will address setting the stage for seamlessly and equitably integrating new vehicle types (increasingly diverse operations), such as large UAS, autonomous freighters, supersonic aircraft, and UAM into these operations.Through the RTCA NextGen Advisory Committee (NAC) 7 , industry has identified this region as a critical choke-point.The ATM-X effort supports FAA technology development activities and priorities to improve traditional operations in the same geographic area. 8,9ile the two use cases focus on different types of airspace users, both use cases will ultimately address setting the stage for seamlessly and equitably integrating newer non-traditional vehicle types (increasingly diverse operations), such as large UAS, autonomous freighters, and UAM vehicles into these operations.
+IV.Sub-project Descriptions Figure 1 depicted the four sub-projects in ATM-X.The UAM sub-project focuses on the emergent users for a representative future operation, and will conduct research, develop and evaluate technologies, and explore architectures for airspace and vertiport management to enable safe UAM missions at a user-specified tempo.The first use case will serve as a concept focus for this research.IDM and IDO focuses on the needs of traditional airspace users along the lines of the second use case.IDM will develop a concept that utilizes state-of-the-art capacity, demand, and weather forecasts in a coordinated fashion across different traffic flow management capabilities to better manage demand/capacity imbalance under adverse weather conditions.IDO will build on IDM progress, along with evaluating how NASA's Airspace Technology Demonstration (ATD) capabilities and others can be applied, to evaluate digital negotiation using trajectory-based technologies in the Northeast Region.The research will investigate improved user collaboration to enable traditional operators' desire for flexibility and predictability.IDO will explore how to transition the existing system to a future service-based architecture.Testbed will develop a capability to simply and easily connect high-fidelity human-in-the-loop (HITL) and automation-in-the-loop simulations and tests.Testbed will support the evaluation of trajectory-based automation and electronic negotiation, collaborative decision making, and connected service-based technologies.Testbed will also provide connectivity between operational FAA, conceptual NASA and other proposed systems.
+A. Initial Urban Air Mobility Operations Sub-ProjectNASA has conducted research addressing a wide range of air traffic management challenges for traditional aircraft since the 1980s.The concepts, technologies, and procedures developed through these efforts have benefited the flying public and the aviation community in the form of more efficient and predictable operations.Recently, NASA has increased focus on developing technologies and standards for new types of aircraft and missions being pursued by the broader aviation community.One notable example is NASA's ongoing effort as part of RTCA Special Committee-228 to develop minimum operational performance standards (MOPS) for detect and avoid (DAA) capabilities for large UAS.As part of this endeavor, NASA helped develop standards for separation, alerting, guidance, surveillance, and displays for DAA systems that the FAA will use to develop technical standards and regulations.NASA has also developed and demonstrated UTM capabilities for low-altitude small UAS (sUAS) operations.More recently, there has been an increased interest in UAM.Significant industry investments have been made toward developing an ecosystem for UAM that includes manufacturers of eVTOL aircraft and builders of vertiports and other infrastructure on the ground.To conduct UAM operations in a safe, secure, and efficient way in the presence of existing airspace users, tools and methods for airspace integration are needed.Many of the technologies and procedures that have been developed to integrate new entrants, such as large and small UAS into the airspace with existing traditional aviation could be applicable to UAM operations and will be leveraged in this sub-project.The Initial UAM Operations Integration sub-project is focused on the low altitude (e.g., below about 2,000 ft) airspace integration aspect of UAM, both to enable early entrants in the airspace and to identify, develop, and evaluate the services, procedures, and tools necessary to support high-demand, mature operations.This sub-project will collaborate closely with other NASA projects to address safety and security issues.Other aspects of UAM, such as vehicle development, battery technology development, ground infrastructure construction, and legal considerations are also important but will be addressed by other NASA projects and external organizations.This sub-project will work with an ARMD level UAM coordination team and support their activities with partners.Phase 1 research will focus on airspace management and safe, secure, and efficient operations into, out of, and within an urban area.The project will support evaluations and field demonstrations of a wide range of technologies, requirements, and procedures, including but not limited to:• Safe mission planning and operations • Noise constraint management • Secure data exchange system architecture • Communications, navigation, and surveillance • Separation assurance • Dynamic scheduling, sequencing, and spacing • Congestion management • Interoperability with other vehicles -large UAS, sUAS, and traditional aircraft operations In addition, a collaborative concept of operations in which the roles, responsibilities, and interactions of pilots, vehicle automation systems, air traffic control, existing airspace users, safety and security systems, and airspace management automation systems will also be defined.The concept of operations serves as a framework for NASA research and development efforts, lab simulations, and live flight demonstrations planned in this first phase.It also provides a benchmark that the UAM community can use to determine technology development priorities, requirements for infrastructure improvements, and achievable airspace capacities.Table 1 shows some of the UAM airspace management technologies that will be evaluated with partners in Phase 1. FY2018 will provide an initial capability that will be matured through evaluations in subsequent years.
+FY2018Initial weather and obstacle aware separation, scheduling, sequencing, and spacing algorithms FY2019Safe mission planning and operations, weather and obstacle-aware software services and operational procedures FY2020 Separation and scheduling, sequencing, and spacing software services Table 1 Representative UAM activities for Fiscal Year (FY) 2018 -2020
+B. Increasing Diverse Operations (IDO) Sub-ProjectNASA has nearly three decades of experience working with the FAA and airspace users in conducting research, technologies and concepts to improve traditional user operations.This sub-project leverages prior developments by NASA and the FAA to continue to push the state-of-the-art in airspace management.To achieve long-term goals, IDO will define an overarching concept/architecture of a service-based system and develop supporting service technologies to enable safe and secure increasingly diverse operations in dense, controlled airspace.These diverse operations will span a range of vehicle performance and design, mission type, and equipage levels.New entrants (such as supersonic, space launch and re-entry, high-altitude long endurance, UAS, and UAM vehicles) are expected to introduce the greatest diversity to operations as their demand for entering controlled airspace increases.However, it is just as important to ensure that the future airspace systems continue to accommodate traditional users while providing appropriate levels of safety and security.During Phase 1, IDO will define a service-based airspace system that improves efficiency and predictability for traditional users but also prepares the system for increased diversity from new entrants.IDO will evaluate the applicability of NASA's ATD capabilities and others that are built on principles of information sharing and time-based scheduling.These capabilities can be integrated or coordinated to evaluate improvements to gate-to-gate operations incorporating user priorities of traditional airspace users.NASA's ATD project encompasses a collection of critical technology development and demonstration activities that addresses near-term, domain trajectory-based operations and provides benefits to traditional air transportation system stakeholders. 10IDO will develop a concept of operations and accompanying system architecture that evaluates the integration of ATD and other capabilities for a future service-oriented airspace system.This concept of operations will be evaluated in a HITL simulation using Northeast Region scenarios.The scenarios will be developed with FAA and airline partners where recent discussions have included dynamic scheduling of metroplex operations in the New York area and integration of arrival scheduling with pre-departure scheduling and en-route routing.This sub-project seeks to improve operations by exploring the enhancement of software airspace management services described by the concept of operations.For example, existing time-based scheduling services may be enhanced to subscribe to a repository of flight trajectories and constraints to increase coordination.Controller tools may be enhanced to provide better awareness of FAA and user coordinated strategic flight trajectories and ensure that control actions dynamically respond to and reinforce, rather than compromise, the strategic plan.A digital negotiation service may be developed to standardize and speed up routine trajectory negotiations currently performed manually and often in an inconsistent manner.Table 2 shows representative IDO activities and a series of evaluations based on a partner informed Northeast Region scenario.The culminating event is a collaborative high-fidelity simulation utilizing the Testbed.
+FY2018Northeast Region simulation scenario identification and definition FY2019 Development of prototype services and procedures for evaluation FY2020High fidelity simulations in Testbed with external partners to evaluate services Table 2 Representative IDO activities for FY (Fiscal Year) 2018 -2020
+C. Integrated Demand Management (IDM) Sub-ProjectIDM was previously a subproject in the recently-completed Shadow Mode Assessment using Realistic Technologies (SMART-NAS) for Safe Trajectory Based Operations 11 project and continues in ATM-X.This sub-project explores operational integration of near-to mid-term NextGen traffic management capabilities to improve NAS performance when the capacity of critical airspace resources is inadequate to meet demand using current systems.Because of the interactions between multiple flights across multiple airspace constraints, successful trajectory-based operations involve coordinated traffic flow management of constraints, as well as trajectories, to provide a continuous, robust solution for a collection of flights.However, in today's air traffic operations, a number of different, uncoordinated and locally-focused systems are involved in managing these interactions, resulting in inefficiencies in flight trajectories as well as inadequate demand / capacity balance.In order to alleviate this problem in the near-to mid-term NextGen timeframe, IDM leverages two main traffic flow management capabilities in the National Airspace System, namely the FAA's Traffic Flow Management System (TFMS) and Time Based Flow Management (TBFM).In the IDM concept, TFMS tools are used to pre-condition traffic into the more tactical TBFM system, enabling TBFM to better manage delivery to the capacity-constrained destination.IDM is focused on coordinating the management strategies employed by the TFMS.TFMS strategically manages aircraft at the origin airports when the demand is expected to exceed capacity at their destination.TBFM tactically manages demand of the actual flow of aircraft near the destination airport.Currently, TFMS and TBFM work largely independently causing additional delay and inefficiencies in operations.The IDM concept also provides a framework to take advantage of many of the past and current NASA ATM research, which developed powerful, integrated operations / tools for managing trajectory constraints, leveraging existing systems, and adding new automation tools / methods where needed.These solutions are complementary, with each focused on a specific portion of the complete flight trajectory.They represent crucial building blocks towards a gateto-gate management solution.IDM provides an integrated solution across the domains to enable improved system performance.The IDM team conducted HITL evaluations of its concepts focused on arrivals to the Newark Liberty Airport (EWR) 12 .Results show the concept achieved the target throughput while minimizing the expected cost associated with overall delays in arrival traffic.Future work based on these results will evaluate the concept for the LaGuardia (LGA) airport with combined EWR and LGA operations and multiple constraints.These evaluations will comprise a combination of fast-time and HITL simulations.As a result of this research and testing, a description of the IDM concept and procedures plus tool specifications will be delivered to the FAA.Table 3 shows representative IDM activities.This shows a series of increasingly complex simulations to evaluate the evolving IDM concept for multiple airports and constraints.
+FY2018Initial report on IDM concept, procedures and tools requirements based on HITL results and stakeholders' feedback FY2019Conduct HITL experiment to evaluate IDM concept in convective weather for two or more airports FY2020Conduct HITL to evaluate final IDM concept for multiple constraints / airports Table 3 Representative IDM activities for FY (Fiscal Year) 2018 -2020
+D. Testbed Sub-ProjectThe Testbed was formerly known as the SMART-NAS Testbed in the SMART-NAS for Safe Trajectory Based Operations Project. 13Testbed will be capable of integrating FAA operational systems with prototype technologies or services and allow new concepts to be evaluated in a realistic environment.The Testbed's role is as an accelerator of concepts and technology development, with a use-case-driven development approach to support NASA research as well as collaborations.The Testbed development was motivated by a survey of simulation capabilities that highlighted significant needs and limitations of existing in simulation systems.These limitations include:• Tools and systems are rarely integrated across different ATM domains During Phase 1, the Testbed and other NASA teams will develop core capabilities, such as:• Back-end, Big-Data analytics tools to generate realistic simulation scenarios using NASA's ATM Sherlock data warehouse 15 • Cloud technologies to securely, reliably, and cost-effectively connect distributed, NASA and non-NASA, real and simulated NAS infrastructure and flight assets • Low-maintenance mechanisms to integrate a wide array of simulation assets without customized one-toone solutions • Ability to create an on-demand, shadow-mode NAS, high-fidelity evaluation from live traffic, weather data, and airspace information • Ability to evaluate collaborative flight management concepts and technologies over an entire flight profile • In-time system-wide safety analytics and key performance indicators to visualize simulation and operational performance • Ability to assess the impact of new vehicle designs on NAS operations The Testbed core capabilities will be developed through a series of two software builds.Build 1 follows the prototype development begun under the SMART-NAS project and focuses on the development of the primary scenario building, cloud-based connectivity and component communications for use by internal NASA stakeholders.Build 2 matures these technologies and develops the required interfaces to connect external stakeholders, thus supporting the Testbed objectives.Towards the end of Phase 1, ATM-X will provide a plan that includes community involvement in both the development, support and distribution of Testbed.Representative Phase 1 work is shown in Table 4.
+FY2018Establish connections to NASA ATM tools and initial connections to partner systems FY2019 Support NASA and partner activities and system evaluations using a service-oriented system FY2020 Expanded activities and system evaluations using improved connectivity, scalability, and usability to external stakeholder systems Table 4 Representative Testbed activities for FY (Fiscal Year) 2018 -2020 V. Summary The ATM-X project is a two phased project to integrate new, diverse entrants into the NAS, while also leveraging NASA's prior ATM achievements that continue to improve traditional airspace operations.In the first phase, ATM-X will collaboratively develop ATM services and evaluate them in HITL simulations and field activities to inform the work for Phase 2. The evaluations are based on and in support of two partner defined use-cases for UAM and traditional Northeast Region operations and will adhere to principles of seamless access, and scalable, flexible, resilient, available, and collaborative operations.Phase 2 will further improve upon the relevant research and services identified and developed in Phase 1 for a possible demonstration of diverse operations.To accomplish Phase 1, ATM-X is structured with four sub-projects: UAM sub-project, IDM, IDO and Testbed.The UAM sub-project will conduct research, develop and evaluate technologies, and explore architectures for airspace and vertiport management to enable safe UAM missions at user-specified tempo.IDM and IDO initially focuses on the needs of traditional airspace users.IDM will develop a concept that utilizes state-of-the-art capacity, demand, and weather forecasts in a coordinated fashion across different traffic flow management capabilities to better manage demand/capacity imbalance under adverse weather conditions.IDO will build on IDM progress, along with evaluating how NASA's ATD capabilities and others can be applied, to evaluate digital negotiation using trajectory-based capabilities in the Northeast Region with improved user collaboration to enable operators' desire for flexibility and predictability.Testbed will continue developing a capability to simply and easily connect high-fidelity HITL and automation-in-theloop simulations, supporting the evaluation of electronic trajectory negotiation, collaborative decision making, and connected trajectory-based technologies.Testbed will also provide connectivity between operational FAA, conceptual NASA and other proposed systems.Fig. 11Fig. 1 Relationship of the Four Sub-Projects in Phase 1.
+Table 11Initial Urban Air Mobility Operations IntegrationConduct research and develop technologies for(UAM sub-project)airspace and vertiport management towards enablingUAM missions at user-specified tempoIncreasing Diverse Operations (IDO)Evaluate digitally-negotiated trajectory-basedcapabilities research that includes leveraging NASA'sAirspace Technology Demonstration technologies inthe Northeast Region with improved user collaborationto enable traditional operators' desire for flexibility andpredictabilityIntegrated Demand Management (IDM)Develop Integrated Demand Management conceptusing coordinated traffic flow management capabilitiesto better manage demand/capacity imbalance duringadverse weatherTestbedDevelop a capability to simply and easily connect high-fidelity simulations to support NASA and communityresearch
+•NASA, the FAA and ATM Industry lack an integrated development, test, and evaluation platform for enabling collaboration across disparate government and industry-developed ATM systems • Simulation preparation and execution is time-consuming, resource intensive, error-prone and limited by the capabilities of individual facilities • Simulations are traditionally comprised of individual brick-and-mortar labs or co-located facilities • Live flight assets and enterprise level hardware and software are not accessible for many efforts seeking to advance ATM technology • Many future concepts are untestable with current systems as documented in the National Research Council Autonomy Report, 2014 14
+		
+		
+			
+
+				
+
+
+	
+		NAS data release policy: Challenges & opportunities
+		
+			JamesEck
+		
+		10.1109/icnsurv.2010.5503288
+	
+	
+		2010 Integrated Communications, Navigation, and Surveillance Conference Proceedings
+		
+			IEEE
+			2016
+		
+	
+	Eck, J., "The Future of the NAS." FAA, 2016
+
+
+
+	
+		Enabling Airspace Integration for High-Density On-Demand Mobility Operations
+		
+			EricRMueller
+		
+		
+			ParmialHKopardekar
+		
+		
+			KennethHGoodrich
+		
+		10.2514/6.2017-3086
+		
+	
+	
+		17th AIAA Aviation Technology, Integration, and Operations Conference
+		
+			American Institute of Aeronautics and Astronautics
+			June, 2017. 14 April 2018
+			
+		
+	
+	Mueller, E., Kopardekar, P., and Goodrich, K., "Enabling Airspace Integration for High-Density On-Demand Mobility Operations," 17th AIAA Aviation Technology, Integration, and Operations Conference, 5-9 June, 2017. 3 https://www.faa.gov/air_traffic/by_the_numbers/ [cited 14 April 2018]
+
+
+
+	
+		7. Building a More Balanced Airline Industry
+		
+			AmericaAirlines
+		
+		10.7591/9780801458330-008
+	
+	
+		Up in the Air
+		
+			Cornell University Press
+			April, 2018
+			
+		
+		
+			U.S. Airline Industry Review
+		
+	
+	Airlines for America, "U.S. Airline Industry Review: Allocating Capital to Benefit Customers, Employees and Investors," April, 2018
+
+
+
+	
+		Unmanned Aircraft System Traffic Management (UTM) Concept of Operations
+		
+			PKopardekar
+		
+	
+	
+		16th AIAA Aviation Technology, Integration, and Operations Conference
+		
+			June, 2016
+			
+		
+	
+	th
+	Kopardekar, P., et al., "Unmanned Aircraft System Traffic Management (UTM) Concept of Operations," 16th AIAA Aviation Technology, Integration, and Operations Conference, 13 -17 th , June, 2016
+
+
+
+	
+		Priorities for Improving Operational Performance in the Northeast Corridor through CY2021
+		
+			Rtca
+		
+	
+	
+		NextGen Advisory Committee
+		
+			March, 2018
+		
+	
+	RTCA, "Priorities for Improving Operational Performance in the Northeast Corridor through CY2021," NextGen Advisory Committee, March, 2018
+
+
+
+	
+		NextGen Proirities -Joint Implementation Plan Update including the Northeast Corridor
+		
+			PWhitley
+		
+		
+			TBristol
+		
+		
+			ABahrami
+		
+	
+	
+		FAA
+		
+			October, 2017
+		
+	
+	Whitley, P., Bristol, T., and Bahrami, A., "NextGen Proirities -Joint Implementation Plan Update including the Northeast Corridor," FAA, October, 2017
+
+
+
+	
+		Presentation to NEC NIWG
+		
+			Faa
+		
+		
+		
+			December, 2017 10. 14 April 2018
+		
+	
+	Initial TBO and NEC. cited 25 Apr 2018
+	FAA, "Initial TBO and NEC," Presentation to NEC NIWG, December, 2017 10 https://www.nasa.gov/aeroresearch/programs/atd/project-description [cited 25 Apr 2018] 11 https://www.nasa.gov/aeroresearch/programs/aosp/smart-nas-project-description [cited 14 April 2018]
+
+
+
+	
+		Integrated Demand Management (IDM) - Minimizing Unanticipated Excessive Departure Delay while Ensuring Fairness from a Traffic Management Initiative
+		
+			Hyo-SangYoo
+		
+		
+			ConnieBrasil
+		
+		
+			NancyMSmith
+		
+		
+			PaulULee
+		
+		
+			ChristophMohlenbrink
+		
+		
+			NathanBuckley
+		
+		
+			AlGlobus
+		
+		
+			GitaHodell
+		
+		10.2514/6.2017-4100
+	
+	
+		17th AIAA Aviation Technology, Integration, and Operations Conference
+		
+			American Institute of Aeronautics and Astronautics
+			June 2017
+			
+		
+	
+	Yoo, H., et al., "Integrated Demand Management: Minimizing Unanticipated Excessive Departure Delay while Ensuring Fairness from a Traffic Management Initiative," 17th AIAA Aviation Technology, Integration, and Operations Conference, 5-9 June 2017
+
+
+
+	
+		Shadow Mode Assessment using Realistic Technologies for the National Airspace System (SMART NAS)<br /> Test Bed Development (Invited)
+		
+			KeePalopo
+		
+		
+			GanoBrotoChatterji
+		
+		
+			MichaelDGuminsky
+		
+		
+			PatriciaCGlaab
+		
+		10.2514/6.2015-2794
+	
+	
+		AIAA Modeling and Simulation Technologies Conference
+		
+			American Institute of Aeronautics and Astronautics
+			June 2015. 2014
+			
+		
+	
+	Autonomy Research For Civil Aviation: Toward A New Era Of Flight
+	Palopo, K., et al., "Shadow Mode Assessment using Realistic Technologies for the National Airspace System (SMART NAS) Test Bed Development," 15th AIAA Aviation Technology, Integration, and Operations Conference, 22-26 June 2015 14 Committee on Autonomy Research for Civil Aviation, "Autonomy Research For Civil Aviation: Toward A New Era Of Flight," National Academies Press, 2014
+
+
+
+	
+		Architecture and capabilities of a data warehouse for ATM research
+		
+			MichelleEshow
+		
+		
+			MaxLui
+		
+		
+			ShubhaRanjan
+		
+		10.1109/dasc.2014.6979560
+	
+	
+		2014 IEEE/AIAA 33rd Digital Avionics Systems Conference (DASC)
+		Colorado Springs, CO
+		
+			IEEE
+			October 2014
+			
+		
+	
+	Eshow, M. M., Lui, M., and Ranjan, S., "Architecture and Capabilities of a Data Warehouse for ATM Research," IEEE/AIAA 33 rd Digital Avionics System Conference, Colorado Springs, CO., 5-9 October 2014
+
+
+				
+			
+		
+	

file121.txt

@@ -0,0 +1,322 @@
+
+	
+	
+		
+I. Introductionhe amount of fuel consumed is an important metric for benefit assessment of air traffic management concepts being considered for improving throughput, increasing capacity and reducing delays.It is also an important metric for environmental impact because for each kilogram of fuel consumed, three kilograms of carbon dioxide, a greenhouse gas, is generated.The main motivations for developing the fuel estimation method is to establish a baseline for the current operations and based on it determine benefits of the proposed four-dimensional trajectory management concepts in terms of fuel usage.Alternative procedures for efficient descent, terminal area scheduling and spacing, and departure release can also be evaluated based on fuel consumption.Four prior related publications on the subject of fuel estimation are cited here as Refs.1-4.References 1 and 2 are focused on departure and arrival fuel consumption below 10,000 feet altitude.Reference 1 compares the International Civil Aviation Organization (ICAO) time-in-mode method based fuel consumption with the actual fuel consumption reported in Flight Data Recorders.Fuel flow-rate patterns were found to be quite different than the ICAO model estimates due to airline climb/descent procedures.This suggests that a fuel consumption model should include aircraft state information such as airspeed.Reference 2 presents thrust specific fuel consumption models for climb and descent.Model parameters are adjusted to fit the aircraft manufacturer data.Thrust specific fuel consumption is multiplied with thrust to determine fuel consumption.Their method assumes nominal climb/descent profiles; it does not consider airline and air traffic control specific operational procedures.Furthermore, the method is only applicable to lower altitudes.Reference 3 describes a closed-form takeoff weight estimation method developed using the constant-altitude-cruise range equation and aircraft design principles.It needs flight-plan data and aircraft performance model to estimate the takeoff weight of the aircraft.The amount of fuel needed for climb, cruise and descent phases of flight and the maximum load factor are computed as a part of the procedure.The method described in this paper differs from Ref. 3 in that it uses the actual flight track data and does not require a model for climb and descent; thus, it is more data driven than model based.Reference 4 describes a fuel estimation procedure using actual trajectory of aircraft, and Base of Aircraft Data (BADA) drag and fuel-flow models.Their procedure is close to the method described here.The main difference is that the fuel estimation procedure is derived from nonlinear equations of motion with point-mass assumptions as opposed to approximations adopted in Ref. 4. Additional contribution of the present work is estimation of aircraft and wind states.Main contribution of this paper is development of the fuel estimation procedure from basic principles without simplifications.The procedure was validated against flight test data provided by the Federal Aviation Administration.A takeoff weight estimation procedure is developed for estimating fuel usage and establishing fuel usage uncertainty bounds when the takeoff weight of the aircraft is unknown.Finally, the adequacy of using position data acquired by air traffic control radar systems for fuel estimation is examined.Results show that in spite of bias, noise and data drop issues, position data could be conditioned for obtaining decent fuel estimates.Section II describes the fuel estimation procedure.The BADA fuel-flow model is discussed in Section III.The equations of motion are given in Section IV.This section also lists an expression for thrust in terms of drag, and aircraft and wind states.Section V provides the BADA drag model.Expressions for lift and bank angle estimation are listed in Section VI.Aircraft state estimation is described in Section VII.Results are discussed in Section VIII.The paper is concluded in Section IX.
+II. Fuel Burn Estimation ProcedureTo determine the amount of fuel consumed, altitude, airspeed and thrust have to be estimated.Altitude is obtained from the trajectory.Airspeed is estimated using a sequence of latitude ( ! ), longitude (! ) and altitude ( h ) reports as a function of time that define the fourdimensional trajectory and wind velocity.Computation of thrust requires an estimate of drag, which depends on lift.Lift depends on estimated aircraft and wind states, and weight.Once lift is determined, the lift induced drag coefficient can be computed.Drag is a function of airspeed, air density, and the drag coefficient, which depends on the aerodynamic configuration of the aircraft.Thrust is determined using estimates of aircraft and wind states, drag and weight.Fuel-flow rate is then obtained using altitude, airspeed and thrust estimates.Weight of the aircraft at a point in time is obtained by subtracting the amount of fuel consumed up to that time from the initial weight (takeoff weight).Fig. 1 shows the steps of the fuel burn estimation procedure.
+III. BADA Fuel Consumption ModelThe BADA fuel consumption model is described for nominal and idle thrust conditions.The nominal fuel-flow rate for jets and turboprops is determined by the product of the thrust specific fuel consumption and thrust, T .Thrust specific fuel consumption for jets is modeled as a linear function of airspeed, V , and for turboprops as a quadratic function of airspeed.Fuel-flow rate is independent of airspeed and thrust for aircraft with piston engines.A generalized expression for the nominal fuel-flow rate for these three different aircraft types can be written in the following form 5 :( )T V f V f f f f nom 2 3 2 1 0 ! + + = (1)where the coefficients in Eq. ( 1) are given in terms of the BADA coefficients 1. Units of the BADA coefficients are provided in the Appendix for completeness.Fuel-flow rate is in kg/s with airspeed in knots and thrust in Newtons.1 f C and 2 f C in TableThe minimum fuel-flow rate for idle thrust is modeled as a linear function of altitude, h , for jet and turboprop engine types and as a constant for piston engine.This model is described by the following equation:h f f f 5 4 min ! = (2)American Institute of Aeronautics and Astronautics 3 Altitude is in feet.The coefficients are again defined in terms of BADA coefficients 2. Units of these BADA coefficients are also listed in the Appendix.Fuel-flow rate coefficients for a jet, a turboprop and a piston aircraft are listed in Table 6 of the Appendix to give the reader a feel for the contribution of these coefficients to fuel-flow rate in Eqs. ( 1) and (2).3 f C and 4 f C in TableThe nominal and the minimum fuel-flow rate models can be combined into a single expression, ( )nom fcr f C f f min, max = (3)The fraction of fuel-flow rate during the cruise phase is fcr C .Its numerical value is 1 during the other flight phases.Equations (1) through (3) show that to estimate fuel-flow rate for jets and turboprops, altitude, airspeed, thrust and the phase of flight (in cruise or not) needs to be known.Altitude is directly available from position reports; airspeed, thrust and the phase of flight have to be estimated.Airspeed can be estimated using the reported position and wind data.Equations of motion, which are discussed in the next section, have to be used for thrust estimation.The amount of fuel consumed can be determined by integrating the fuel-flow rate as! = f t f dt f m 0 (4)
+Engine Type4 f 5 f Jet 3 60 1 f C ! " # $ % & ! ! " # $ $ % & ! " # $ % & 4 3 60 1 f f C C Turboprop 3 60 1 f C ! " # $ % & ! ! " # $ $ % & ! " # $ % & 4 3 60 1 f f C C Piston 3 60 1 f C ! " # $ % & 0 Table 1.Nominal fuel-flow rate model coefficients.
+Engine Type0 f 1 f 2 f 3 f Jet 0 1 4 10 6 1 f C ! " # $ % & ' ! ! " # $ $ % & ! " # $ % & ' 2 1 4 10 6 1 f f C C 0 Turboprop 0 0 1 7 10 6 1 f C ! " # $ % & ' ! ! " # $ $ % & ! " # $ % & ' 2 1 7 10 6 1 f f C C Piston 1 60 1 f C ! " # $ % & 0 0 0 Figure 2. Velocity triangle.where f t is the flight time.
+IV. Equations of MotionThe motion of aircraft, modeled as a point mass, is often described by the following three equations (see Ref.
+6):gn V h R ) ( 1 + = ! & (5) ge V h R ! " cos ) ( 1 + = & (6) and h V h = & (7)! is the latitude, ! is the longitude, h is the geometric altitude and R is the mean radius of the Earth.gn V and ge V are the north and east components of the ground-relative aircraft velocity.h V is the climb or descent rate depending on whether it is positive or negative.The horizontal velocity of the aircraft with respect to the ground is the resultant of the horizontal components of the airmass-relative velocity of the aircraft and the wind velocity.This relationship is shown in Fig. 2, where The magnitude of the airmass-relative acceleration resulting from the thrust, drag, lift and gravitational forces on the aircraft modeled as a point mass is! ! " ! " ! # sin cos sin cos cos sin cos h e n W W W g m D T V & & & & $ $ $ $ $ = (8)where V is airmass-relative speed (true airspeed), T is thrust, D is drag, ! is angle-of-attack, m is mass, g is acceleration due to gravity and ! is flight path angle.Note that! cos V V s =(9)The kinetic equations for airmass-relative heading angle and flight path angle areµ µ # " cos cos cos sin cos sin sin sin V W V W mV L T e n & & & $ + + = (10) and V W V W V W V g mV L T h e n ! ! " ! " ! µ µ # ! cos sin sin sin cos cos cos cos sin & & & & $ + + $ + = (11)µ is bank angle and L is lift.Equations ( 9), ( 10) and ( 11) are derived assuming flat Earth, constant gravitational acceleration and slowly changing mass.The altitude rate, Eq. ( 7), can be written in terms of the airspeed, flight path angle and the vertical component of the wind velocity ash h W V V h + = = ! sin & (12)Since wind varies both with position and time, the time derivative of the north, east and up components of the wind velocity can be determined ash and e n i h h W W W t W W i i i i i , ; = ! ! + ! ! + ! ! + ! ! " " # # (13)Observe that !& , !& and h & are defined in Eqs. ( 5) through (7).Assuming the angle of attack to be zero in Eq. ( 8), ( ) ( )! " ! # $ ! % ! & ' ( ( ) * + + , -. . + + . + + + = 2 1 sin cos V W h W W V W h W g V m D T h e n h h / /(14)This expressions shows that thrust estimate depends on drag, mass, altitude rate, airspeed, rate of change of airspeed, wind terms and the airmass-relative heading angle.Dropping the wind terms, the following simplified expression is obtained:V h mg V m D T & & + + = (15)It is now easy to see that the thrust required for balancing the right hand side of Eq. ( 15) during deceleration and descent can be less than the minimum thrust generated by the engines due to errors in the drag model and aircraft weight.A minimum thrust model is required in these instances.It can be constructed by equating Eq. ( 1) to Eq. ( 2) assuming the BADA fuel model to be consistent.Thus for jets and turboprops, the minimum thrust is obtained as,2 3 2 1 5 0 4 min V f V f f h f f f T ! + ! ! = (16)Note that Eq. ( 16) cannot be used for piston engines because nominal, minimum and cruise fuel-flow are specified to be a constants for piston engines in BADA.
+V. Drag ModelAerodynamic drag force is obtained as the product of the drag coefficient, D C , and the dynamic pressure asS V C D D 2 2 1 ! = (17)where ! is the density of air and S is the wing reference area.D C is given as the sum of the zero-lift drag coefficient, 0 D C , and the induced drag coefficient, which is a quadratic function of the lift coefficient, L C .Thus,2 2 0 L D D D C C C C + = (18) 0 D C and 2 DC are functions of aerodynamic configuration of the aircraft.BADA coefficients associated with the aerodynamic configuration are listed in Table 3. Traditionally drag coefficients are given as a function of Mach number and Reynolds number.BADA models these values as constants; it does not take Mach and Reynolds number effects into account.Note that the additional termLDG D C ! , 0represents drag rise due to deployment of the landing gear.During the approach and landing configurations, drag coefficients are adjusted for flap setting.One of difficulties of drag computation is determining the aerodynamic configuration.BADA specifies conditions based on stall speeds and maximum altitude thresholds that have to be met based on airspeed and altitude to determine the aerodynamic configuration.The only remaining parameter that needs to be specified for drag computation is L C , which can be obtained using the definition of the lift force asS V L C L 2 2 ! = (19)The lift force needed is related to heading angle and flight path angle rates as shown in Eqs.(10) and (11), therefore assumptions have to be made about the trajectory being followed.This is discussed in the next section.
+VI. Trajectory AssumptionsTo stay on course in a wind field, the aircraft has to crab such that the across-track component of the wind is cancelled.The airmass-relative heading angle needed to stay on the path specified by the course angle g ! is obtained from the two relations based on Fig. 2: Table 3. Drag coefficients as a function of aerodynamic configuration.
+Aerodynamic Configuration0 D C 2 D C Takeoff TO D C , 0 TO D C , 2 Initial Climb IC D C , 0 IC D C , 2 Clean CR D C , 0 CR D C , 2 Approach AP D C , 0 AP D C , 2 Landing LDG D LD D C C ! + , 0 , 0 LD D C , gn n s V W V = + ! cos (20) and ge e s V W V = + ! sin (21) as ! ! " # $ $ % & ' ' = ' n gn e ge W V W V 1 tan ( (22)Resultant magnitude of the horizontal component of the airmass-relative aircraft velocity is2 2 ) ( ) ( e ge n gn s W V W V V ! + ! = (23)Combining Eq. ( 23) with Eqs. ( 9) and ( 12), the following expression for true airspeed is obtained in terms of aircraft and wind velocity states:2 2 2 ) ( ) ( ) ( h e ge n gn W h W V W V V ! + ! + ! = &(24)Lift force, which is needed for L C computation in Eq. ( 19), and the bank angle can now be computed using Eqs.(W V W V W V W V W V W V ! + ! ! ! ! ! ! " (25)Substituting !& from Eq. (25) in Eq. (10) and using the relations in Eqs. ( 9), (20), ( 21) and (23),ge gn S V P V P L L & & 2 1 sin + = = µ (26) with 2 2 1 ) ( ) ( ) ( e ge n gn e ge W V W V W V m P ! + ! ! ! = (27) and 2 2 2 ) ( ) ( ) ( e ge n gn n gn W V W V W V m P ! + ! ! = (28)To evaluate the other component of the lift force vector, the first step consists of differentiating Eq. ( 9) or Eq. ( 12) to get !& .Differentiating Eq. ( 12), American Institute of Aeronautics and Astronautics8 ! ! ! cos sin V V W h " " = (29)Substituting in Eq. (11) and using Eqs.( 9), ( 12), ( 20) and ( 21W h W V W V W V W V W W V m P ! + ! + ! ! + ! ! = & (31) 2 2 2 2 2 4 ) ( ) ( ) ( ) ( ) ( ) ( h e ge n gn e ge n gn h e ge W h W V W V W V W V W W V m P ! + ! + ! ! + ! ! = & (32) 2 2 2 2 2 5 ) ( ) ( ) ( ) ( ) ( h e ge n gn e ge n gn W h W V W V W V W V m P ! + ! + ! ! + ! = & (33) 2 2 2 2 2 6 ) ( ) ( ) ( ) ( ) ( ) ( h e ge n gn e ge n gn n gn W h W V W V W V W V W V m P ! + ! + ! ! + ! ! ! = &(34)and2 2 2 2 2 7 ) ( ) ( ) ( ) ( ) ( ) ( h e ge n gn e ge n gn e ge W h W V W V W V W V W V m P ! + ! + ! ! + ! ! ! = & (35)Lift force and the bank angle can be determined using Eq. ( 26) and (30);2 2 C S L L L + = (36) ! ! " # $ $ % & = ' C S L L 1 tan µ (37)L C can now be obtained using Eq. ( 19) with lift force determined using Eq. (36).Drag force can be determined using Eqs.( 18) and (17).Finally, thrust can be computed using Eq. ( 14).Note that the airspeed and airmass-relative acceleration in Eq. ( 14) can be replaced by ground-relative terms using Eqs.(20), (21), ( 23) and (24) as described in the Appendix.The fuel-flow rate is determined using Eq.(3) via Eqs.( 1) and (2), and the fuel consumed is obtained using Eq.(4).
+VII. State EstimationThe procedure for estimating aircraft states, which are needed in the steps described in the previous sections is outlined in this section.Observations for estimation of aircraft states are given as a temporal sequence of latitudes, longitudes and altitudes that constitutes the four-dimensional trajectory of the aircraft.Given this sequence of observations, a state estimator such as a Kalman Filter can be designed using the state equations, Eqs. ( 5) - (7).Alternatively, filters and smoothers can be used for state estimation as described in Ref. 7.Figure 3 shows a Proportional-Integral-Derivative (PID) controller based estimator design for the altitude channel.The objective is to estimate the altitude rate and vertical acceleration from altitude time history.Estimated values are denoted by the superscript "^."The controller assumes a double integrator model of aircraft with altitude rate and altitude as measurements for feedback.The commanded acceleration results in altered altitude rate and altitude.The proportional gain, P K , integral gain, I K , and the derivative gain, D K , of the controller can be chosen by placing the poles of the closed-loop system in the left half of the s-plane.Optimal gains can be chosen by equating the coefficients of the characteristic polynomial to the Butterworth or integral of time multiplied by the absolute value of error (ITAE) polynomials listed in Ref. 8.A discrete version of the estimator in Fig. 3 is obtained by approximating the first and second derivatives of altitude using Taylor Series approximation about time-step k , which is separated from the next time-step 1 + k by time t ! ,as follows. .Once altitude is estimated using Eq. ( 40), the altitude rate and vertical acceleration can be estimated using Eqs.(38) and (39).One could estimate the altitude rate and acceleration by using the observed altitudes instead of the estimated altitudes.This however, would result in noisy estimates because noise in the altitude measurement would be amplified by the differencing process.Reducing noise in the altitude measurement by using the PID filter prior to The estimator implemented by Eqs.(38) through ( 41) is also used for the latitude and longitude channels with latitude and longitude measurements.These independent estimators provide estimates of angles, angular rates and angular accelerations:t k h k h k h ! " + = + ) ( ) 1 ( ) 1 ( & (38) and 2 ) 1 ( ) ( 2 ) 1 ( ) 1 ( ˆt k h k h k h k h ! " + " + = + & & (39) Thus, ) 1 ( ) ( ) 1 ( ) ( ) 2 ( ) 1 ( ) ( ) 1 ( ˆ3 2 2 3 2 t K t K t K k e t k h k h t K k h t K t K! ˆ, ! & ˆ, ! & & ˆ, !ˆ, ! & ˆ and ! & & ˆ.The horizontal components of the aircraft velocity vector can now be estimated using Eqs.( 5) and (6).Finally, the acceleration terms can be estimated using derivatives of Eqs. ( 5) and ( 6) as follows:    large errors occurred at takeoff and just after landing due to sudden change in velocity.! ! ) ( ˆh h R V gn + + = (42) and ! " ! " " !To compute the airmass-relative aircraft velocity terms and the wind terms needed for lift, drag, thrust and fuel-flow computations, horizontal components of the wind velocity were obtained from the 4/17/2009 hourly RUC data.Vertical component of the wind was assumed to be zero.The spatial and temporal partial derivatives of the ground-relative wind velocity are computed using finite-differences (see Eq. 38 for example) along the FDR reported trajectory.Latitude, longitude and altitude rates obtained via the state estimators are used with the partial derivatives to obtain the total derivatives via Eq.(13).Horizontal components of wind velocity and aircraft state estimates were used to estimate true airspeed using Eq.(24).This true airspeed was converted to calibrated airspeed using the standard atmosphere model.Pressure and temperature values derived from RUC data can also be used for this conversion.Figure 8 shows the time histories of the estimated calibrated airspeed and the indicated airspeed from the FDR. Figure 9 shows the difference between the estimated calibrated airspeed and the indicated airspeed derived from the FDR.Mean and standard deviation of the errors are -3.3 knots and 8.6 knots, and the extremal values are -64.2knots and 32.0 knots.As in Fig. 7, large errors occurred at takeoff and upon landing.Results presented in the figures above illustrate aircraft and wind state estimation accuracy.The next set of figures demonstrates the accuracy of fuel burn estimation.Estimated aircraft states, wind states and weight were used to estimate lift using Eq.(36).Takeoff weight of the aircraft was specified to be 39,362 kg (86,778 lb), which was the actual takeoff weight of the FAA aircraft employed for the flight test.Subsequently, the estimated amount of fuel burnt was subtracted from the takeoff weight to estimate weight as a function of time.17) was then used with the estimated aircraft states, wind states and weight to estimate thrust using Eqs.( 14) and ( 16).V & was set to zero for cruise.Thrust estimates were found to be noisy, so they were smoothed using the procedure described in Section VII.The smoothed estimates were then used in Eq. ( 1) to estimate the fuel-flow rate.
+Fuel Estimation ValidationEstimated fuel-flow rate and fuel-flow rate from the FDR during the climb phase are shown in Fig. 10.Notice the low fuel-flow rate in Fig. 10 and the flat altitude in Fig. 5 for the first 10 minutes; they correspond to taxi on the ground.During the cruise phase are shown in Fig. 11.Fuel-flow rates for the descent phase are given in Fig. 12. Observe the big spike in fuel-flow rate from FDR after the 312 minute mark; it is most likely due to increased thrust accompanied with thrust reverser deployment for speed reduction to taxi speed.Based on this observation, the negative estimated thrust is considered to be positive thrust with thrust reverser deployed after landing.The resulting fuel-flow rate value was estimated to be 0.38 kg/s compared with 0.72 kg/s reported in FDR data.For reference, aircraft altitude is about 14,770 feet at the 301 minute mark, 8,000 feet at the 305 minute mark and zero at the 312 minute mark.The aircraft needs to be below 8,000 feet for approach configuration and at or below 3,000 feet for landing configuration.As expected, thrust and fuel-flow rate were found to be strongly correlated.Finally, Fig. 13 shows the estimate of the amount of fuel consumed and the FDR reported values as a function of time.It is difficult to assess the error from this figure therefore Fig. 14 is provided to show the relative error.Relative error drops below 20% seven minutes into flight when the aircraft is at about 11,000 feet altitude.Mean and standard deviation of the fuel burn error were found to be -1.4% and 10.6%.Extremal values were determined to be -79.4% and 36.1%.values occur prior to takeoff.The total amount of fuel consumed during the flight was estimated to be 8,099 kg, which is only two kilograms more than the FDR value.To get this close match, the Bombardier RJ-900 Regional Jet model fuelflow coefficients were multiplied by a factor of 0.853.This value was obtained by trial and error.The fuel consumed estimate is lowered to 8,072 kg with a factor 0.85.The error with respect to the FDR value of 25 kg is reasonable based on the observed uncertainty of 50 pounds (23 kg) in the FDR fuel consumed data.The other meaningful measure is the mean of the relative error shown in Fig. 14.Considering relative error beyond 100 minutes, the mean value is 0.86%; thus, it is fair to assume that the error in the estimated amount of fuel consumed is within 1% of the actual amount of fuel consumed in the flight test.This result validates the fuel estimation procedure described in the paper.
+Model SimplificationTo determine if the simpler model used in Ref. 4 is adequate for fuel estimation, lift was set equal to weight in all phases of flight and thrust was modeled using Eqs.( 15) and ( 16).This means that the wind terms in Eq. ( 14) were dropped.As in the complete model, described in the previous section, V & was set to zero for cruise.Other than replacing the lift and thrust models with simpler models, all the steps described for obtaining the results in the previous section were followed.Results obtained on the flight test data matched the results in the previous section.Differences between these two sets of results were found to be negligible.Mean and the standard deviation of the fuel burn error were found to be -1.35% and 10.65% compared to -1.42% and 10.62% for the data in Fig. 14.Based on these results, the simpler lift and thrust models are adequate for fuel burn estimation.
+Takeoff Weight Estimation and UncertaintyThe results described in the previous section were based on the actual takeoff weight of the test aircraft.In most instances, the actual takeoff weight of the aircraft will not be known therefore, a procedure is needed for estimating the takeoff weight of the aircraft.One possibility is to use the procedure described in Ref. 3. This technique uses the constant-altitude range equation and aircraft design principles for estimating the takeoff weight.The method also needs distance to the top-of-climb point and the weight at this location for given cruise-altitudes and cruise-speeds.This information is derived by simulating the climb trajectory according to BADA aircraft performance and procedure models.The second possibility is to assume an initial weight without fuel and then estimate the amount of fuel needed for flying from the airport of origin to the airport of destination.Since the amount of fuel needed has to be carried onboard the aircraft, a bit more fuel is consumed for carrying this additional weight.A few iterations of adding the fuel needed to the takeoff weight and computing the fuel needed for the flight should yield a good estimate of the takeoff weight.This approach has been employed for generating the results discussed below.The main benefit is less dependence on BADA and more reliance on actual trajectory data.To start the iterations, the takeoff weight was set to the maximum zero-fuel weight of 25,401 kg (56,000 lb) based on manufacturer data (see Ref. 10).Maximum zero-fuel weight includes the structural weight of the aircraft, crew, maximum payload and everything other than the fuel.The zero-fuel weight can be adjusted to a lower value based on assumed load-factor.For example, a load-factor of 0.8 means that payload is assumed to be 80% of the maximum payload.Although the actual zero-fuel weight of the test aircraft 23,509 kg (51,828 lb) was known, it was not used because this would not be known for a typical flight.The assumption is that the aircraft was carrying the full payload.One would need to reduce the payload if the takeoff weight exceeds the maximum takeoff weight or when the destination cannot be reached with maximum fuel.A flowchart in Ref. 3 describes these conditions.Next, the simplified lift and thrust models of the previous section were used with the fuel estimation procedure to determine the amount of fuel needed for the flight and the average fuel burn rate during cruise.The fuel needed was 6,777 kg (14,941 lb) and the average fuel burn rate was 17.8 kg/min (39.2 lb/min).The average burn rate was used to determine the amount of reserve fuel.Federal Aviation Regulations require domestic flights conducted under Instrument Flight Rules to have enough fuel to fly to the first airport of intended landing, then fly to an alternative airport (if conditions require an alternative airport), then for 45 minutes thereafter at normal cruising speed.The weight of the reserve fuel was determined to be 1,599 kg (3,526 lb) assuming 90 minutes of additional flight time.The initial takeoff weight of 25,401 kg was augmented with 6,777 kg of fuel needed for the flight and 1,599 kg of reserve fuel for the next iteration.This process was repeated for the subsequent iterations.Takeoff weight is shown as a function of iterations in Fig. 15.Observe that there is very little change in the takeoff weight estimate after the fourth iteration.After 10 iterations, takeoff weight estimated value is 34,812 kg (76,748 lb).The actual takeoff weight of the Global 5000 was 39,362 kg (86,778 lb), which is 11.6% more than the estimated takeoff weight.The main reason for the difference is that 7,756 kg (17,099 lb) of extra fuel was carried during the flight test.This example shows that accurate takeoff weight estimation is difficult.If the aircraft has greater range capability, it can carry more fuel than that required for the flight.Airlines sometime ferry fuel depending on where fuel is cheapest to purchase.Since the amount of fuel burned is a function of the takeoff weight, any uncertainty in takeoff weight translates into uncertainty in the estimate of fuel burned.To explore this aspect a bit more, the fuel estimation procedure was repeated with maximum takeoff weight of 41,957 kg (92,500 lb) specified by the manufacturer (see Ref. 10).The results are summarized in Table 4.The first row of Table 4 shows results for takeoff weight estimated using the iterative procedure.The second row shows results obtained using the actual flight test weight.The third row shows results with maximum takeoff weight.Second column shows the difference of the takeoff weight with respect to the actual takeoff weight.The third column shows a single value derived from flight test data.Estimated fuel consumption corresponding to the takeoff weights is shown in column four.Column five lists the fuel consumption error with respect to the measured fuel consumption.The second and the fifth columns show that takeoff weight errors translate to fuel consumption errors.The smallest and the largest fuel consumption values in column four represent fuel estimation uncertainty bounds.Analysis to this sort can be beneficial in determining the bounds of environmental impact related to fuel consumption.
+Fuel Estimation Using ASDI DataThe ultimate objective of this paper is to enable fuel estimation using data from air traffic data sources.The primary sources of trajectory data are the Host Computers in the Air Route Traffic Control Centers (ARTCC).These data are provided at a 12-second interval.Trajectory data from Host Computers are consolidated and provided at a one-minute interval as Airline Situation Display to Industry (ASDI) data.Fuel estimation results obtained with the trajectory of the flight test aircraft derived from 4/17/2009 ASDI data are discussed in this section.Track data corresponding to the FAA's Global 5000 aircraft tail number were extracted and compared with position data recorded by the FDR.The first ASDI track data was 3.9 nautical-miles away from the airport, based on the first FDR position data, and at an altitude of 3,100 feet.Similarly, the last ASDI track data was 3.8 nauticalmiles away from the airport, based on the last FDR position data, and at an altitude of 793 feet.In addition, the following ASDI position data quality issues were also identified.While the mean, median and mode temporal separation between two successive track positions were 60 seconds, the maximum and minimum separation were 184 seconds and 30 seconds.The standard deviation was 8 seconds.These observations confirm that ASDI and Host Computer track data are not totally synchronous and periodically suffer from data drops.To determine the accuracy of reported position, FDR position data were first interpolated using spline interpolation and sub-sampled at the ASDI track times.Great circle distances between the resulting FDR position data and the corresponding ASDI position data were then determined.The mean and the standard deviation of the errors were determined to be 1.23 and 0.62 nautical-miles.Minimum and maximum were 0.13 and 3.05 nautical-miles.The main point is that position data acquired by radar is affected by bias, noise and quantization errors.To deal with data quality issues, outliers were removed from ASDI position data; then data were interpolated with a spline fit and sub-sampled at a 60-second interval.The resulting track data were then input to the fuel estimation procedure.The initial weight was set to the actual takeoff weight of 39,362 kg (86,778 lb) minus 408 kg (900 lb), where 408 kg of fuel was burnt to reach the altitude of 3,076 feet based on FDR data.There is a difference of 24 feet in the FDR and ASDI reported altitudes at the first position in ASDI data.The difference in the two altitudes at the last location is 245 feet with FDR altitude of 548 feet and ASDI altitude of 793 feet.The amount of fuel consumed was estimated to be 7,741 kg (17,066 lb) compared to the actual fuel consumption of 7,688 kg (16,949 lb).The estimation error is 0.7% with respect to the FDR reported fuel consumption.Next, the iterative weight estimation procedure was initiated with the maximum zero-fuel weight of 25,401 kg (56,000 lb).After 10 iterations the initial weight of aircraft was estimated to be 34,616 kg (76,315 lb) and the fuel consumed was estimated to be 7,276 kg (16,041 lb).This represents an error of -5.4% with respect to the FDR value of 7,688 kg.Although, the error is more, the iterative weight computation procedure is preferred because initial weight data will not be available as it was for the flight test.The other aspect is that if Host Computer data are used, only a part of the trajectory will be available.Starting with zero-fuel weight at the starting location accounts for fuel burned to fly up to that location to some extent.Estimation results can be expected to improve with Host Computer data because of faster update interval of 12-seconds.
+IX. ConclusionsThis paper described a procedure for estimating fuel consumption based on actual trajectory, and drag and fuelflow models.The method consists of estimating aircraft and wind states and using them to determine lift, drag, thrust and fuel-flow.Fuel consumption estimates generated for a Bombardier Global 5000 flight from Atlantic City to Los Angeles were compared with the Flight Data Recorder values, obtained during a flight test conducted by the Federal Aviation Administration, to validate the method.Results show that fuel usage can be estimated within 1% of the actual value when the takeoff weight is known.The procedure was simplified by setting lift equal to weight and by removing the wind terms from thrust.This simplification did not degrade fuel estimation accuracy.A procedure for estimating takeoff weight was then introduced.Starting with an initial estimate of takeoff weight, this procedure used reserve fuel requirements to iteratively improve the takeoff weight and fuel estimates.The method was found to converge within five iterations.It was shown that fuel usage uncertainty bounds can be determined by varying the takeoff weight.Finally, the adequacy of using trajectory data obtained by air traffic control systems was examined.Trajectory data for the Atlantic City to Los Angeles flight obtained from Airline Situation Display to Industry data were used for estimating fuel usage.Although these data suffered from bias, noise, asynchronous update, and data drops, it was possible to condition the data for obtaining reasonable fuel estimates.Fuel usage could be estimated within 5.4% of the actual value using Airline Situation Display to Industry data with simplified models and the iterative takeoff weight estimation method.W h W V W V W h W h W V W V W V W V V ! + ! + ! ! ! + ! ! + ! !2W V W V W V W W V W W W ! + ! ! + ! = + & & & &" " (A-3)V is given in terms of ground-relative quantities in Eq. (24).Thrust can therefore be computed in terms of ground- relative terms.Units of BADA fuel-flow coefficients are given in Table 5.Fuel-flow rate coefficients for a jet, a turboprop and piston engine types are given in Table 6.Table 6.Fuel-flow rate coefficients for a jet, a turboprop and a piston aircraft.
+Aircraft Type ManufacturerFigure 1 .1Figure 1.Fuel burn estimation procedure.
+Ware the magnitudes of the horizontal components of the ground-relative aircraft velocity and the wind velocity, and s V is the magnitude of the horizontal component of the airmass-relative aircraft velocity.g ! is the heading angle of the ground-relative aircraft velocity with respect to the local north direction.! and w ! are the heading angles of the airmass-relative aircraft velocity and ground-relative wind velocity also with respect to the local north direction.n W , e W and h W as the north, east and up components of the wind velocity vector.
+Figure 3 .3Figure 3. Altitude estimator.
+Figure 4 .4Figure 4. Actual latitude/longitude position history.
+Figure 5 .5Figure 5. Actual altitude time history.
+Figure 7 .7Figure 7. Groundspeed estimation error time history.
+Figure 6 .6Figure 6.FDR reported groundspeed time history.
+Figure 8 .8Figure 8.Time histories of estimated calibrated airspeed and indicated airspeed from FDR.
+Figure 9 .9Figure 9.Estimated calibrated airspeed error time history.
+Figure 11 .11Figure 11.Estimated and FDR fuel-flow rate during cruise.
+Figure 10 .10Figure 10.Estimated and FDR fuel-flow rate during climb.
+Figure 13 .13Figure 13.Estimated and FDR reported fuel consumption time history.
+Figure 12 .12Figure 12.Estimated and FDR fuel-flow rate during descent.
+Figure 14 .14Figure 14.Percentage error with respect to actual fuel usage.
+Figure 15 .15Figure 15.Convergence of takeoff weight estimates.
+Table 2 .2Minimum fuel-flow rate model coefficients.
+Table 4 .4Takeoff weight and fuel consumption uncertainty.Takeoff% WeightMeasured FuelEstimated Fuel% ErrorWeightErrorConsumptionConsumption34,812 kg-11.67,610 kg-6.0139,362 kg08,097 kg8,101 kg0.0541,957 kg6.68,407 kg3.83
+Table 5 .5Units of BADA fuel-flow rate coefficients.Engine TypeCf1Cf2Cf3Cf4CfcrJetkg/(min*kN)knotskg/minfeetdimensionlessTurbopropkg/(min*kN*knot) knotskg/minfeetdimensionlessPistonkg/min--kg/min--dimensionlessCf1Cf2Cf3Cf4CfcrCRJ-900Bombardier0.61472369.758.2151355,9101EMB-120 BrasiliaEmbraer4.5662664.156.555943,0481PA-28-161 CherokeePiper0.44515--0.30872--0.87274
+			& & & &
+			& & & & & &
+			& & & & &
+			& & & & &
+			& & & & & & & &
+			& & & & & &
+			& & & & & &
+			& & & & & & & & & & (A-2)The airmass-relative heading terms can be replaced using Eqs.(20), (21) and (23) as follows.
+		
+		
+
+			
+AcknowledgementsThe author thanks Confesor Santiago of NASA Ames Research Center (he was formerly at Federal Aviation Administration) and Mike Paglione of Federal Aviation Administration, and Robert Oaks of General Dynamics Information Technology for providing the flight test data and taking the time to answer questions.
+			
+
+			
+With the aircraft states computed using the altitude and the latitude/longitude estimators, and the wind related estimates obtained using Eq. ( 13), lift, drag, thrust and fuel-flow rate can be computed.The north and east components of the wind velocity vector can be obtained as a function of time, latitude, longitude and altitude from the Rapid Update Cycle (RUC) data, which are provided by the National Oceanic and Atmospheric Administration (NOAA).The vertical wind velocity can be computed by post processing RUC data using the relation described on page 480 of Ref. 9. h W is small relative to n W and e W ; it can be assumed to be zero.
+VIII. ResultsThis section is organized into five subsections.The first subsection on validation describes the flight test conditions and compares the estimated states with the actual states from the Flight Data Recorder (FDR).Fuel estimation accuracy is examined for the climb, cruise and descent phases of the flight in the second subsection.The next subsection discusses model simplification and its effect on fuel usage estimate.A procedure for takeoff weight estimation that starts by setting the takeoff weight to the maximum zero-fuel weight and then iteratively improves the estimate by adding reserve fuel and fuel consumed is discussed in the fourth subsection.Suitability of using radar-based position data for fuel estimation is explored in the fifth subsection.
+Aircraft State Estimation ValidationTo validate the fuel estimation procedure described in Fig. 1, FDR data from an actual flight of the Federal Aviation Administration (FAA) owned Bombardier Global 5000 aircraft from Atlantic City International airport (ACY) in New Jersey to Los Angeles International airport (LAX) in California on 4/17/2009 were used.These data were sampled at 4 second intervals.The dry weight of the aircraft was 23,509 kg (51,828 lb) and the initial fuel weight was 15,853 kg (34,950 lb).The total amount of fuel burnt during the flight test was 8,097 kg (17,850 lb).The FDR provided latitude/longitude position history is shown in Fig. 4 and the altitude time history is shown in Fig. 5.The temporal sequence of latitude, longitude and altitude derived from the FDR data were input to the latitude, longitude and altitude estimators, that were discussed in the previous section, to estimate latitude, longitude and altitude rates.Components of the groundrelative aircraft velocity computed via Eqs.( 5) and (6) with these rates were then used to estimate the groundspeed.FDR reported groundspeed is shown in Fig. 6 and the error of the estimated groundspeed with respect to it is shown as a function of flight time in Fig. 7. Groundspeed estimation error was found to have a mean of -0.5 knots, standard deviation of 3.6 knots and extremal values of -62.7 knots and 36.1 knots.These
+AppendixThe airmass-relative terms in the thrust equation, ( ) ( )			
+			
+
+				
+
+
+	
+		Analysis of Departure and Arrival Profiles Using Real-Time Aircraft Data
+		
+			JudithPatterson
+		
+		
+			GeorgeJNoel
+		
+		
+			DavidASenzig
+		
+		
+			ChristopherJRoof
+		
+		
+			GreggGFleming
+		
+		10.2514/1.42432
+	
+	
+		Journal of Aircraft
+		Journal of Aircraft
+		0021-8669
+		1533-3868
+		
+			46
+			4
+			
+			July-August 2009
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	Patterson, J., Noel, G. J., Senzig, D. A., Roof, C. J., and Fleming, G. G., "Analysis of Departure and Arrival Profiles Using Real-Time Aircraft Data," Journal of Aircraft, Vol. 46, No. 4, July-August 2009, pp. 1094-1103.
+
+
+
+	
+		Modeling of Terminal-Area Airplane Fuel Consumption
+		
+			DavidASenzig
+		
+		
+			GreggGFleming
+		
+		
+			RalphJIovinelli
+		
+		10.2514/1.42025
+	
+	
+		Journal of Aircraft
+		Journal of Aircraft
+		0021-8669
+		1533-3868
+		
+			46
+			4
+			
+			July-August 2009
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	Senzig, D. A., Fleming, G. G., and Iovinelli, R. J., "Modeling of Terminal-Area Airplane Fuel Consumption," Journal of Aircraft, Vol. 46, No. 4, July-August 2009, pp. 1089-1093.
+
+
+
+	
+		Closed-Form Takeoff Weight Estimation Model for Air Transportation Simulation
+		
+			Hak-TaeLee
+		
+		
+			GanoChatterji
+		
+		10.2514/6.2010-9156
+	
+	
+		10th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference
+		Fort Worth, TX
+		
+			American Institute of Aeronautics and Astronautics
+			September 13-15, 2010
+		
+	
+	AIAA 2010-8164
+	Lee, Hak-Tae., and Chatterji, G. B., "Closed-Form Takeoff Weight Estimation Model for Air Transportation Simulation," AIAA 2010-8164, Proc. 10th AIAA Aviation Technology, Integration and Operations Conference (ATIO), Fort Worth, TX, September 13-15, 2010.
+
+
+
+	
+		Prototype Implementation and Concept Validation of a 4-D Trajectory Fuel Burn Model Application
+		
+			RobertOaks
+		
+		
+			HollisRyan
+		
+		
+			MikePaglione
+		
+		10.2514/6.2010-8164
+	
+	
+		AIAA Guidance, Navigation, and Control Conference
+		Toronto, Ontario, Canada
+		
+			American Institute of Aeronautics and Astronautics
+			August 2-5, 2010
+		
+	
+	Oaks, R. D., and Paglione, M., "Prototype Implementation and Concept Validation of a 4-D Trajectory Fuel Burn Model Application," AIAA 2010-8164, Proc. AIAA Guidance, Navigation, and Control Conference, Toronto, Ontario, Canada, August 2-5, 2010.
+
+
+
+	
+		EUROCONTROL moves forward with modernisation project
+		
+			Eurocontrol
+		
+		10.1108/aeat.2009.12781dab.014
+		No. 2009-009
+	
+	
+		Aircraft Engineering and Aerospace Technology
+		
+			BP
+		
+		
+			F
+		
+		0002-2667
+		
+			81
+			4
+			March 2009
+			Emerald
+			Bretigny-sur-Orge, France
+		
+	
+	EEC Technical/Scientific Report
+	Eurocontrol Experimental Centre
+	Eurocontrol, "Base of Aircraft Data (BADA) Aircraft Performance Modelling Report," EEC Technical/Scientific Report No. 2009-009, Eurocontrol Experimental Centre, B. P. 15, F-91222 Bretigny-sur-Orge, France, March 2009.
+
+
+
+	
+		En-route flight trajectory prediction for conflict avoidance and traffic management
+		
+			GBChatterji
+		
+		
+			BSridhar
+		
+		
+			KDBilimoria
+		
+		10.2514/6.1996-3766
+	
+	
+		Guidance, Navigation, and Control Conference
+		San Diego, CA
+		
+			American Institute of Aeronautics and Astronautics
+			July 29-31, 1996
+		
+	
+	Chatterji, G. B., Sridhar, B., and Bilimoria, K. D., "En-route Flight Trajectory Prediction for Conflict Avoidance and Traffic Management," AIAA 96-3766, Proc. AIAA Guidance, Navigation, and Control Conference, San Diego, CA, July 29-31, 1996.
+
+
+
+	
+		Short-term trajectory prediction methods
+		
+			GanoBChatterji
+		
+		10.2514/6.1999-4233
+	
+	
+		Guidance, Navigation, and Control Conference and Exhibit
+		Portland, OR
+		
+			American Institute of Aeronautics and Astronautics
+			1999
+		
+	
+	AIAA 99-4233
+	Chatterji, G. B., "Short-Term Trajectory Prediction Methods," AIAA 99-4233, Proc. AIAA Guidance, Navigation, and Control Conference, Portland, OR, 1999.
+
+
+
+	
+		American politics system. By Hugh York, McGraw-Hill and the Party system. By Hugh A. Bone. New York, McGraw-Hill Book Company, Inc., 1949. viii, 777 pp. $5.50
+		
+			JJD'azzo
+		
+		
+			HHoupis
+		
+		10.1002/ncr.4110390314
+	
+	
+		National Municipal Review
+		Nat Mun Rev
+		0190-3799
+		1931-0250
+		
+			39
+			3
+			
+			1988
+			Wiley
+			New York, NY
+		
+	
+	D'Azzo, J. J., and Houpis, H., Linear Control System Analysis & Design, McGraw-Hill Book Company, New York, NY, 1988.
+
+
+
+	
+		Mesoscale Weather Prediction with the RUC Hybrid Isentropic–Terrain-Following Coordinate Model
+		
+			StanleyGBenjamin
+		
+		
+			GeorgAGrell
+		
+		
+			JohnMBrown
+		
+		
+			TatianaGSmirnova
+		
+		
+			RainerBleck
+		
+		10.1175/1520-0493(2004)132<0473:mwpwtr>2.0.co;2
+		
+	
+	
+		Monthly Weather Review
+		Mon. Wea. Rev.
+		0027-0644
+		1520-0493
+		
+			132
+			2
+			
+			February 2004
+			American Meteorological Society
+		
+	
+	cited 20 July 2011
+	Benjamin, S. G., Grell, G. G., Brown, J. M., Smirnova, T. G., and Bleck, R., "Mesoscale Weather Prediction with the RUC Hybrid Isentropic-Terrain-Following Coordinate Model," Monthly Weather Review, Vol. 132, February 2004, pp. 473-494. 10 URL: http://www2.bombardier.com/en/3_0/3_2/pdf/global_5000_factsheet.pdf [cited 20 July 2011].
+
+
+				
+			
+		
+	

file122.txt

@@ -0,0 +1,386 @@
+
+	
+	
+		
+Introductionhe motivation for developing an aircraft performance model of a multirotor Urban Air Mobility (UAM) aircraft is to enable generation of trajectories, which are needed for concept evaluation studies and in decision support tools.One of the challenges facing the future Air Traffic Management (ATM) system is integrating new entrants like UAM, Unmanned Aerial Systems (UAS), supersonic aircraft and space-launch vehicles in the U. S. National Airspace System (NAS) with legacy commercial subsonic and General Aviation (GA), while accommodating the business and operational objectives of all stakeholders.The ability to simulate UAM trajectories is essential because they will be used both by the flight operator and the ATM service provider for planning and control purposes.For example, the flight operator will use generated trajectories to determine route of flight, energy consumption, cruise speed, cruise altitude and time of arrival considering forecast winds aloft and alternative landing sites (airports and vertiports) that can be reached from locations along the route for creating flight-plan alternatives.The service provider will employ trajectories for separation assurance and traffic flow management.Separation assurance requires predicting conflicts, generating resolution advisories and evaluating them.Traffic flow management requires forecasting traffic demand at constrained resources-airspace and landing sites-and for determining flow control initiatives such as ground-hold, metering and scheduling needed for preventing the available capacity from being exceeded.In the past several years, different designs have been proposed for electric vertical takeoff and landing (eVTOL) aircraft by the industry.Designs by Ehang (https://www.ehang.com/ehangaav/),Lilium (https://www.lilium.com/),Joby (https://www.jobyaviation.com/),Kitty Hawk (https://kittyhawk.aero/),Boeing-Aurora (https://www.aurora.aero/)and Airbus-Acubed (https://acubed.airbus.com/)have included rotary wing aircraft with multiple rotors, fixed-wing multi-engine tiltrotor, fixed-wing tiltrotor with rear push-propellers, fixed-wing aircraft with multiple lift fans (upward pointing fixed propellers) and rear push-propellers, and multi-engine tiltwing.This paper presents a model of the rotary wing aircraft with multiple rotors and employs it for trajectory synthesis.A trajectory can be generated by driving the mathematical model of the aircraft dynamics with a control system to follow the desired path.The output of the feedback control system for eliminating the difference between the desired state and the estimated state of the aircraft is fed as input to the model of the aircraft dynamics for temporal evolution of the state and outputs.This approach is adequate for traffic simulation and prediction needed for ATM applications.The other related applications of interest to the aviation community are estimator and control system design for estimating the state and controlling the motion of the physical aircraft, respectively.Both these applications require a mathematical model of the aircraft dynamics.Several six-degree-of-freedom models with different levels of complexity that include both the rotational and translational dynamics of a quadrotor aircraft, and controllers for them, are available in Ref. 1 through 5.The control problem has been solved in several different ways.Reference 1 employs nonlinear programming to generate the trajectory that minimizes the total impulse while staying outside a geographical region and subjecting the observer to less than or equal to the specified sound pressure level.Reference 2 uses an adaptive backstepping procedure for horizontal and vertical path tracking and attitude control.Reference 3 uses piecewise Proportional-Integral (PI) control in the along-track direction and Proportional-Integral-Derivative (PID) control in the cross-track direction for path tracking, which generates attitude reference commands.The attitude control law using feedback linearization and acceleration feedback is then used to track the attitude reference commands to orient the vehicle's thrust for generating the acceleration needed for tracking the desired path.Reference 5 linearizes the dynamics and control about an operating point to develop a Linear-Quadratic-Regulator (LQR) for flight control.The main difficulty with including rotational dynamics for trajectory generation, as in Refs. 1 through 5, is that the moment of inertia tensor needs to be known.With multirotor UAM still being designed and being developed, these data are unavailable.Furthermore, aircraft manufacturers are hesitant to share data they consider proprietary and not required for the operator to fly the aircraft.In addition, an attitude controller for trajectory generation for ATM applications is unnecessary because the angular rates and the resulting attitude are not observable from position data obtained from air traffic control radar systems or from position data determined onboard the aircraft and broadcast to the ground.Only six states-latitude, longitude, altitude, speed, heading and climb/descent rate-are observable from surveillance data; therefore, they are the ones employed for ATM decision support.For example, Time-Based Flow Management (TBFM), which is used for scheduling arrivals to major U. S. airports, uses a trajectory synthesis procedure based on a point-mass model that only considers the translational dynamics of the aircraft to predict estimated times of arrival at the metering locations (see Ref. 6).Systems for simulating air traffic such as NASA's Future ATM Concepts Evaluation Tool (FACET) (see Ref. 7), Airspace Concept Evaluation System (ACES) (see Ref. 8) and ATM-X Testbed use Eurocontrol's Base-of-Aircraft-Data (BADA) (see Ref. 9) aircraft performance model for generating the flight trajectory without considering rotational dynamics.Due to these reasons, this paper only models the translational dynamics of multirotor aircraft and derives the controls using the states and the state derivatives, where the state derivatives are obtained using Proportional and Proportional-Derivative (PD) controllers.The Proportional controller is used for tracking the airspeed and the PD controller is used for tracking the horizontal path.The vertical profile is specified.It is shown that this model driven by the non-linear controller developed in this paper can generate trajectories for UAM mission requirements outlined in Ref. 10.The rest of the paper is organized as follows.Section II provides a brief description of a conceptual aircraft model described in Ref. 11.The equations of motion are discussed in Section III.Computation of control variables-thrust, thrust vector angle and bank angle-is provided in Section IV.Prescription of heading and flight-path angles for following horizontal and vertical paths in the presence of wind is examined in Section V. Section VI describes the trajectory generation process and presents results for a flight scenario.The paper is concluded in Section VII.
+II. Multirotor Aircraft ModelParameters needed for generating the trajectory using the aircraft performance model presented in this paper are obtained from the battery-powered electric quadcopter concept vehicle described in Ref. 11.To size and analyze the aircraft designs, Ref. 11 used the NASA Design and Analysis of Rotorcraft (NDARC) tool.The NDARC tool provides models for rotors and lifting surfaces, engines and motors, and energy storage and conversion for requirements and technology trades for aircraft design.The quadcopter concept vehicle employs motors, shafts and gearboxes to power the four rotors.The aircraft performance model in this paper does not consider power topology of the quadcopter for modeling the power consumption and thrust generation by individual rotors; the total thrust and power consumption are modeled.The quadcopter aircraft parameters are listed in Table 1.
+III. Equations of MotionThe motion of aircraft, modeled as a point-mass, is described by the following three equations (see Ref. 12):gn V h R ) ( 1 + =   (1) ge V h R   cos ) ( 1 + =  (2) and h V h =  (3)where  is the latitude,  is the longitude, h is the geometric altitude and R is the mean radius of the Earth.The magnitude of the airmass-relative acceleration resulting from the thrust, drag, and gravitational forces on the multirotor aircraft modeled as a point-mass, defined in Fig. 2, is cos sinTD Vg m   - =- (4)where V is the airmass-relative speed (true airspeed), T is the thrust, D is the drag,  is the thrust vector angle with respect to the airmass-relative velocity, m is the mass, g is the acceleration due to gravity and  is the airmass- relative flight-path angle.Note that Eq. ( 4) assumes the wind acceleration to be zero.s V is defined in terms of true airspeed and flight-path angle as cos V V s =(5)The north, east and vertical components of the ground-relative aircraft velocity in terms of the true airspeed, flightpath angle, heading angle and wind terms are Using Eq. ( 6) and ( 7), groundspeed is obtained as2 2 ) sin cos ( ) cos cos ( e n g W V W V V + + + =     (9)The state equations for airmass-relative heading angle and flight-path angle assuming wind acceleration to be zero can be written as follows:sin sin cosT mV    = (10)and sin cos cosTg mV V     =- (11)where  is the airmass-relative bank angle.Equations ( 4), ( 10) and ( 11) are derived assuming flat Earth, constant gravitational acceleration and slowly changing mass.Mass is constant for battery-powered electric aircraft.To summarize, the state equations for the point-mass model of a multirotor aircraft are given by Eqs.(1) through (3), Eq. ( 4) and Eqs.(10) and (11), where the states are: ,  , h ,V ,  and  .The controls are: T ,  and  , which can be determined as shown in the next section.
+IV. Control ComputationLet the commanded airspeed, heading angle and flight-path angle be denoted by  4), ( 10) and ( 11), the following equations can be written for computing the control variables.cos sinc c c T mV D mg  = + +(12)sin sin cosc c c c T mV     =(13)and sin cos cosc c c c T mV mg     =+(14)Given that three equations are available for three unknowns, the unknowns can be determined as follows.Dividing Eq. ( 13) by Eq. ( 14), the commanded bank angle can be determined as Squaring and adding Eq. ( 13) and ( 14), taking the square root and dividing by Eq. ( 12), the thrust vector angle can be determined as22 1 ( cos ) ( cos ) tan sin cc c c m V V g mV D mg       -   + +  =  ++  (16)and finally, squaring and adding Eqs. ( 12) through ( 14) and taking the square root of the result,c pv c V K V V =- (18)where pv K is the proportional gain and V is the estimated airspeed based on measured airspeed.Similarly, a proportional-derivative controller could be employed for generating commands for following the commanded heading angle.With p K  as the proportional gain and d K  as the derivative gain of the heading controller, and  and  as the estimated heading angle and the heading-angle rate, respectively, ()c p c d KK      = --(19)In this study, the vertical profile is defined by specifying c  in the different phases of flight; it is assumed that the commanded flight-path angle is instantly achieved.Thus,  , the estimated flight-path angle, is set equal to c  .Furthermore, the commanded flight-path angle rate is set to zero,0 c  =(20)The commanded angular acceleration in Eq. ( 19) is integrated forward in time to determine the commanded headingangle rate for use in Eqs.(15) through (17).Note that limits can be placed on the bank angle and acceleration commands generated by Eq. (15) and Eq. ( 18), respectively, to accommodate passenger ride quality requirements.Application of proportional control, especially in Eq. ( 19), can be a bit tricky for large angular differences because of the two ways of turning from one direction to the other-clockwise and counterclockwise, one of which results in a smaller angular rotation; therefore, the smaller angular difference and the direction of the turn needs to be determined.It is straightforward to specify the airspeed in different phases of flight provided enough thrust can be generated for overcoming the drag, gravitational and wind forces for accelerating the aircraft.Specification of heading and flightpath angle is based on the horizontal and vertical paths to be followed in the presence of wind.This is discussed in the next section.
+V. Reference Command GenerationTo stay on course in a wind field, the aircraft has to crab such that the across-track component of the wind is cancelled.This assumes that the aircraft has enough speed to counter the wind.The airmass-relative heading angle needed to stay on the path specified by the course angle g  is obtained from the two relations based on Fig. 1:g g n V W V    cos cos cos = + (21) and g g e V W V    sin sin cos = + (22) as         - + = -      cos cos sin sin 1 V W W g e g n g (23)from the trigonometric identity resulting from the difference of Eq. ( 21) multiplied by sin g  and Eq. ( 22) multiplied by cos g  .To compute the desired airmass-relative heading angle using Eq. ( 23), g  has to be specified using a navigation procedure like great-circle navigation to follow the desired route (horizontal path) of flight.The closedloop great-circle navigation law is given in the following form in Ref. 13:      - - - = - ) cos( sin cos cos sin cos ) sin( tan 1           c c c c c g (24)where the current position of the aircraft is ) , (
+ and the commanded position is ) , ( c c   for heading to the next waypoint along the route or directly to the destination.Airmass-relative heading angle command can now be computed using estimated values of latitude and longitude- and  -in Eq. ( 24), and airspeed and flight-path angle in Eq. (23).Thus, 1 sin cos sin cosn g e g cg WW V    - -  =+  (25)This commanded heading angle value is used in Eq. ( 19).Like the heading angle that needs to be specified for following the horizontal path, flight-path angle needs to be specified for following the vertical path.For example, to maintain level flight in cruise, Eq. ( 8) prescribes the flightpath angle to be1 sin h c W V  - =-  (26)which means that a vertical rate has to be generated to counter vertical wind.The commanded airmass-relative flightpath angle can also be computed for following a climb/descent path prescribed via ground-relative flight-path angle.The ground-relative flight-path angle can be specified as1 tan c g GC hh d  - - =   (27)where c h is the commanded altitude, h is the estimated (current) altitude, and the great-circle distanceGC d is   1 cos sin sin cos( ) cos cos GC c c c dR       - = + -(28)The numerator and denominator in Eq. ( 27) are altitude-to-go and distance-to-go, respectively.Using Eqs. ( 8) and ( 9), the relation between  and g  is 22 sin tan ( cos cos ) ( cos sin )h g ne VW V W V W       + = + + + (29)Because the commanded heading angle is available from Eq. ( 25), commanded airspeed c V is specified and g  is available from Eq. ( 27),  -and equivalently c  -is the only unknown in Eq. ( 29).If the vertical component of the wind is ignored, Eq. ( 29) reduces to a quadratic, which can be easily solved to determine  computed via Eq. ( 29 is assumed to be instantly achieved; it is used directly in Eqs. ( 15) through (17) for computing the controls.For vertical descent to the airport/vertiport surface-final descent,c c h n e V h W W W = - + +(34)The commanded descent rate c h in Eqs. ( 33) and (34) needs to be specified for landing the aircraft and reducing the descent rate to zero (or close to zero) on touchdown.Starting with a descent rate of h at altitude h , the deceleration required for reducing the descent rate to zero after traversal of distance h is obtained using the kinematic relation between acceleration, initial and final speeds and distance as follows.( )2 2 c h h h =(35)The feedback control law given by Eq. ( 35) assumes positive descent rateh has a negative value-and positive h .Because the deceleration required depends on the initial descent rate according to Eq. ( 35), the initial descent rate needs to be such that the deceleration does not exceed the specified deceleration limit.If the starting altitude for vertical descent for landing is li h and the deceleration limit is lim a , the initial descent rate according to Eq. ( 35) needs to belim 2 li li h a h =- (36)This completes the description of the variables that need to be specified for computing the controls in all phases of flight.The trajectory can be determined using either the thrust, thrust vector angle and bank angle-directly as controlsor that resulting from models of thrust, thrust vector angle and bank angle dynamics-driven by the commanded controls-in the equations of motion.The trajectory generation process and an example of the generated trajectory along with the time histories of the control variables are discussed in the next section.
+VI. Trajectory GenerationThe complete procedure for trajectory generation is summarized in Fig. 3.The process is initiated by reading the flight plan, which provides basic information about the type of aircraft, equipage, origin, destination, route of flight, cruise altitude and cruise speed.Flight plans also specify alternative airports and provide additional information required for operating in the U. S. airspace.The flight procedure (also called airline procedure) describes how the aircraft is to be flown.For example, it might stipulate that after takeoff from the vertiport, the aircraft will climb vertically to 50 feet altitude maintaining a climb rate of 500 feet/minute.It will then climb to the cruise altitude of 2,000 feet while maintaining a 10-degree climb angle and a speed of 60 knots.On reaching the cruise altitude, the aircraft will accelerate to the cruise speed of 98 knots and maintain it, and so on.The next step is to read the initial conditions like the location of the flight, altitude and speed for example.After reading the required data, the recursive part of the process is begun by determining the mode of flight, where the modes of flight include: on ground, takeoff and initial climb, climb, cruise, initial descent, approach and final descent to landing.These modes are determined as a function of altitude, speed and location of the flight with respect to reference locations such as the origin and destination of flight.Based on the mode of flight, reference commands are generated as shown in Section V using the measured and estimated states, and the desired states needed for tracking the path and speed profiles.The reference commands and the estimated states are then used to determine the controls as discussed in Section IV.The computed controls are input to the equations of motion, described in Section III, and integrated forward in time to determine the true states.Because the true states of the aircraft are unavailable, a combination of sensors such as Global Positioning System (GPS), accelerometers and gyros are needed along with state estimators running on onboard computers to estimate them.The estimated states are used for onboard control computations for controlling the motion of the physical aircraft.In the ATM systems, observable states are estimated using the aircraft position data acquired by surveillance with radar and transponderbased systems and received from aircraft equipped with an Automatic Dependent Surveillance system.Depending on the degree of realism desired, the "Estimate states" step in Fig. 3 can be as simple as (a) setting the estimated states to the true states, or it can be a bit more complex as (b) adding noise and bias to the true states according to estimation error distributions, or it can be as realistic as (c) modeling the surveillance sensors with their sources of errors and using an Extended Kalman Filter for state estimation.The estimated and the true states are stored for further analysis.The estimated states are then used for determining the mode of flight in the next recursive step.The recursive process is terminated when the mode of flight transitions to "on ground" at the destination.The trajectory consisting of the temporal history of the positionlatitude, longitude and altitude-is output in the final step of the trajectory generation process outlined in Fig. 3.The procedure summarized in Fig. 3 was utilized for computing the trajectory of the six-passenger multirotor aircraft with takeoff mass of 2,940 kg (see Table 1 for other parameters) flying from Palo Alto Airport (PAO), California to San Martin Airport (E16), California.The two airports are 34 nautical miles apart.PAO is located at 37.46 degrees latitude and -122.11degrees longitude, and E16 is located at 37.08 degrees latitude and -121.60 degrees longitude.The simulated horizontal trajectory is shown in Fig. 4 and the vertical trajectory is shown in Fig. 5.The first four minutes and the last five minutes of flight-path angle time history are shown in Figs.6a and6b, respectively.Figure 6a shows the aircraft climbs vertically (90 degrees), climbs at a 10-degree flight-path angle and then levels off for cruise.Figure 6b shows the aircraft initially descends at a -10-degree ground-relative flight-path angle (see Eq. ( 27)) after the end of cruise.The aircraft then reduces speed to the final descent speed (according to Eqs. (34) and (36)) and descends at a steeper flight-path angle (about -30 degrees) to reduce the descent speed to zero on touchdown   7a and7b show the first four minutes and the last five minutes of the thrust time history.Observe the initial variation in thrust prior to one minute in Fig. 7a.It is due to the bank angle required for reorienting the heading from the initial heading to that required for countering the crosswind component of the wind and heading towards the destination.This figure also shows the increase in thrust between two and three minutes for accelerating from 60 knots to 98 knots in cruise.Figure 7b shows an increase in thrust between 25 and 26 minutes to slowdown from 98 knots to 60 knots, and between 28 and 29 minutes to slowdown from 60 knots to the final descent speed.Figures 8a and8b show the corresponding first four minutes and the last five minutes of the thrust vector angle w.r.t.horizontal (sum of  and  ; see Fig. 2) time history.Figure 8a shows the thrust vector angle during initial climb, climb and cruise.Observe the thrust vector angle exceeds 90 degrees between 25 and 26 minutes and between 28 and 29 minutes in Fig. 8b.Angle greater than 90 degrees is for pointing the thrust in the opposite direction of motion to slowdown the aircraft.Figures 9a and9b show the early and the later parts of the bank    angle time history.The initial bank angle is for turning the aircraft from its initial heading to that required for countering the crosswind and flying towards the destination.Note the correspondence between the bank angle in Fig.     9a and the thrust prior to one minute in Fig. 7a.Figures 10 and 11 show the airspeed and groundspeed time histories, respectively.Observe in Figs. 10 and 11 that in the approach segment prior to final vertical descent, the airspeed is over 20 knots while the groundspeed is less than three knots against a headwind of 20 knots.The groundspeed is zero during final descent; airspeed is reduced to zero on touchdown.
+VII. ConclusionsThe results in this paper demonstrate that the performance of the point-mass model of the multirotor electric vertical takeoff and landing aircraft driven by the controllers developed in this paper is suitable for generating trajectories for meeting urban air mobility mission requirements.The ability to generate these trajectories will enable concept evaluation and development of decision support tools for accommodating urban air mobility operations in the air traffic management system.Vare the north and east components of the ground-relative aircraft velocity.h V is the ground-relative climb or descent rate depending on whether it is positive or negative.The horizontal velocity of the aircraft with respect to the ground is the resultant of the horizontal components of the airmass-relative aircraft velocity and the ground-relative wind velocity.This relationship is shown in Fig.1, where g V and s W are the magnitudes of the horizontal components of the ground-relative aircraft velocity and the wind velocity, and s V is the magnitude of the horizontal component of the airmass-relative aircraft velocity.g  is the heading angle of the ground-relative aircraft velocity with respect to the local north direction. and w  are the heading angles of the airmass-relative aircraft velocity and ground-relative wind velocity, respectively, also with respect to the local north direction.north, east and up components of the ground-relative wind velocity vector.
+Figure 1 .1Figure 1.Velocity triangle.
+, respectively.Let the commanded thrust, thrust vector angle and the bank angle be denoted by c T , c  and c  , respectively.Based on the relationships in Eqs. (
+Figure 2 .2Figure 2. Forces on the multirotor aircraft.
+Drag-D -in Eqs.(16) and (17) is a function of V and density of air, which is determined based on estimated altitude.The state variables in Eqs.(15) through (17) are estimated quantities.needed in the above equations can be determined by defining control laws for controlling the corresponding state variables.Cruise control for the airspeed to follow the commanded airspeed leads to ()
+c.Keeping the vertical component of wind results in a quartic equation.Observe that without the wind terms, Eq. (29) simplifies to cg
+gVhas to be maintained at -90 degrees.= means that both the north and east components of the ground-relative aircraft velocity must be zero.Setting says the airmass-relative heading angle should be either in the direction of the horizontal component of the wind velocity or opposite to it.See Fig.1for a pictorial depiction of the horizontal component of wind velocity.Flying into the wind-opposite direction to the horizontal component of the wind velocity-enables zero groundspeed to be achieved by applying the horizontal component of the airmass-relative velocity in the forward direction.Combining Eq. (32) with Eqs.(8) and (3),
+Figure 3 .3Figure 3. Trajectory generation process.
+Figure 4 .Figure 5 .45Figure 4. Horizontal trajectory.Figure 5. Vertical trajectory.
+Figure 7b .7bFigure 7b.Last five minutes of the thrust time history.
+Figure 7a .7aFigure 7a.First four minutes of the thrust time history.
+Figure 6a .6aFigure 6a.First four minutes of the flight-path angle time history.
+Figure 6b .6bFigure 6b.Last five minutes of the flight-path angle time history.
+Figure 10 .10Figure 10.Airspeed time history.Figure 11.Groundspeed time history.
+Figure 9a .9aFigure 9a.First four minutes of the bank angle time history.
+Figure 9b .9bFigure 9b.Last five minutes of the bank angle time history.
+Figure 8a .8aFigure 8a.First four minutes of thrust vector angle time history.
+Figure 8b .8bFigure 8b.Last five minutes of the thrust vector angle time history.
+Table 1 . Quadcopter eVTOL conceptual model parameters.1ParameterValueStructural Mass (excluding the battery)1,684 kgMass of Single Passenger91 kgMaximum Number of Occupants (includes the pilot)6 personsBattery Mass710 kgMaximum Mass (with six occupants)2,940 kgUseful Battery Capacity (80% Battery Capacity)295,778 watt hoursMaximum Deliverable Power501,110 wattsDrag Coefficient1.1984 dimensionlessReference Area1 square meter
+		
+		
+			
+
+				
+
+
+	
+		Multidisciplinary Optimization of Urban-Air-Mobility Class Aircraft Trajectories with Acoustic Constraints
+		
+			RobertDFalck
+		
+		
+			DanielIngraham
+		
+		
+			EliotAretskin-Hariton
+		
+		10.2514/6.2018-4985
+	
+	
+		2018 AIAA/IEEE Electric Aircraft Technologies Symposium
+		Cincinnati, OH
+		
+			American Institute of Aeronautics and Astronautics
+			July 9-11, 2018
+		
+	
+	Falck, R. D., Ingraham, D., and Aretskin-Hariton, E., "Multidisciplinary Optimization of Urban-Air-Mobility Class Aircraft Trajectories with Acoustic Constraints," Proc. AIAA/IEEE Electric Aircraft Technologies Symposium, Cincinnati, OH, July 9-11, 2018.
+
+
+
+	
+		Modeling and Adaptive Flight Control for Quadrotor Trajectory Tracking
+		
+			HakimBouadi
+		
+		
+			FMora-Camino
+		
+		10.2514/1.c034477
+	
+	
+		Journal of Aircraft
+		Journal of Aircraft
+		0021-8669
+		1533-3868
+		
+			55
+			2
+			
+			March-April, 2018
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	Bouadi, H., and Mora-Camino, F., "Modeling and Adaptive Flight Control for Quadrotor Trajectory Tracking," Journal of Aircraft, Vol. 55, No. 2, March-April, 2018.
+
+
+
+	
+		Quadrotor Helicopter Trajectory Tracking Control
+		
+			GabrielHoffmann
+		
+		
+			StevenWaslander
+		
+		
+			ClaireTomlin
+		
+		10.2514/6.2008-7410
+	
+	
+		AIAA Guidance, Navigation and Control Conference and Exhibit
+		Honolulu, HI
+		
+			American Institute of Aeronautics and Astronautics
+			August 18-21, 2008
+		
+	
+	Hoffmann, G. M., Waslander, S. L., and Tomlin, C. J., "Quadrotor Trajectory Tracking Control," Proc. AIAA Guidance, Navigation, and Control Conference, Honolulu, HI, August 18-21, 2008.
+
+
+
+	
+		Quadrotor Helicopter Flight Dynamics and Control: Theory and Experiment
+		
+			GabrielHoffmann
+		
+		
+			HaomiaoHuang
+		
+		
+			StevenWaslander
+		
+		
+			ClaireTomlin
+		
+		10.2514/6.2007-6461
+	
+	
+		AIAA Guidance, Navigation and Control Conference and Exhibit
+		Hilton Head, SC
+		
+			American Institute of Aeronautics and Astronautics
+			August 20-23, 2007
+		
+	
+	Hoffmann, G. M., Huang, H., Waslander, S. L., and Tomlin, C. J., "Quadrotor Flight Dynamics and Control: Theory and Experiment," Proc. AIAA Guidance, Navigation, and Control Conference and Exhibit, Hilton Head, SC, August 20-23, 2007.
+
+
+
+	
+		Dynamics and Optimal Control of Quadrotor Platform
+		
+			RadoslawZawiski
+		
+		
+			MarianBlachuta
+		
+		10.2514/6.2012-4915
+	
+	
+		AIAA Guidance, Navigation, and Control Conference
+		Minneapolis, MN
+		
+			American Institute of Aeronautics and Astronautics
+			August 13-16, 2012
+		
+	
+	Zawiski, R., and Blachuta, M., "Dynamics and Optimal Control of Quadrotor Platform," Proc. AIAA Guidance, Navigation, and Control Conference, Minneapolis, MN, August 13-16, 2012.
+
+
+
+	
+		The Dynamic Planner: The Sequencer, Scheduler, and Runway Allocator for Air Traffic Control Automation
+		
+			GLWong
+		
+		cited: 11/20/2019
+		
+	
+	
+		NASA TM-2000-209586, National Aeronautics and Space Administration
+		Ames Research Center, Moffett Field, CA
+		
+			April 2000
+			
+		
+	
+	Wong, G. L., "The Dynamic Planner: The Sequencer, Scheduler, and Runway Allocator for Air Traffic Control Automation," NASA TM-2000-209586, National Aeronautics and Space Administration, Ames Research Center, Moffett Field, CA 94035-1000, April 2000, URL: http://www.aviationsystemsdivision.arc.nasa.gov/publications/full_list_by_author.shtml [cited: 11/20/2019]
+
+
+
+	
+		FACET: Future ATM Concepts Evaluation Tool
+		
+			KarlDBilimoria
+		
+		
+			BanavarSridhar
+		
+		
+			ShonRGrabbe
+		
+		
+			GanoBChatterji
+		
+		
+			KapilSSheth
+		
+		10.2514/atcq.9.1.1
+		cited: 11/20/2019] 8
+		
+	
+	
+		Air Traffic Control Quarterly
+		Air Traffic Control Quarterly
+		1064-3818
+		2472-5757
+		
+			9
+			1
+			
+			2001
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	Bilimoria, K., Sridhar, B., Chatterji, G., Sheth, K., and Grabbe, S., "FACET: Future ATM Concepts Evaluation Tool," Air Traffic Control Quarterly, Vol. 9, No. 1, pp. 1-20, 2001, URL: http://www.aviationsystemsdivision.arc.nasa.gov/publications/full_list_by_author.shtml [cited: 11/20/2019] 8
+
+
+
+	
+		Build 4 of the Airspace Concept Evaluation System
+		
+			LarryMeyn
+		
+		
+			RobertWindhorst
+		
+		
+			KarlinRoth
+		
+		
+			DonaldVan Drei
+		
+		
+			GregKubat
+		
+		
+			VikramManikonda
+		
+		
+			SharleneRoney
+		
+		
+			GeorgeHunter
+		
+		
+			AlexHuang
+		
+		
+			GeorgeCouluris
+		
+		10.2514/6.2006-6110
+	
+	
+		AIAA Modeling and Simulation Technologies Conference and Exhibit
+		Keystone, Colorado
+		
+			American Institute of Aeronautics and Astronautics
+			Aug. 21-24, 2006
+		
+	
+	Meyn, L., Windhorst, R., Roth, K., Drei, D. V., Kubat, G., Manikonda, V., Roney, S., Hunter, G., Huang, A., and Couluris, G., "Build 4 of the Airspace Concept Evaluation System," Proc. AIAA Modeling and Simulation Technologies Conference and Exhibit, Keystone, Colorado, Aug. 21-24, 2006.
+
+
+
+	
+		Sensitivity Analysis of Aviation Environmental Impacts for the Base of Aircraft Data (BADA) Family 4
+		10.2514/6.2021-0457.vid
+	
+	
+		Eurocontrol Experimental Centre
+		
+			10/04
+			July 2004
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	"User Manual for Base of Aircraft Data (BADA) Revision 3.6," Eec note no. 10/04, Eurocontrol Experimental Centre, July 2004.
+
+
+
+	
+		A Proposed Approach to Studying Urban Air Mobility Missions Including and Initial Exploration of Mission Requirements
+		
+			MDPatterson
+		
+		
+			KRAntcliff
+		
+		
+			LWKohlman
+		
+	
+	
+		Proc. American Helicopter Society International 74th Annual Forum & Technology Display
+		American Helicopter Society International 74th Annual Forum & Technology DisplayPhoenix, AZ
+		
+			May 14-17, 2018
+		
+	
+	Patterson, M. D., Antcliff, K. R., and Kohlman, L. W., "A Proposed Approach to Studying Urban Air Mobility Missions Including and Initial Exploration of Mission Requirements," Proc. American Helicopter Society International 74th Annual Forum & Technology Display, Phoenix, AZ, May 14-17, 2018.
+
+
+
+	
+		VTOL Urban Air Mobility Concept Vehicles for Technology Development
+		
+			ChristopherSilva
+		
+		
+			WayneRJohnson
+		
+		
+			EduardoSolis
+		
+		
+			MichaelDPatterson
+		
+		
+			KevinRAntcliff
+		
+		10.2514/6.2018-3847
+	
+	
+		2018 Aviation Technology, Integration, and Operations Conference
+		Atlanta, GA
+		
+			American Institute of Aeronautics and Astronautics
+			June 25-29, 2018
+		
+	
+	Silva, C., Johnson, W., Antcliff, K. R., and Patterson, M. D., "VTOL Urban Air Mobility Concept Vehicles for Technology Development," Proc. AIAA Aviation Technology, Integration, and Operations Conference, Atlanta, GA, June 25-29, 2018.
+
+
+
+	
+		Fuel Burn Estimation Using Real Track Data
+		
+			GanoBChatterji
+		
+		10.2514/6.2011-6881
+	
+	
+		11th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference
+		Virginia Beach, VA
+		
+			American Institute of Aeronautics and Astronautics
+			September 20-22, 2011
+		
+	
+	AIAA 2011-6881
+	Chatterji, G. B., "Fuel Burn Estimation Using Real Track Data," AIAA 2011-6881, Proc. AIAA Aviation Technology, Integration, and Operations Conference, Virginia Beach, VA, September 20-22, 2011.
+
+
+
+	
+		En-route flight trajectory prediction for conflict avoidance and traffic management
+		
+			GBChatterji
+		
+		
+			BSridhar
+		
+		
+			KDBilimoria
+		
+		10.2514/6.1996-3766
+	
+	
+		Guidance, Navigation, and Control Conference
+		San Diego, CA
+		
+			American Institute of Aeronautics and Astronautics
+			July 29-31, 1996
+		
+	
+	Chatterji, G. B., Sridhar, B., and Bilimoria, K. D., "En-route Flight Trajectory Prediction for Conflict Avoidance and Traffic Management," AIAA 96-3766, Proc. AIAA Guidance, Navigation, and Control Conference, San Diego, CA, July 29-31, 1996.
+
+
+				
+			
+		
+	

file123.txt

@@ -0,0 +1,342 @@
+
+	
+	
+		
+I. Introductionirspace sectors have evolved over decades to assist the human controller organize flights for safe and efficient operations through the airspace.Unfortunately, the resulting sector design's inflexibility makes it difficult for it to adapt to changing weather and traffic conditions.With limited means for redistributing capacity in the airspace, traffic flow management techniques, such as delaying aircraft on the ground, are employed to reduce traffic in the affected airspace.Since this leads to delays, reconfiguring the airspace to dynamically adjust its capacity to where and when it is most needed has been proposed as an alternative. 1][4][5][6][7][8][9] This paper examines whether airspace partitions created with several days of data are robust, where robust means that they can be used on other similar days, and it examines the benefit of using different partitions at different times of the day.The focus of earlier airspace partitioning research was on partitions generated with at most one-day of traffic data; the issue of whether the partitions could be used with traffic data from other days was not of concern.The benefit, measured by reduced sector-hours, of using different configurations generated by combining sectors and by combining altitude slices has been examined in Refs. 10 and 11, respectively.This same metric has been used here to show the tradeoff between reduced sector-hours and the number of times the partitions are changed in a day.The Mixed Integer Linear Programming (MILP) method described in Ref. 9 is used with traffic data from ten high-volume low-delay days to design sectors in Fort Worth, Cleveland and Los Angeles centers.These centers were chosen because they are located in different regions of the U. S. and experience very different traffic patterns.A comparison of peak traffic-counts in the sectors for traffic from 57 days including the ten days used in the design shows that the sector configurations in these centers are robust.Results show that sector configurations created with two-hour traffic data can be used for duration of six to twelve hours without exceeding the peak traffic-count requirement.Most of the sector-hour reduction is obtained by using one sector configuration during the daytime hours and one during the nighttime hours compared to using a single configuration for the entire day.Further reduction is achieved if three sector configurations are used during the day.Section II describes the actual air traffic dataset consisting of 57 high-volume low-delay days, out of which, ten days are used for creating the sector configurations.The entire dataset is used for evaluating the sector configurations.Section III discusses the MILP method, and Section IV describes the robust sectorization and validation method for creating sectors.Validation results are discussed in Section V. Tradeoff between sectorhours and the number of configuration changes is discussed in Section VI, and the paper is concluded in Section VII.
+II. Air Traffic DatasetThe analysis and results discussed in this article are based on air traffic data from high-volume low-delay days.High-volume traffic is usually associated with weekdays.Delays are low on days on which the flights are relatively unaffected by weather and congestion caused rerouting and ground holds.Most aircraft stay on their filed route of flight and are on-time with respect to their schedule.To identify such days, delay data for all the days in 2007 were obtained from the Federal Aviation Administration's Air Traffic Operations Network (OPSNET) database.The days were then categorized based on total domestic departure-counts and total time delay in minutes using the multiple-metric K-Means classification method described in Ref. 12. Days were separated into nine groups based on the combination of traffic-volume ("low-volume," "medium-volume," "high-volume") and delay ("lowdelay," "medium-delay," and "high-delay").Figure 1 shows the scatter plot of the 57 days that were selected for this study.Ten of the 57 days, marked in circles, were used for designing the sectors.We will refer to these days as the training set.The ones marked with triangles are the remaining 47 days that were used for evaluating the robustness of these sectors, referred to as test days.Figure 2 shows the average, upper and lower bounds of the number of aircraft in the Fort Worth Center airspace as a function of time for the ten training days.The numbers of training and test days for each day of the week are listed in Table 1.Aircraft position data from the ten training days for each two-hour time period were used in the MILP sector design method, described in Ref. 9, to create 12 sector configurations spanning the 24-hour time period.The MILP sector design method is briefly discussed next.
+III. Mixed Integer Linear Programming MethodThe MILP method discussed in this section assumes a hexagonal tessellation of the airspace.Such a tessellation with tiles marked with numbers uniquely identifying them is shown in Fig. 3. Tiles marked with the letter " s " are special tiles called "seed" tiles.The setup phase of the algorithm counts the total number of aircraft located within the tile along with the total number of aircraft that cross each of the six sides of the tile for the duration of interest.The direction of tile boundary crossing is ignored.The seed tiles are also selected at this point.The optimization process clusters the hexagonal tiles together to form sectors by using a connection variable that represents a directional link between two tiles.This variable contains not only the identity of the linked tiles, but also the accumulated sum of aircraft counts of every tile upstream of that link.This accumulation of aircraft counts   is terminated at the "sink" tile, which is selected during the optimization process from among the pre-determined seed tiles.In this way, the value of the final link going into the sink tile, plus the sink tile's aircraft count equals the total aircraft count of that cluster of tiles.Figure 3 shows a notional solution of the optimization in which the directions of the links between adjacent tiles, all the way to the sink tile 1 are marked with arrows.All the tiles that contribute links to a particular sink tile, s ˆ, are said to belong to one region (sector) of airspace.Note that sink tile does not refer to an actual destination of aircraft.Rather it is a mathematical construct to aid in the formulation of the optimization problem.The solution phase of the algorithm is implemented by six basic constraints and an optimization function.The first constraint ensures that the link variable captures the accumulated number of aircraft upstream in the contiguous cluster of tiles.This is basically a conservation of aircraft constraint between a tile's incoming links, the one outbound link, and its own contribution of aircraft counts.The second constraint predicates that the total number of aircraft captured by the sum of incoming links to a sink tile, plus its own contribution of aircraft counts is constrained to be within 5% of the average number of aircraft, where the average number of aircraft is the sum of the number of aircraft in all tiles for the duration of interest divided by the desired number of sectors.This constraint leads to the creation of sectors with nearly equal numbers of aircraft.The third constraint asserts that the number of sink tiles should equal the desired number of sectors.The fourth constraint establishes that all non-sink tiles (including seed tiles that do not end up becoming sinks) have a single outbound link to an adjacent tile.The fifth constraint specifies that there is no outbound link from a sink tile.Finally, one of the most compelling reasons for using this method of tile clustering is that tile contiguity can be enforced by only allowing links to be formed between adjacent cells.This is the sixth constraint.In practice, this constraint can be implicitly enforced in the data structure utilized by the other constraints.The optimization function consists of the sum of the weighted outbound link values from each tile to its adjacent tiles.The weights are given by the boundary crossing counts computed during the setup phase.These weights are used to ensure that the link directions resulting from minimization of the optimization function are aligned along the major flows seen in the air traffic data.Other details of the MILP method are given in Ref. 9. A notional solution of the optimization shown in Fig. 3 can be viewed as a Directed Acyclic Graph (DAG) rooted at the sink tile 1 in Fig. 4. The graph is directed because the outbound links are defined from a tile to its adjacent tile; it is acyclic because single outbound-links (no backward links between adjacent tiles), conservation of aircraft, and single sink tile per sector prevent the formation of loops.Once the tiles are associated with sectors, a boundary smoothing algorithm described in Ref. 9 is used for generating the final sector boundaries.
+IV. Robust Sectorization and ValidationThis section describes the method for designing sectors using several days of air traffic data, selecting few sector configurations for the 24-hour period, and validating the design.The design is validated by playing back the test traffic data through the designed sectors and determining that the design criteria are not violated.The method is summarized in a block diagram in Fig. 5.The examples and results presented in this and subsequent sections are based on high-altitude traffic, which is above 24,000 feet altitude.6.The minimum and maximum numbers of aircraft during this interval were 23 and 81.This example shows that the traffic-counts in a sector can be unacceptably high when airspace is partitioned into few sectors.To ensure that the traffic-counts stay below a specified threshold in most instances, the airspace needs to be partitioned into more sectors.This is the motivation for step 508 that increases m by two.The previous steps are repeated to create histograms of the type in Fig. 6.    7. The last value of the graph in Fig. 7 is 2346, which is the total number of traffic-count samples in Fig. 6.Based on the last value, the 99.9 percentile value is 2344.The traffic-count corresponding to 2344 is 78 aircraft.This location is marked by an ' !' in Fig. 7.The central idea here is that if the Fort Worth Center airspace were to be partitioned into two sectors during the 6 p.m. to 8 p.m. CST time-interval, the probability is 99.9 percent that the traffic-count would be at or below 78 aircraft in a sector.Lower percentile values can be chosen to remove outlier traffic-count values.The process of computing the cumulative frequency and selecting a traffic-count value corresponding to the specified percentile is repeated for each of the nine sector configurations.The values obtained for nine sector configurations for the first two-hour time period (6 p.m. to 8 p.m. CST) are shown in Fig. 8.The number of sectors needed for ensuring 99.9% probability of traffic-counts staying below a specified traffic-count threshold can be obtained from the data presented in this figure .For example, at least 12 sectors would be needed if a threshold of 20 aircraft were chosen.This example shows that given a percentile value and a design threshold, a sector configuration can be chosen for the time-interval of interest.Step 510 checks if a sector configuration has been selected for the last two-hour interval.If not, t is incremented by two-hours in step 511, and a new timeinterval is determined in step 502.The entire process discussed thus far is repeated for this new time-interval.The result is a selection of 12 sector configurations, one for each two-hour time-interval, in step 509. Figure 9 shows a bar chart of the number of sectors in the configurations selected in the Fort Worth Center.Observe that the number of sectors correlates to the traffic-count shown in Fig. 2.In step 512, two or three sector configurations are chosen from the available 12.This selection is accomplished by organizing the configurations into a few groups and then identifying one representative configuration for each group.The K-Means algorithm discussed in Ref. 12 is used to organize the configurations into groups based on the number of sectors.For example, sector configurations for the first, second, and tenth two-hour time periods shown in Fig. 9 are placed in the first group, 3 rd through 6 th are placed in the second group and the remaining are placed in the third group, when three groups are desired.Based on these three groups, the sector configuration for the first two-hour period (6 p.m. to 8 p.m. CST) is selected for the duration of the first four-hours from 6 p.m. to 10 p.m. CST.Similarly, the sector configuration of six sectors for the 4 a.m. to 6 a.m.time-interval is applied for the eight-hour period spanning the 10 p.m. to 6 a.m.interval.Finally, the third configuration of 16 sectors for the 10 a.m. to 12 p.m. time interval is selected for the twelve-hour period from 6 a.m. to 6 p.m. CST.Note that the sector configuration for the tenth two-hour time period (12 p.m. to 2 p.m. CST) is a member of the first group since it has twelve sectors, but it lies between two members of the third group (10 a.m. to 12 p.m. and 2 p.m. to 4 p.m. configurations).Regardless, the representative member of the third group is used to cover this interval.Selected sector configurations and durations of their application for the Cleveland, Los Angeles and Fort Worth Centers are summarized in Table 2.Once representative sectors are selected in step 512, histograms of the type given in Fig. 6 are created for them in step 514 using aircraft position data from training set and test set days derived from step 513.In step 515 the    cumulative frequency values are computed based on the histograms provided by step 514 (see Fig. 7).These values are then used for determining traffic-counts corresponding to the percentile value (for example, 99.9) used in the design.The sector design is validated by determining if this traffic-count is above or below the specified threshold capacity value (for example, 20 aircraft) used in design.
+V. Validation ResultsResults of validation using three sector configurations of the Fort Worth, Cleveland and Los Angeles centers listed in Table 2 are described in this section.The three sector configurations of the Fort Worth Center are shown in Figs.10-12.Traffic data from the ten training and 47 test set days were played back through these configurations for the time durations noted in Table 2 to compute traffic-counts in the sectors.Histograms were then created with these traffic samples.Cumulative frequency values were computed using these histograms, and 99.9 percentile traffic-counts were determined.Figure 13 shows the histogram of 162,204 traffic-count samples for the Sector Configuration I shown in Fig. 10.The maximum number of aircraft in a sector was found to be 28 aircraft.The 99.9 percentile traffic-count was found to be 20 aircraft; it is marked by the vertical line in Fig. 13.Observe that the value of 20 aircraft is same as the design threshold value in Fig. 8, therefore the sector configuration in Fig. 10 can be applied for the 6 p.m. to 10 p.m. CST duration.This example shows that a sector configuration developed with traffic data from a smaller timeinterval can be applied to a larger time-interval without violating the design criteria.Figures 14 and15 show histograms derived from traffic data from the 57 days and the sector configurations II and III shown in Figs.11 and12.Total numbers of traffic-count samples were 163,158 and 653,632, and the peak traffic-counts in a sector were 42 and 31 aircraft for these two sector configurations, respectively.The 99.9 percentile traffic-counts were determined to be 21 and 18 as shown in Figs. 14 and15.Although the 99.9 percentile traffic-count value of 21 for Configuration II was found to be one above the design value, instances of traffic-count of 21 were found to be small with 99.87 percentile value of 20 aircraft.Given that the traffic-counts in most instances are at or below the design value, Configuration II and III can be used for the desired eight-hour and twelve-hour periods.In situations where the traffic-count is forecasted to be much higher than what the sector was designed for, traffic flow management techniques can be used to moderate the demand.The validation results given here suggest that this would be required infrequently.Validation results for Cleveland, Los Angeles and Fort Worth centers are summarized in Table 3.The last row of
+VI. Sector Configuration Change FrequencyResults presented in the previous section indicate that a single sector configuration used during the busy part of the day can be used for the entire day without exceeding the traffic-count limits.These configurations have the most number of sectors compared to other configurations designed for lower traffic-volume.For example, Configuration III shown in Fig. 12 has 16 sectors compared to Configuration II shown in Fig. 11 that has six sectors.Given that each sector requires resources in terms of equipment and air traffic controllers, it is desirable to have as few sectors as possible for handling the expected traffic.Thus, from a resource utilization perspective, sector configurations should be changed as frequently as possible.Although sector configuration change is permitted in the current air traffic control environment, it is difficult to do so frequently because of safety issues of transitioning from one configuration to the next.Change during a busy period is workload intensive because aircraft have to be handed over to neighboring sectors. 11If done in an uncoordinated manner, aircraft would be within the geometric confines of one sector while being controlled by another sector.Configuration change is difficult even if it is timed with a shift change when a new controller assumes separation responsibility for the sector.Regulations require the sector controller to ensure that the incoming controller has complete situational awareness prior to transfer.This is difficult to achieve if the sector configuration changes upon transfer.Due to these practical impediments, sector configuration change should be considered only when there is a significant benefit.The number of sector-hours has been proposed as a benefit metric for comparing different sector configurations in Refs. 10 and 11.It is obtained by summing the product of the number of sectors in each time-interval with the time-interval duration in hours.Following this definition, 256 sector-hours are obtained for the sector configuration change strategy in Fig. 9 with 12 sector configurations.If Configuration III (16 sectors) were used in the Fort Worth Center for the entire day, 384 sector-hours would be spent.The ratio of the sector-hours between a single sector configuration and 12 sector configurations changed once every two-hours is therefore 1.5; sector-hours can be reduced by 50%.Several different configuration change schedules for the Fort Worth Center are provided in Table 4.The numbers of sectors for the two-hour periods are shown in the table.The first row indicates that the same configuration is used throughout the day.The last row of the table contains the same information as the bar chart in Fig. 9; it shows that sector configurations are changed 11 times: 16 to 12, 12 to 10, 10 to 6, 6 to 2, 2 to 4, 4 to 6, 6 to 14, 14 to 16, 16 to 12, 12 to 14, and 14 to 16.Similar schedules were also created for Cleveland and Los Angeles Centers, and sector-hours were computed for each schedule.384 sector-hours were obtained in the Cleveland Center with 16 sectors used for the entire day; 324 sector-hours were obtained with 16 sectors from 5:00 a.m. to 11:00 p.m. and 6 sectors from 11:00 p.m. to 5:00 a.m.EST.For three configuration changes with 14 sectors during 7:00 p.m. to 11:00 p.m., 6 sectors during 11:00 p.m. to 5:00 a.m., and 16 sectors during 5:00 a.m. to 7:00 p.m. EST, 316 sector-hours were obtained.These sector-hours are lower than 435 sector-hours for the current high-altitude operations in the Cleveland Center reported in Ref. 10. On an average 22 sectors are used for daytime (6:00 a.m. to 11:00 p.m. EST) operations and 11 sectors are used for nighttime operations (11:00 p.m. to 6:00 a.m.EST) in the Cleveland Center.Lower sector-hours were obtained in The results summarized in Fig. 16 show that two configuration changes are needed for reducing the sectorhours from about 50% to 19% in Fort Worth Center, 23% in Cleveland Center and 26% in Los Angeles Center above the minimum sector-hours achievable with the 12 two-hour sectorizations.These results suggest that the current practice in most centers of using one configuration for the daytime hours and one for the nighttime hours is a reasonable one.Sector-hours are further reduced to 20% in Cleveland and Los Angeles Centers and 13% in Fort Worth Center with three configuration changes.If four configuration changes are allowed, the sector-hours are at most 15% above that achieved with the two-hour sectorizations.In summary, results presented in Table 3 and in Fig. 16 advocate both, from safety (99.9 percentile peak traffic-count) and resource utilization (sector-hours) perspectives, that two to three sectors configurations are adequate for a good-weather day.Significant reduction in sector-hours is obtained by using Configuration III during daytime hours and Configuration II during nighttime hours in the three centers.Further reduction is obtained if Configuration I is used during the times listed in Table 2.Although sector-hours can be reduced even more by changing sector configurations according to Fig. 16, the frequency of change should be guided by practical considerations, especially during busy traffic periods.
+VII. ConclusionsA robust sectorization and validation method for partitioning airspace into sectors based on several days of air traffic data was described in the paper.Traffic data from ten days out of a set of 57 high-volume low-delay days in 2007 were used for designing sectors in the Cleveland, Fort Worth and Los Angeles center airspace for each twohour period of the day using the method.Of the twelve sector configurations for each day, three were chosen to span the 24-hour time period.Traffic data from the entire dataset were played back though the three selected sector configurations, and histograms of traffic-counts were computed.These distributions show that the probability of traffic-counts exceeding the threshold value used in the sector design is less than one percent.Examples demonstrate that sector configurations created using two-hour time-interval traffic data from several days can be applied over much longer time-intervals from six-hour to 12-hour durations without violating the design criteria.Sector-hours were computed for several sector configuration change schedules to establish a tradeoff with respect to the number of configuration changes during the day.It was determined that most of the benefit as measured by sector-hours is derived by using two configurations, one during daytime hours and one during the nighttime hours.Further benefit is obtained by using one additional configuration.
+VIII. ReferencesFigure 1 .1Figure 1.High-volume low-delay days.
+Figure 2 .2Figure 2. Upper and lower bounds.
+Figure 3 .3Figure 3. Tessellation of the airspace.Figure 4. Directed Acyclic Graph resulting from optimization.
+Figure 4 .4Figure 3. Tessellation of the airspace.Figure 4. Directed Acyclic Graph resulting from optimization.
+Step 507 transfers control to Step 509 once the nine histograms are obtained with airspace partitioned into two through 18 sectors.The sector selection step 509 is used to select a sector configuration with the appropriate number of sectors for the chosen T .The sector cumulative frequency is computed for each of the nine histograms by summing the frequency along the traffic-count bins.The cumulative frequency graph for the histogram in Fig.6is shown in Fig.
+Figure 5 .5Figure 5. Robust sectorization and validation method.
+Figure 6 .6Figure 6.Histogram of traffic-counts during the two-hour period with Fort Worth Center airspace partitioned into two sectors.
+Figure 7 .7Figure 7. Cumulative frequency of traffic-counts during the two-hour period with Fort Worth Center airspace partitioned into two sectors.
+Figure 8 .8Figure 8. 99.9 percentile traffic-counts during the two-hour period with nine different configurations of Fort Worth Center airspace.
+Figure 9 .9Figure 9. Selected sector configurations of the Fort Worth Center airspace for the twohour time-intervals.
+Figure 10 .10Figure 10.Fort Worth Sector Configuration I based on 6 p.m. to 8 p.m. CST traffic data from training set days.
+Figure 11 .11Figure 11.Fort Worth Sector Configuration II based on 4 a.m. to 6 a.m.CST traffic data from training set days.
+Figure 15 .15Figure 15.Histogram of traffic-counts from 6 a.m. to 6 p.m. CST with Fort Worth Center Configuration III.
+Figure 12 .12Figure 12.Fort Worth Sector Configuration III based on 10 a.m. to 12 p.m. CST traffic data from training set days.
+Figure 13 .13Figure 13.Histogram of traffic-counts from 6 p.m. to 10 p.m. CST with Fort Worth Center Configuration I.
+Table 1 . Numbers of training and test days corresponding to days of the week. Day of week Training days Test days1Monday25Tuesday315Wednesday114Thursday19Friday34Total10
+Table 2 .2Selected sector configurations for Cleveland, Fort Worth and Los Angeles centers.CenterFigure 14.Histogram of traffic-counts from 10 p.m. to 6 a.m.CST with Fort Worth Center Configuration II.
+Table 3 .399.9 percentile traffic-counts in the chosen sector configurations for Cleveland, Fort Worth and Los Angeles centers.table lists 99.9 percentile traffic-counts obtained with Sector Configuration III used for the entire day.Cleveland Center and Los Angeles Center results, like the Fort Worth Center results, suggest that sector configurations developed with traffic data from several days over smaller time-intervals can be used over larger time-intervals on similar days of traffic without peak traffic-counts significantly exceeding the design threshold.ConfigurationCenterCleveland Fort Worth Los AngelesI182017II222113III181818III all day1817
+Table 4 .4Sector configuration change schedules for Fort Worth Center.Number of ChangesChange ScheduleSector-hours02
+		
+		
+			
+
+				
+
+
+	
+		Initial Concepts for Dynamic Airspace Configuration
+		
+			ParimalKopardekar
+		
+		
+			KarlBilimoria
+		
+		
+			BanavarSridhar
+		
+		10.2514/6.2007-7763
+	
+	
+		7th AIAA ATIO Conf, 2nd CEIAT Int'l Conf on Innov and Integr in Aero Sciences,17th LTA Systems Tech Conf; followed by 2nd TEOS Forum
+		Belfast, Northern Ireland
+		
+			American Institute of Aeronautics and Astronautics
+			September 18-20, 2007
+		
+	
+	Kopardekar, P., Bilimoria, K., and Sridhar, B., "Initial Concepts for Dynamic Airspace Configuration," Proc. 7th AIAA Aviation Technology, Integration and Operations Conference (ATIO), Belfast, Northern Ireland, September 18-20, 2007.
+
+
+
+	
+		Temporal and Spatial Distribution of Airspace Complexity for Air Traffic Controller Workload-Based Sectorization
+		
+			ArashYousefi
+		
+		
+			GeorgeDonohue
+		
+		10.2514/6.2004-6455
+	
+	
+		AIAA 4th Aviation Technology, Integration and Operations (ATIO) Forum
+		Forum, Chicago, IL
+		
+			American Institute of Aeronautics and Astronautics
+			September 20-22, 2004
+		
+	
+	Yousefi, A., and Donohue, G. L., "Temporal and Spatial Distribution of Airspace Complexity for Air Traffic Controller Workload-Based Sectorization," AIAA 2004-6455, Proc. 4 th AIAA Aviation Technology, Integration and Operations (ATIO) Forum, Chicago, IL, September 20-22, 2004.
+
+
+
+	
+		An Efficient Method for Airspace Analysis and Partitioning Based on Equalized Traffic Mass
+		
+			AKlein
+		
+	
+	
+		Proc. 6 th FAA and Eurocontrol ATM Conference
+		6 th FAA and Eurocontrol ATM ConferenceBaltimore, MD
+		
+			June 2005
+		
+	
+	AIAA 2004-6455
+	Klein, A., "An Efficient Method for Airspace Analysis and Partitioning Based on Equalized Traffic Mass," AIAA 2004- 6455, Proc. 6 th FAA and Eurocontrol ATM Conference, Baltimore, MD, June 2005.
+
+
+
+	
+		Genetic algorithms for partitioning air space
+		
+			DDelahaye
+		
+		
+			J-MAlliot
+		
+		
+			MSchoenauer
+		
+		
+			J-LFarges
+		
+		10.1109/caia.1994.323662
+	
+	
+		Proceedings of the Tenth Conference on Artificial Intelligence for Applications
+		the Tenth Conference on Artificial Intelligence for ApplicationsSan Antonio, TX
+		
+			IEEE
+			March 1-4, 1994
+			
+		
+	
+	Delahaye, D., Alliot, J-M., Schoenauer, M., and Farges J-L., "Genetic Algorithms for Partitioning Air Space," Proc. 10 th IEEE Conference on Artificial Intelligence for Applications, San Antonio, TX, pp. 291-297, March 1-4, 1994.
+
+
+
+	
+		Dynamic Airspace Configuration Management Based on Computational Geometry Techniques
+		
+			JoeMitchell
+		
+		
+			GirishkumarSabhnani
+		
+		
+			RobertHoffman
+		
+		
+			JimmyKrozel
+		
+		
+			ArashYousefi
+		
+		10.2514/6.2008-7225
+	
+	
+		AIAA Guidance, Navigation and Control Conference and Exhibit
+		Honolulu, Hawaii
+		
+			American Institute of Aeronautics and Astronautics
+			August 18-21, 2008
+		
+	
+	Mitchell, J. S. B., Sabhnani, G., Krozel, J., Hoffman, R., and Yousefi, A., "Dynamic Airspace Configuration Management Based on Computational Geometry Techniques," Proc. AIAA Guidance, Navigation, and Control Conference and Exhibit, Honolulu, Hawaii, August 18-21, 2008.
+
+
+
+	
+		Airspace sectorization with constraints
+		
+			HuyTrandac
+		
+		
+			PhilippeBaptiste
+		
+		
+			VuDuong
+		
+		10.1051/ro:2005005
+	
+	
+		RAIRO - Operations Research
+		RAIRO-Oper. Res.
+		0399-0559
+		1290-3868
+		
+			39
+			2
+			
+			June 23-27, 2003
+			EDP Sciences
+			Budapest, Hungary
+		
+	
+	Trandac, H., Baptiste, P., and Duong, V., "Optimized Sectorization of Airspace with Constraints," Proc. 5 th Eurocontrol/FAA ATM R&D Seminar, Budapest, Hungary, June 23-27, 2003.
+
+
+
+	
+		Airspace Sector Redesign Based on Voronoi Diagrams
+		
+			MinXue
+		
+		10.2514/6.2008-7223
+	
+	
+		AIAA Guidance, Navigation and Control Conference and Exhibit
+		Honolulu, Hawaii
+		
+			American Institute of Aeronautics and Astronautics
+			August 18-21, 2008
+		
+	
+	Xue, M., "Airspace Sector Redesign Based on Voronoi Diagrams," Proc. AIAA Guidance, Navigation, and Control Conference and Exhibit, Honolulu, Hawaii, August 18-21, 2008.
+
+
+
+	
+		A Weighted-Graph Approach for Dynamic Airspace Configuration
+		
+			StephaneMartinez
+		
+		
+			GanoChatterji
+		
+		
+			DengfengSun
+		
+		
+			AlexandreBayen
+		
+		10.2514/6.2007-6448
+		AIAA 2007-6448
+	
+	
+		AIAA Guidance, Navigation and Control Conference and Exhibit
+		Hilton Head, SC
+		
+			American Institute of Aeronautics and Astronautics
+			August 20-23, 2007
+		
+	
+	Martinez, S. A., Chatterji, G. B., Sun, D., and Bayen, A. M., "A Weighted-Graph Approach for Dynamic Airspace Configuration," Proc. AIAA Guidance, Navigation, and Control Conference and Exhibit, Hilton Head, SC, AIAA 2007-6448, August 20-23, 2007.
+
+
+
+	
+		Analysis of an optimal sector design method
+		
+			MichaelDrew
+		
+		10.1109/dasc.2008.4702801
+	
+	
+		2008 IEEE/AIAA 27th Digital Avionics Systems Conference
+		St. Paul, MN
+		
+			IEEE
+			October 2008
+		
+	
+	Drew, M., "Analysis of an Optimal Sector Design Method," Proc. 27 th Digital Avionics Systems Conference, St. Paul, MN, October 2008.
+
+
+
+	
+		Algorithms for Combining Airspace Sectors
+		
+			MichaelBloem
+		
+		
+			ParimalKopardekar
+		
+		
+			PramodGupta
+		
+		10.2514/atcq.17.3.245
+	
+	
+		Air Traffic Control Quarterly
+		Air Traffic Control Quarterly
+		1064-3818
+		2472-5757
+		
+			17
+			3
+			
+			2009
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	Bloem, M., Gupta, P., and Kopardekar, P., "Algorithms for Combining Airspace Sectors," Air Traffic Control Quarterly, Vol. 17, No. 3, 2009, pp. 245-268.
+
+
+
+	
+		Flight Level-based Dynamic Airspace Configuration
+		
+			KennethLeiden
+		
+		
+			StevenPeters
+		
+		
+			StaceyQuesada
+		
+		10.2514/6.2009-7104
+		AIAA 2009-7104
+	
+	
+		9th AIAA Aviation Technology, Integration, and Operations Conference (ATIO)
+		Hilton Head, SC
+		
+			American Institute of Aeronautics and Astronautics
+			September 21-23, 2009
+		
+	
+	Leiden, K., Peters, S., and Quesada, S., "Flight Level-based Dynamic Airspace Configuration," Proc.9 th AIAA Aviation Technology, Integration, and Operations Conference (ATIO), Hilton Head, SC, AIAA 2009-7104, September 21-23, 2009.
+
+
+
+	
+		Characterization of Days Based on Analysis of National Airspace System Performance Metrics
+		
+			GanoChatterji
+		
+		
+			BassamMusaffar
+		
+		10.2514/6.2007-6449
+		AIAA 2007-6449
+	
+	
+		AIAA Guidance, Navigation and Control Conference and Exhibit
+		Hilton Head, SC
+		
+			American Institute of Aeronautics and Astronautics
+			August 20-23, 2007
+		
+	
+	Chatterji, G. B., and Musaffar, B., "Characterization of Days Based on Analysis of National Airspace System Performance Metrics," Proc. AIAA Guidance, Navigation, and Control Conference and Exhibit, Hilton Head, SC, AIAA 2007-6449, August 20-23, 2007.
+
+
+				
+			
+		
+	

file124.txt

@@ -0,0 +1,316 @@
+
+	
+	
+		
+I. Introductionhis paper is motivated by the need for selecting days with particular air traffic flow patterns and operational characteristics, as encapsulated in the performance metrics, for validating simulation models and evaluating next generation air traffic system concepts.Evaluation of system-wide impacts in terms of cost and benefits with one or two days of data, or with several days of data with similar traffic conditions, is of limited utility.Such evaluations therefore, have to be conducted with a set containing days with distinct characteristics.In order to balance the effort required against the quality of results achieved for these types of simulations and evaluations, a small set of days that covers all the possible traffic conditions is desirable.The multiple-metric classification method proposed in this paper makes it possible to create such a set of days.
+TPrior effort on the selection of days for validating simulation models is described in Refs. 1 and 2. Reference 1 contains a detailed description of the data and the procedure used, while Ref. 2 is a summary of the same.The approach consisted of using the K-Means algorithm, first proposed in Ref. 3, to partition the set of days into six significant groups, each with at least 2% of the days, and one outlier group for days that could not be assigned to the six significant groups.Each group was separated from others in terms of a single Euclidean distance metric composed of the eight chosen metrics.Based on analysis of the metrics associated with each of the six significant groups, they concluded that Ground Delay Program (GDP) minutes and total operations count, a measure of trafficvolume, were the primary determinants of group membership.Threshold values were computed for these two metrics and used within a decision-tree for labeling a given day as a typical day characterized by one of the six groups.The main limitation of the method is that the Euclidean distance metric, constructed by adding quadratic terms corresponding to metrics with different scales and units, partitions days in the transformed domain of the combined metrics.This obscures the relation of the groups to the individual metrics.Thus, grouping with a finer level of granularity cannot be achieved with this method.The method proposed in this paper overcomes the limitations of the previous approach by creating groups based on each metric individually using the K-Means algorithm.Each day is then tagged with a composite ID consisting of IDs of the groups it belongs to based on different metrics.For example, if a day is a member of Group 1 based on Metric 1, a member of Group 1 based on Metric 2, and a member of Group 3 based on Metric 3, it is tagged with the composite group ID of (1,1,3), where the first index indicates grouping based on Metric 1, the second based on Metric 2 and the third based on Metric 3.All days with the same tag are then placed in one group.A salient feature of the proposed algorithm is that the linguistic description of the group labels based on each metric is retained in the composite label of the final grouping.For example, if groups 1, 2 and 3 mean "low," "medium" and "high," respectively, the composite label (1,1,3) means "low" based on the first metric, "low" based on the second metric and "high" based on the third metric.Thus, the fidelity of partitions of individual metrics is retained in the final grouping.The rest of the paper is organized as follows.Major trends observed in the 517 days of NAS delay data are described in Section II.Total time delay in minutes is used as a distance metric within the K-Means algorithm to partition the set of 517 days into ten groups in Section III.Convergence characteristics of the K-Means algorithm and summary statistics of the ten groups are provided in this section.A multiple-metric classification technique that builds on the single-metric classification technique is then developed in Section IV.Two examples of grouping of days are provided in Section IV to highlight the salient features of the algorithm.Conclusions are discussed in Section V.
+II. National Airspace System Delay and Traffic-Volume CharacteristicsTo keep track of operational efficiency of the air traffic system, the Federal Aviation Administration (FAA) and the Bureau of Transportation Statistics (BTS) keep records of a multitude of metrics including delay, number of operations, conditions at airports, and traffic management initiatives in databases.Several of the frequently used databases are: Aviation System Performance Metrics (ASPM), Air Traffic Control System Command Center (ATCSCC) Logs, BTS data, Enhanced Traffic Management System (ETMS) and OPSNET.Detailed descriptions of the contents of these databases are available in Refs. 1 and4.All the analysis and the results in this paper are based on OPSNET data, which are available via http://www.apo.data.faa.gov.OPSNET data only include delays of fifteen minutes or more experienced by Instrument Flight Rule (IFR) flights that are reported by the FAA facilities.These data do not include delays caused by mechanical or other aircraft operator problems.Speed reductions and pilot initiated deviations around weather are also not reported.Taxi times spent under non-FAA facilities, for example under company/airport ramp towers, are not included in delay reports. 5ASPM also provides delay data that are computed based on the Out-Off-On-In (OOOI) data provided by nine commercial and cargo carriers, which can also be utilized for analyzing days via the methods discussed in this paper.Although the trends in ASPM and OPSNET data are similar, the two databases contain very different types of data that make comparisons between them difficult.They are both useful depending on the analysis desired.OPSNET delay data are provided in a tabular form; numbers of aircraft delayed are reported by category, by class and by cause.Delays by category consist of numbers of aircraft that were subject to departure delays, arrival delays, enroute delays and traffic management system (TMS) delays.The distinction between the enroute and TMS delays is discussed later in this section.Delays by class consist of numbers of air carrier, air taxi, general aviation and military aircraft that were delayed.Delays by cause consist of numbers of aircraft that were delayed due to weather, terminal volume, center volume, equipment limitations, runway issues and "other" issues.International delays are included in the "other" category.In addition to these, average time delay in minutes and total time delay in minutes are included in the table.Seventeen variables of OPSNET national delay data for two days are summarized in Table 1.This table shows that the numbers of aircraft delayed by category (departure + arrival + enroute + TMS) add up to the total number of aircraft delayed.Similarly, the numbers of aircraft delayed by class and by cause also add up to the total number of aircraft delayed during the day of operation.Observe that the average delay is obtained by dividing the total time delay in minutes by the total number of delayed aircraft.There are three significant trends that are easily seen in Table 1.First, the sum of departure and TMS delayed flights account for most of the delayed flights.Second, most aircraft are delayed due to weather.Third, as expected, total time delays are proportional to total numbers of aircraft delayed.To understand NAS delay characteristics, OPSNET delay data covering a period of 517 days (17 months) spanning the period from January 1, 2003 through May 31, 2004 were analyzed.Figure 1 shows a scatter plot of the percentages of aircraft delayed due to weather as a function of days.The mean percentage of aircraft delayed due to weather was found to be 66% and the standard deviation was found to be 21% for this dataset.Additional statistical characteristics are summarized in Table 2.These results show that the number of aircraft delayed due to weather represents a large fraction of the number of aircraft that experience delay in the NAS, a finding consistent with Ref. 6, which states that weather is responsible for approximately 70% of NAS delays.The data shown earlier in Fig. 1 was reorganized as a function of total number of aircraft that experienced delay.These transformed data are shown in Fig. 2. Figure 2 shows that on days when a large number of aircraft are delayed, weather is the dominant cause of delays.Percentages of aircraft delayed due to weather are widely scattered when fewer aircraft are delayed, which indicates that factors other than weather are also responsible for delays on those days.Figure 3 shows the number of aircraft delayed due to weather versus the total number of aircraft delayed.Viewing the sample points with respect to the diagonal line across the figure, it is clear that a high degree of correlation exists between the number of aircraft delayed due to weather and the total number of aircraft delayed in the NAS.Assuming both the number of aircraft delayed due to weather and the total number of aircraft delayed are random variables, the correlation coefficient was computationally determined to be 0.95.Correlation between the number of aircraft delayed due to weather and the total time delay in minutes due to all reportable causes (see the last row of Table 1 for an example) was found to be 0.94.The causes of delay other than weather were also studied.Their statistics are summarized in Table 3 along with those of weather delays.Correlation coefficients 1 ρ in Table 3 are defined with respect to the total number of aircraft delayed, and correlation coefficients 2 ρ are defined with respect to the total accrued time delay in minutes.As evident from the correlation coefficient value of 0.21 in this table, the number of aircraft delayed due to volume has a weak linear correlation with the total number of aircraft delayed in the NAS.Correlation is even lower, 0.11, with respect to the total time delay in minutes.Similarly, the value of the correlation coefficient between the number of aircraft delayed due to equipment, runway and other non-US facilities, and the total number  To determine relative contributions of delays attributed to departure, enroute and arrival phases of flights, and to TMS restrictions, percentages of aircraft delays by category were calculated for the 517-day dataset.It was determined that on an average, on any given day, 47% of the aircraft that are delayed in the NAS, are delayed during departure, 1% during enroute and 14% during arrival phases of flight.The average percentage of aircraft delayed due to TMS was 39%.Analysis of the data showed that, on average, departure delayed flights and TMS delayed flights roughly account for 86% of the flights that are delayed.Additional statistics that characterize these delays are summarized in Table 4.Note that the values of the correlation coefficients 1 ρ and 2 ρ listed in the table are with respect to the total number of aircraft delayed and the total accrued time delay in minutes, respectively.Since traffic management initiatives such as GDP and GS are applied to aircraft while they are on the ground, and rerouting and holding while they are airborne, TMS delays include both ground and airborne delay components.To separate TMS delay into ground delay and airborne delay components, analysis of GDP and GS data, that are also available via OPSNET, is needed.Like data in Table 1, these data are also provided in a tabular format with the following items: 1) date, 2) number of aircraft delayed, 3) total delay in minutes, and 4) average time delay in minutes.For example, GDP and GS data for two days, 10 April 2004   1.Ground delay and airborne delay components can be computed using NAS delay data (see delays by category in Table 1) and the GDP and GS delays (see Table 5) as follows.Let, ,, and be the numbers of aircraft delayed during departure, due to GDP, and due to GS, respectively.The total number of aircraft delayed on the ground is thend n GDP n GS n GS GDP d G n n n n + + = (1)Number of aircraft delayed during the airborne phase can be obtained after subtracting GDP and GS components from TMS delays as, ) (  6 show that on average 74% of the aircraft that experience delay are delayed on the ground, compared to an average of 26% that are delayed while airborne.The last row of Table 6 shows that on some days NAS conditions are unusual in that a large percentage of delayed aircraft experience airborne    In addition to the NAS delay metrics discussed in this section, past studies such as Ref. 1, 2 and 7 have used metrics of traffic volume to select days for analysis.There is consensus in the literature that the "traffic volume" and the delay taken together characterize NAS operations, therefore traffic volume metrics are discussed next.OPSNET database includes Towers: Summary Report, Instrument Operations: Summary Report and Centers: Summary of Domestic Operations Report that contain traffic volume data.These three reports count traffic from different perspectives.One is unable to separate departure counts from arrival counts in the Towers: Summary Report and in the Instrument Operations: Summary Report.A departure at one facility is counted as a departure at that facility and as an arrival at a different facility.Since departures and arrivals are counted together twice in these reports, the total number of operations excluding the overflight operations have to be halved to estimate the number of departures or the number of arrivals.The Centers: Summary of Domestic Operations Report directly provides a count of the number of departures from airports within the ARTCCs.Since departure count eventually drives the overflight count and the arrival count, it represents the traffic demand.Due to this reason, departure count from the Centers: Summary of Domestic Count Report has been used in this paper.Table 7 lists the departure counts and the overflight counts for the two days.Departure counts excluding military flights for the two days obtained by summing the air carrier, air taxi and general aviation departures are 31,959 and 42,062.GS GDP TMS e a A n n n n n n - - + + = (2)A histogram of the total domestic departure counts for the 517 days of data is shown in Fig. 5.The minimum and the maximum numbers of departures were found to be 25,677 on 11/27/2003 (Thanksgiving holiday) and 51,399 on 5/27/2004 (Thursday).Observe that the histogram is bimodal which indicates that days can be classified into two categories -low departure count day and high departure count day.Reference 1 noted similar observations and offered evidence that the bimodal distribution is primarily due to the weekday versus weekend traffic levels.This section described several delay and traffic volume metrics that are available in OPSNET data.Summary statistics described in the tables and the patterns observed in the figures suggest that these metrics can be used for distinguishing one day of NAS operations from another day of NAS operations.To illustrate the use of a metric for classifying days of operations, total time delay in minutes, in Eq. ( 3), is used as a distance metric within the K-Means method in the next section.T n
+III. Single-Metric ClassificationThe motivation for assigning or labeling days into groups with associated properties, such as mean delay values, is to aid selection of prototype days for analysis.For example, a few days from a group of days with large delays and from a group of days with small delays can be selected for system-wide studies using the National Aeronautics and Space Administration's air traffic simulation, concept evaluation, and decision support tools such as the Airspace Concept Evaluation System (ACES), the Center TRACON Automation System (CTAS) and the Future ATM Concepts Evaluation Tool (FACET). 8-11 All classification processes use metrics, or features, of the data to partition it into groups.A popular classification method, known as the K-Means method, partitions data such that the means associated with the groups are as widely separated as possible. 3The method labels the data elements based on their closeness to the group American Institute of Aeronautics and Astronautics means for reducing the group variance.The K-Means algorithm consists of two steps: 1) the initialization step, and 2) the iterative step.Data elements that are far apart from each other are chosen as the initial means of the groups during the initialization step.Each element is then assigned to the group that it is closest to, based on its distance with respect to the group's initial mean value.Group means are then recomputed based on the elements assigned.Each element is then reassigned to its closest group based on its distance with respect to the recomputed mean values.This process of computation of the means and reassignment of elements to groups is continued in subsequent iterative steps.Convergence is achieved when the numerical values of the group means do not change with reassignment of the elements.Iterations are halted once convergence is achieved.To further clarify the initialization and the iterative steps of the K-Means algorithm, consider a vector with the following elements [0, 0.5, 0.8, 1.2, 5, 7,12,15,20,25].If two groups are desired, the elements with values closer to 0 are assigned to the first group and the elements with values closer to 25 are assigned to the second group.Thus, elements one through seven are assigned to Group One and elements eight through ten are assigned to the Group Two in the initialization step.With this assignment of the elements to the groups, average values of the first group and the second group are 3.79 and 20 and the standard deviation values are 4.47 and 5. Reassignment of the elements based on the recomputed means results in the first six elements being assigned to Group One and the last four elements being assigned to Group Two in the first iterative step.Group means are recomputed in the next iterative step.These means are 2.42 and 18 and the standard deviations are 2.87 and 5.72.The next iterative step results in the same means and the standard deviations as those in the prior step; final grouping is therefore achieved in the previous step.For this example, the K-Means algorithm partitions the data into two groups in three steps.If three groups are desired for the above example, a value from the vector that is far away from both 0 and 25 needs to be selected as the initial value for the third group.Observe that this value is 12 since its minimum distance to 0 (12 units) and 25 (13 units) is a maximum compared to the minimum distances of other elements to 0 and 25.Other values in the vector are less than 12 units with respect to either 0 or 25.Once these initial group means are chosen, the subsequent iterative steps are the same as those described in the previous paragraph.It should be noted that good initial conditions are needed because the K-Means algorithm is sensitive to initial conditions.The K-Means algorithm was used for classifying 517 days into ten groups using total time delay in minutes as the discriminating metric.The choice of ten groups was arbitrary.The algorithm converged in thirteen iterations.Its convergence characteristics are shown in Figure 6.Properties of the ten groups, based on total time delay statistics of the elements assigned to the groups, are summarized in Table 8.The second column of the table shows the number of days in the group, with the group ID given in the first column.Columns three and four show the average delay and the standard deviation of the delays in minutes.Columns five and six show the minimum and maximum delays in minutes of the days belonging to the particular groups.The data in this table  7. Note that the extent of the abscissa is limited to the range of the delay data.Sixteen days (seven days in Group 8, seven days in Group 9 and two days in Group 10) that experienced large delays are listed in Table 9.Of the 517 days grouped by the K-Means algorithm, the least delay of 1,686 minutes occurred on 1/11/2003 (a Saturday) and most delay of 186,313 minutes occurred on 5/13/2004 (a Thursday).
+Table 9. Days with large delays.Results presented in this section demonstrate the use of the K-Means algorithm for partitioning a set of days into groups organized in order of a single metric like total time delays.The next section describes a labeling technique that enables use of the single-metric K-Means classification technique for achieving classification based on multiple metrics.
+IV. Multiple-Metric ClassificationMotivation for multiple-metric classification stems from the desire for finer levels of partitioning.For example, a group with large mean delay contains days when aircraft were delayed due to weather and also days when aircraft were delayed due to runway conditions.In order to discern which ones were affected by weather and which ones were affected by runway conditions, one would need metrics such as numbers of aircraft delayed due to weather and due to runway conditions, in addition to total delays.Fidality Classification based on multiple metrics has been traditionally accomplished by weighing and combining several metrics into a single metric, and then using it in a K-Means algorithm.For example, if day ' ' was characterized by metrics:, , …, and if day 'q f 1 , q u 2 , q u f q u ,r ' was similarly characterized by , , …, , a weighted quadratic function of the form1 , r u 2 , r u f r u , 2 , , 1 , ) ( l r l q f l l r q u u w d - = ∑ ≤ ≤ (4)can be defined as the distance metric between days and q r .Note that through are the weights corresponding to the different metrics.Interpretation of the distance metric, Eq. ( 4), for grouping days with the K-Means algorithm is straightforward with as the mean of measure l in groupm k n j p u w d l j l k f l l j k ≤ ≤ ≤ ≤ - = ∑ ≤ ≤ 1 ; 1 ; ) ( 2 , , 1, where is the number of metrics, m is the number of days and is the number of groups.f nAlthough the distance metric, Eq. ( 5), enables transformation of a multiple-metric classification problem into a single-metric classification problem, its deficiencies are noteworthy.Limitations stem from the fact that metrics have different scales and units, and that only their combined contribution is available in the distance metric; classification is insensitive to individual contribution.For example, consider the two metrics in Table 1: 1) number of aircraft affected by departure delays and 2) total time delay.Units of the two metrics are quite different, number of aircraft and minutes.The scales are also different by an order of magnitude; 257 aircraft impacted by departure delays versus 12,616 minutes of total time delay on 10 April 2004.To compensate for these differences, the associated weights have to be scaled correctly, and their units have to be chosen appropriately to enable summation of quantities with disparate units.References 1 and 2 suggest that the inverse of the statistical variance of the metric should be used to weigh its contribution.Even with this scaling, a meaning cannot be ascribed to the grouping in the native domain of the metrics.In order to overcome the limitations of the weighted quadratic distance function used in the prior approach of Refs. 1 and 2, a multiple metric classification technique that treats each metric independently of others in an dimensional metric space is proposed.Since each metric is treated independently, the single metric K-Means algorithm described earlier in Section III can be used for assigning days to groups based solely on each metric.IDs of these groups are then coordinates in the -dimensional metric space.For the sake of discussion, consider the problem of classifying days into four groups using two metrics.Using the K-Means algorithm twice, days are first assigned to four groups based on Metric One, and then to four groups based on Metric Two.The resulting sixteen possible groups are labeled using two indices as follows: (1, 1), (1, 2), (1, 3), (1, 4), (2, 1), (2, 2), (2, 3), (2, 4), (3, 1), (3, 2), (3, 3), (3,4), (4, 1), (4, 2), (4, 3) and (4,4).The first index denotes group ID based on Metric One and the second index denotes group ID based on Metric Two.Thus, a day which is assigned to the second group based on Metric One and to the first group based on Metric Two is a member of group (2, 1).Since a unique group is labeled using two indices in this two metric classification problem, the combined group IDs are coordinates in a twodimensional metric space.f f Generalization of the technique to metrics such that days are classified into groups using metric results inf l n l ∏ ≤ ≤ = f l l g n n 1 (6)number of possible groups.Equation (6) shows that if the same number of groups is desired for each metric, the number of possible groups is given in terms of the power of .For example if groups are desired with each metric, the number of possible groups is .Thus, it is seen that the growth in the number of groups is explosive with increasing number of metrics.Should one conclude that the growth is unbounded based on this observation, or is there an upper bound on the number of groups?The answer is provided by the following.If each day is classified into its own group, one would have the same number of groups as the number of days; hence, number of days is the upper bound.This fact also implies that if is the number of days and , there are at least number of empty groups, groups without any members.Removing these empty groups, the number of possible groups, , is given asf n f n m m n g > m n g - g n ∏ ≤ ≤ = f l l g n m n 1 ), min( (7) where each .m n l ≤ Is it possible that several of the groups counted in Eq. ( 7) are empty?One can demonstrate this to be true by constructing the following examples.Consider the problem of classifying ten days into two sets of five groups using American Institute of Aeronautics and Astronautics two metrics.Following the nomenclature of Eq. ( 7), 2 = f for the two metrics, 5 1 = n using Metric One, 5 2 = n using Metric Two, and for the ten days.Substituting these numerical values in Eq. ( 7) it is seen that .Assume that the first two days are assigned to Group 1, the third and the fourth to Group 2, the fifth and the sixth to Group 3, the seventh and the eighth to Group 4, and the final two to Group 5 based on Metric One, and also based on Metric Two.In this scenario, groups with members are (1, 1), (2, 2), (3,3), (4,4) and (5, 5).All other groups labeled with two different indices, such as (1, 2), (1, 3) and (5, 1), are empty.7).If the first five days are assigned to Group 1 and the next five to Group 2 based on both the metrics, groups (1, 1) and (2, 2) are non-empty while groups (1, 2) and (2, 1) are empty.These two examples clearly show that it is always possible to have empty groups.An aspect of multiple-metric classification that has not been discussed so far is the semantics associated with the group IDs.Without a linguistic meaning, it is difficult to interpret what do group IDs such as (1, 1) and (1, 2) mean.One of the ways of attributing a meaning to the indices is to order them according to the increasing values of the group means.For example if total time delay in minutes was the metric being considered, the index with the least value would correspond to the group with the minimum mean total time delay while the index with the highest value would correspond to the group with the maximum mean total time delay.From an implementation perspective, once classification into specified number of groups is accomplished with the K-Means algorithm using a single metric, and group means are computed based on the metric values of the assigned members, the group means are sorted in an increasing order.Indices of the groups are then re-labeled to reflect the sorted order.This procedure is repeated for each metric to obtain the complete set of indices required for labeling the groups.Three metrics-1) total domestic departure counts, 2) number of aircraft delayed on the ground, and 3) number of aircraft delayed in the air were used as the basis for classification in the multiple-metric K-Means algorithm described in this paper.Recollect that the total domestic departure counts were obtained from the Centers: Summary of Domestic Counts Report discussed in Section II.Numbers of aircraft delayed on the ground and delayed in the air were computed using Eqs.( 1) and (2).Days were initially organized into three groups using the single-metric K-Means algorithm.When number of aircraft delayed in the air was used as a metric, 393 days were assigned to Group 1, 128 to Group 2 and one to Group 3. The mean and the standard deviation values derived from the number of aircraft delayed in the airborne phase metric of the assigned members for these groups are listed in Table 10.The sole member of Group 3, 7/15/2005, had excessive amount of airborne delay of 1891 minutes.1661 aircraft were delayed on the ground on this day.Since this day is an outlier, it has the effect of increasing the standard deviation of the other groups.Due to this reason, it was removed from the dataset.American Institute of Aeronautics and Astronautics The analysis was repeated with the remaining 521 days.The resulting grouping showed that 6/5/2005 became the sole member of Group 3 with 970 aircraft delayed in the airborne phase and 753 aircraft delayed on the ground.Table 11 summarizes these results.Comparing Table 10 to Table 11, it is seen that the removal of 7/15/2005 data lowers the standard deviation values of the groups.Since 6/5/2005 is an outlier day, it was also removed from the dataset.Classification based on number of aircraft delayed in the airborne phase for the remaining 520 days are summarized in Table 12.Observe that the standard deviation values decrease further for groups 1 and 2. It increases for Group 3. The mean values decrease for all three groups.Days belonging to Group 1 can be thought of as days with low number of aircraft delayed while airborne; in Group 2 as days with medium number of aircraft delayed and the ones in Group 3 as days with large number of aircraft delayed.Similar categorization based on number of aircraft delayed on the ground partitions the days in Groups 1 through 3, whose statistics are summarized in Table 13.Results obtained using total domestic departure counts are provided in Table 14.Note that the days were classified into two groups based on the bimodal distribution seen in Fig. 5.Comparing Tables 12 and13, it is seen that the trends are similar with the largest number of days assigned to groups with lower mean and lower standard deviation values.The trends are different in Table 14.More days are assigned to Group 2 with higher mean departure counts.Given that three groups were created using two metrics and two groups using one metric, the total number of possible composite groups, determined using Eq. ( 7), is 18.The range of IDs for these groups is (1, 1, 1) to (2,3,3).With the first index being the group number associated with the total domestic departure counts metric, the second index being the group number associated with the number of aircraft delayed on the ground metric, and the third index being the group number associated with the number of aircraft delayed during the airborne phase metric, each day in the set of 520 days has a three index group ID associated with it.Many of the salient features of multiple-metric classification algorithm are apparent from data in Table 15.A majority of days, 94 and 89, are assigned to Group (1, 1, 1) and Group (2, 1, 1).These groups represent days on which few aircraft were delayed.The difference between them is that Group (1, 1, 1) represents low-volume days while Group (2, 1, 1) represents high-volume days.Table 15 shows that a large number of days were assigned to groups (2, 2, 1) and (2, 2, 2) that represent high-volume days on which many aircraft were delayed on the ground.There are several groups with few days assigned to them; five groups had three or less than three days assigned to them.Table 16 lists the corresponding dates.Member days belonging to small groups can be considered to be special.Days in Group 3 with Group ID (1, 1, 3) experienced relatively low total departure counts, fewer aircraft affected by ground delay, and higher number of aircraft affected by airborne delay.Two members of Group 6 had more aircraft delayed on the ground compared to members of Group 3. Three member days of Group 7 had many more aircraft delayed on the ground and few during the airborne phase.The sole member of Group 9 had many aircraft delayed while on ground and while in the air.The member days of Group 12 experienced high departure counts, relatively few aircraft delayed on the ground, and a large number of aircraft delayed during the airborne phase.Another example of multiple-metric classification using total domestic departure counts and delays by cause: 1) weather, 2) volume, 3) equipment, runway and other, as metrics for partitioning 522 days into groups is summarized in Table 17.Numbers of aircraft delayed due to terminal volume and due to center volume (see "Delays by Cause" in Table 1) were combined to obtain the number of aircraft delayed due to volume.Similarly, numbers of aircraft delayed due equipment, runway and other issues were added together for obtaining the number of aircraft impacted due to these causes.As in the previous example, days were categorized into groups with the K-Means algorithm using each of these four metrics one at a time.Observe that in this example, days are partitioned into 37 groups out of 54 possible groups.
+American Institute of Aeronautics and AstronauticsAmerican Institute of Aeronautics and Astronautics The results of the two examples considered here show that 1) days can be classified into the specified number of groups based on each individual metric, 2) the individual metric group labels can be used for creating multiplemetric group labels, and 3) linguistic description of the individual metric grouping is retained in the composite group label.These examples also illustrate that the multiple-metric classification method does not require that the number of groups be the same based on each metric for creating composite group IDs.In the first example, days were organized in two groups using the total number of departure counts metric and in three groups using the number of aircraft delayed on the ground and the number of aircraft delayed in the air.This technique of maintaining different numbers of groups along different axes of the metric space is in contrast with the method described in Ref. 1 and 2 that only partitions data along the single distance metric.Results demonstrate that the multiple-metric classification method generates groups with several members and also groups with few members; thus, identifying both nominal and off-nominal days.By selecting a typical day from each group, and then using traffic data corresponding to those days, enough data diversity can be assured for validation of simulations and for Monte Carlo type of benefits analysis of novel air traffic management concepts.Resulting benefits metrics can be weighed with number of members in the group that each day is associated with for estimating overall benefits.
+V. ConclusionsConsistent with other studies, analysis of 517 days of National Airspace System (NAS) delay data, which were obtained from the Federal Aviation Administration's Air Traffic Operations Network (OPSNET) database, showed that weather is the predominant causal factor for delays; equipment and runway conditions, and traffic-volume are the other major causal factors.It was also determined the departure and traffic management system delays account for about 86% of the aircraft that are delayed.Ground Delay Program and Ground Stop delay data, also obtained from OPSNET, were combined with the NAS delay data to obtain the number of aircraft delayed on the ground and in the air.The results obtained indicate that on an average 74% of the delayed aircraft are delayed on the ground while only 24% are delayed in the air.
+American Institute of Aeronautics and AstronauticsThe daily total time delay in minutes was used as a discriminating metric within the K-Means algorithm for partitioning 517 days into ten groups.Mean time delay values associated with the resulting groups, computed using time delay values of the member days, arranged in increasing order were found to be approximately equidistant from the preceding and succeeding mean values.Differences between the standard deviation values associated the groups were also found to be small.Most of the days were assigned to groups with smaller mean time delays.Days with large delays were also identified by the algorithm.A multiple-metric algorithm was synthesized with the single-metric K-Means algorithm at its core.The technique consists of creating groups using each metric individually as a distance metric within the single-metric K-Means algorithm.Member days are labeled with the group numbers associated with the metrics.Final grouping is achieved by assigning days with a common label to the same group, such that groups are labeled by the same number of indices as the number of metrics.The multiple-metric algorithm was applied to the problem of organizing the 522 days into groups using a) total domestic departure counts, b) number of aircraft delayed on the ground and c) number of aircraft delayed in the air as the three metrics.Two days that were found to be outliers were removed and the remaining 520 days were classified into 18 groups.Six of the 18 groups had six or fewer days as members.Although these groups represent unusual days, their inclusion in a set of days that represents diverse air traffic conditions is essential for evaluating concepts and validating simulations.The other 12 groups had 14 or more days as members.Another example of multiple-metric classification of 522 days into groups with a) total domestic departure counts, b) weather delays, c) volume delays and d) equipment, runway and other delays as the chosen metrics was presented.In this instance, 37 groups out of 54 possible groups had member days.Of the 37 groups, 20 groups had five or fewer days as members.Comparing the results obtained via the two examples, it was seen that different sets of days can be created and certain unusual days can be identified based on the choice of metrics.The two examples serve as illustrations of the ability of the multiple-metric algorithm to create datasets, consisting of days classified into groups, with enough data diversity for concept evaluation and simulation validation.Figure 2 .2Figure 2. Percentage of aircraft delayed due to weather as a function of total number of aircraft delayed.
+Figure 3 .3Figure 3. Proportion of number of aircraft delayed due to weather to the total number of aircraft delayed in the NAS.
+the numbers of aircraft delayed in arrival phase, in enroute phase, and due to TMS, number of aircraft delayed in the NAS.Results obtained using Eqs.(1) through (3) with 517 days of OPSNET data are shown in Fig. 4; n and values are plotted against values.This figure shows that when more aircraft are delayed, a significantly higher number of them are delayed on the ground compared to in the air.Statistical trends summarized in Table
+Figure 4 .4Figure 4. Proportion of number of aircraft delayed on the ground and in the air to the total number of aircraft delayed in the NAS.
+delay.Of the 665 aircraft that were delayed on 2/22/2003, 513 aircraft (77%) were delayed during their airborne phase of flight.The airborne and ground delay values for this day are marked with a large 'X' and a large 'O' in Fig.4.
+Figure 5 .5Figure 5. Histogram of 517 days of total domestic departure counts.
+Figure 6 .6Figure 6.Convergence characteristics of the K-Means algorithm as it partitions the 517 days into ten groups.
+can be constructed to show that empty groups are possible even when .
+Table 1 . OPSNET NAS delay summary data.1Data Variable4/10/20044/13/2004Delays by CategoryTotal # of Aircraft3912,312Departure257651Arrival101391Enroute012TMS331,258Delays by ClassAir Carrier3381,769Air Taxi26474General Aviation2769Military00Delays by CauseWeather2352,049Terminal Volume5927Center Volume4113Equipment126Runway3024Other25173Time DelayAverage Time (min.)32.2753.51Total Time (min.)12,616123,709American Institute of Aeronautics and Astronautics
+Table 2 . Statistical characteristics of percentages of aircraft delayed due to weather.2CharacteristicAircraft delayed by weatherMean66%Standard deviation21%Minimum5%Maximum98%Median70%
+Figure 1. Percentage of aircraft delayed due to weather.American Institute of Aeronautics and Astronautics of aircraft delayed in the NAS was found to be 0.27.It was found to be 0.16 with respect to the total time delay in minutes.In the hierarchy of prime causal factors for delays, equipment and runway conditions follow weather.Results presented in this section suggest that delay metrics that encapsulate weather characteristics are likely to be useful in the classification of days.Delays attributed to weather, volume, andequipment, runway and other causes are realized viadeparture, arrival, enroute and TMS restrictions.Departure delays incur by holding aircraft at thegate, on the taxiway, short of the runway, and on therunway. Arrival delays accrue when aircraft aredelayed in the arrival Center's airspace or inTerminal Radar Approach Control airspace due torestrictions at arrival airports. Enroute delays occurwhen aircraft are held as a result of initiativesimposed by a facility for traffic managementreasons such as volume regulation, frequencyoutage and weather. The other major category ofdelays in the OPSNET data is TMS delays, whichresult from national or local Center (coordinatedwith Air Traffic Control System Command Center)traffic flow management initiatives such as GroundDelay Programs, local Ground Stops (GS),Departure Sequencing Programs, Enroute SpacingPrograms, Arrival Sequencing Programs, airborneholding, rerouting, Miles-in-Trail, Minutes-in-Trailand Fuel Advisory.
+Table 3 . Summary of weather, volume, and equipment, runway and other delay characteristics.3CharacteristicWeatherEquip.,VolumeRunway& OtherMean66%20%14%Standard21%16%11%deviation 1 ρ0.950.270.210.940.160.112 ρ
+Table 4 . Summary of departure, enroute, arrival and TMS delay characteristics.4Characteristic Departure TMS Arrival EnrouteMean47%39%14%1%Standard17%17%8%1%deviation 1 ρ0.820.860.550.450.730.860.510.492 ρAmerican Institute of Aeronautics and Astronautics and 13 April 2004, are shown in Table5.NAS delay data for the same two days were previously itemized in Table
+Table 6 . Summary of aircraft delayed on ground versus aircraft delayed in the air.6CharacteristicDelayed onDelayed in AirGroundMean74%26%Median76%24%Standard Deviation11%11%Minimum23%7%Maximum93%77%
+Table 5 . OPSNET GS and GDP delay data.5Data Variable4/10/20044/13/2004Ground Stops# of Aircraft Delayed398Minutes of Delay2927,079Average Delay97.3372.23Ground Delay Program# of Aircraft Delayed61,044Minutes of Delay24485,166Average Delay40.6681.57Total Delays Due to GS and GDP# of Aircraft Delayed91,142Minutes of Delay53692,245Average Delay59.5580.77American Institute of Aeronautics and Astronautics
+Table 7 . Centers: Summary of Domestic Operations Report.7Data Variable4/10/20044/13/2004DeparturesAir Carrier17,12220,231Air Taxi9,13212,831General Aviation5,7059,000OverflightsAir Carrier23,02123,753Air Taxi3,5564,318General Aviation2,7634,763
+Table 8 . Summary of properties of the ten groups.8American Institute of Aeronautics and Astronautics show that there are fewer days in groups associated with large delays.For example, group number ten consists of only two days compared to group number one with 145 days.Observe that the mean values associated with the groups are approximately equally spaced and that the standard deviation values are fairly close to each other.Standard deviation values can be expected to increase with fewer groups.Probability density functions corresponding to Gaussian distributions with the group means and standard deviations listed in Table8are shown in FigureGroupNumberMeanStandardMinimumMaximumIDof DaysDelayDeviationDelayDelay(min.)(min.)(min.)(min.)114511,3024,3101,68618,834212626,6264,48919,10734,62839243,0265,09534,94752,11345761,2395,19852,56270,01153880,0905,42171,18690,464632 102,8205,94494,023112,031711 124,6523,612119,692131,17287 141,6335,854133,884148,34197 163,5015,887156,717172,211102 183,4264,083180,539186,313
+Table 10 . Three groups based on number of aircraft delayed in the airborne phase.10To illustrate the utility of the multiple-metric classification technique, an example of classifying 522 days, which included the 517 days discussed previously and five special days used in earlier studies, into groups is presented next.These five special days are 5/17/2002, 4/17/2005, 4/21/2005, 6/5/2005 and 7/15/2005.5/17/2002 is the ACES baseline day.The other four days were used earlier in Ref. 7. They were categorized as a low-volume low-weather day, high-volume low-weather day, low-volume high-weather day and high-volume high-weather day, respectively in Ref. 7.GroupNumberMeanStandardMinimumMaximumNumberof DaysNumberDeviationNumberNumberDelayedDelayedDelayed13931244972182128314100222970311891018911891
+Table 13 . Three groups based on number of aircraft delayed on the ground.13GroupNumberMeanStandardMinimumMaximumNumberof DaysNumberDeviationNumberNumberDelayedDelayedDelayed12493791604864621859131736511,2933861,6893281,3092,778
+Table 12 . Groups based on number of aircraft delayed in the airborne phase (excluding 7/15/2005 and 6/5/2005).12GroupNumberMeanStandardMinimumMaximumNumberof DaysNumberDeviationNumberNumberDelayedDelayedDelayed129110336715821742143815929535538271304604
+Table 11 . Three groups based on number of aircraft delayed in the airborne phase (excluding 7/15/2005).11GroupNumberMeanStandardMinimumMaximumNumberof DaysNumberDeviationNumberNumberDelayedDelayedDelayed1386123477213213430481214604319700970970
+Table 14 . Two groups based on total domestic departure counts.14GroupNumberMeanStandardMinimumMaximumNumberof DaysDelayDeviationDelayDelay(min.)(min.)(min.)(min.)116834,7922,92025,67740,528235246,3352,24340,59651,759
+Table 15 . Final grouping with three-metric classification.15Organizing the resultingtriple index group IDsusing a "dictionary sort" algorithm, each unique group and its members areGroup NumberGroup ID Number of Days1 μμ2μ3σ1σ2σ3determined. The values of1 (1, 1, 1)94 33,97031083 2,718 15837the metrics of the members2 (1, 1, 2)14 33,698444 224 3,179 14941are determination of minimum, then used for3 (1, 1, 3) 4 (1, 2, 1)2 34,650 27 36,060373 348 6,588 120 880 110 2,565 15241 33maximum,meanand5 (1, 2, 2)19 36,816951 221 1,916 15937standard deviation values associated with the groups.6 (1, 2, 3) 7 (1, 3, 1)2 36,583 3 35,219 1,713 131 4,418 301 951 390 4,185 30595 8Results of this process for8 (1, 3, 2)6 37,067 1,586 213 2,296 20430thethree-metric9 (1, 3, 3)1 36,485 1,462 355000classification being discussed here, are problem, outlined in Table 15. Group means for the three metrics are listed in the columns labeled 1 μ , 2 μ and 3 μ ; standard deviation values are listed in the columns labeled as 1 σ , 2 σ and 3 σ .10 (2, 1, 1) 11 (2, 1, 2) 12 (2, 1, 3) 13 (2, 2, 1) 14 (2, 2, 2) 15 (2, 2, 3) 16 (2, 3, 1) 17 (2, 3, 2) 18 (2, 3, 3)89 45,495 47 46,195 3 43,944 60 47,330 50 46,393 27 45,934 18 46,686 1,614 116 2,190 326 383 110 2,200 151 483 204 2,008 117 505 447 2,494 148 898 113 1,961 188 911 205 2,383 156 953 359 2,201 196 38 47,027 1,728 231 1,947 380 20 46,539 1,721 408 2,537 26131 33 52 31 36 58 31 41 81
+Table 16 . Days in small groups.16GroupGroup ID DateNumber3(1, 1, 3)2/22/20033(1, 1, 3)5/16/20046(1, 2, 3)6/14/20036(1, 2, 3)3/28/20047(1, 3, 1)9/14/20037(1, 3, 1)5/23/20047(1, 3, 1)5/30/20049(1, 3, 3)1/4/200412(2, 1, 3)5/2/200312(2, 1, 3)5/5/200312(2, 1, 3)9/30/2003
+Table 17 . Final grouping obtained using departure counts and delays by cause metrics.17Group NumberGroup IDNumber of Days1 μμ2μ3μ4σ1σ2σ3σ41 (1, 1, 1, 1)86 33,6482684972 2,932 16036442 (1, 1, 1, 2)22 35,24335760 252 2,172 17439643 (1, 1, 1, 3)2 33,69115479 498175 1685844 (1, 1, 2, 1)7 35,796355 21289 1,339 15961345 (1, 1, 2, 2)3 37,448402 198 259794 25128616 (1, 1, 2, 3)2 36,120352 229 4958894020 1447 (1, 1, 3, 1)1 35,484296 401 17300008 (1, 1, 3, 2)1 39,454394 541 23800009 (1, 2, 1, 1)25 35,7709635090 2,920 209354610 (1, 2, 1, 2)7 36,85496080 238 2,653 302386811 (1, 2, 1, 3)2 37,02196273 587378 38530 17012 (1, 2, 2, 1)2 37,118960 291 105 3,430 35312813 (1, 2, 2, 2)2 38,540741 306 24965941 100 16414 (1, 2, 3, 2)1 36,485898 710 209000015 (1, 3, 1, 1)4 37,143 1,56740 148 3,049 132322516 (1, 3, 2, 1)2 37,814 1,867 177 1295802711117 (2, 1, 1, 1)78 45,50832580 104 1,774 171263618 (2, 1, 1, 2)21 45,74033085 234 2,349 173214719 (2, 1, 1, 3)8 45,69037091 458 1,609 210248320 (2, 1, 2, 1)29 46,356337 19498 2,204 181463621 (2, 1, 2, 2)21 47,496319 215 274 2,293 129516522 (2, 1, 2, 3)7 47,610314 195 491 1,867 148365823 (2, 1, 3, 1)2 45,004411 412 147 6,233 270331324 (2, 1, 3, 2)2 50,358192 382 265 1,98185181125 (2, 1, 3, 3)1 48,983407 551 540000026 (2, 2, 1, 1)64 46,0909408096 2,318 217303927 (2, 2, 1, 2)27 46,39297796 250 2,481 220284828 (2, 2, 1, 3)3 47,117 1,172 106 555 2,916 33124 29329 (2, 2, 2, 1)19 46,817898 20596 1,910 216513830 (2, 2, 2, 2)17 47,587 1,021 205 262 2,533 241535831 (2, 2, 2, 3)3 47,360765 424 407 4,315 142 337 11332 (2, 2, 3, 3)2 45,890744 429 434 1,87064207133 (2, 3, 1, 1)26 46,221 1,9406680 1,646 341304234 (2, 3, 1, 2)13 47,142 1,89788 252 2,020 279266135 (2, 3, 1, 3)2 45,527 1,943 109 410 3,341 5868336 (2, 3, 2, 1)5 47,467 1,735 209 120482 227343737 (2, 3, 2, 2)3 49,992 2.055 160 280 1,271 9151770
+Table 1818lists the group membership of holidays and special days -ACES baseline day, Joint Planning and Development Office (JPDO) baseline day and days studied in Ref. 7. Group IDs fromTable 15 are listed under Group ID 1 heading and from Table 17 under Group ID 2 heading.Results in this table show that the domestic departure counts are generally lower on holidays.Group ID 1 (2, 2, 2) indicates that the ACES baseline day has high departure counts, moderate number of aircraft delayed on the ground, and moderate number of aircraft delayed in the air; Group ID 2 (2, 2, 1, 1) indicates high departure counts, moderate number of aircraft delayed due to weather, low number of aircraft delayed due to volume and low number of aircraft delayed due to equipment, runway and other issues.The two group IDs for the JPDO baseline day indicate high departure counts, moderate number of aircraft delayed on ground, low number of aircraft delayed in the air, low number of aircraft delayed due to weather, moderate number of aircraft delayed due to volume and high number of aircraft delayed due to equipment, runway and other conditions.Ref. 7 considered 4/17/2005 to be a low departure count, low-delay due to weather day.The group IDs in Table 18 label this day as a high departure count, low-delay due to weather day.The departure count of 40,653 on this day is at the lower end of the high departure count group can be inferred from the statistics given in
+Table 18 . Classification of holidays and special days.18Number DateSignificanceDay of WeekGroup ID 1Group ID 215/17/2002 ACES Baseline DayFriday(2, 2, 2)(2, 2, 1, 1)21/1/2003 New Year's DayWednesday(1, 1, 1)(1, 1, 1, 1)31/20/2003 Martin Luther King DayMonday(1, 1, 1)(1, 1, 1, 1)42/17/2003 President's DayMonday(1, 1, 1)(1, 1, 1, 1)55/26/2003 Memorial DayMonday(1, 1, 1)(1, 1, 1, 1)67/4/2003 Independence DayFriday(1, 1, 1)(1, 1, 1, 1)79/1/2003 Labor DayMonday(1, 2, 1)(1, 2, 1, 1)8 10/13/2003 Columbus DayMonday(2, 1, 1)(2, 1, 2, 1)9 11/11/2003 Veterans DayTuesday(2, 2, 1)(2, 1, 2, 1)10 11/25/2003 Two Days Before ThanksgivingTuesday(2, 3, 1)(2, 1, 3, 3)11 11/27/2003 Thanksgiving DayThursday(1, 1, 2)(1, 1, 1, 1)12 12/25/2003 Christmas DayThursday(1, 1, 1)(1, 1, 1, 1)131/1/2004 New Year's DayThursday(1, 1, 1)(1, 1, 1, 1)141/19/2004 Martin Luther King DayMonday(2, 1, 1)(2, 1, 2, 2)152/16/2004 President's DayMonday(2, 2, 2)(2, 2, 2, 2)162/19/2004 JPDO Baseline DayThursday(2, 2, 1)(2, 1, 2, 3)175/31/2004 Memorial DayMonday(1, 3, 2)(1, 3, 2, 1)184/17/2005 Ref. 7 L/L DaySunday(2, 1, 1)(2, 1, 1, 2)194/21/2005 Ref. 7 H/L DayThursday(2, 1, 2)(2, 1, 3, 2)206/5/2005 Ref. 7 L/H DaySundayNot included (1, 3, 1, 1)217/15/2005 Ref. 7 H/H DayFridayNot included (2, 3, 2, 2)
+Table 14 .14Classification of 4/21/2005 as high departure count, low-delay due to weather day is in agreement with Ref. 7 except that many aircraft were delayed due to volume on this day.The results in Table 18 for 6/5/2005 and 7/15/2005 are in agreement with Ref. 7. Both these days experienced an inordinate amount of airborne and ground delays due to weather.A number of aircraft were also delayed due to volume, and equipment, runway and other issues on 7/15/2005.
+		
+		
+			
+
+				
+
+
+	
+		Aggregate Statistics of the National Airspace System
+		
+			JimmyKrozel
+		
+		
+			BobHoffman
+		
+		
+			StevePenny
+		
+		
+			TarynButler
+		
+		10.2514/6.2003-5710
+	
+	
+		AIAA Guidance, Navigation, and Control Conference and Exhibit
+		Herndon, VA
+		
+			American Institute of Aeronautics and Astronautics
+			20170. October 2002
+		
+	
+	Suite 200
+	Krozel, J., Hoffman, J., Penny, S., and Butler, T., "Selection of Datasets for NAS-Wide Simulation Validations," Metron Aviation, Inc., 131 Elden St., Suite 200, Herndon, VA 20170, October 2002.
+
+
+
+	
+		A Cluster Analysis to Classify Days in the National Airspace System
+		
+			BobHoffman
+		
+		
+			JimmyKrozel
+		
+		
+			StevePenny
+		
+		
+			AnindyaRoy
+		
+		
+			KarlinRoth
+		
+		10.2514/6.2003-5711
+		AIAA-2003-5711
+	
+	
+		AIAA Guidance, Navigation, and Control Conference and Exhibit
+		Austin, TX
+		
+			American Institute of Aeronautics and Astronautics
+			August 11-14, 2003
+		
+	
+	Hoffman, B., Krozel, J., Roy, A.., and Roth, K., "A Cluster Analysis to Classify Days in the National Airspace System," AIAA-2003-5711, Proceedings of AIAA Guidance, Navigation, and Control Conference, Austin, TX, August 11-14, 2003.
+
+
+
+	
+		Some Methods for Classification and Analysis of Multivariate Observations
+		
+			JBMacqueen
+		
+	
+	
+		Proceedings of the 5 th Berkeley Symposium on Mathematical Statistics and Probability
+		the 5 th Berkeley Symposium on Mathematical Statistics and ProbabilityBerkeley
+		
+			University of California Press
+			1967
+			
+		
+	
+	MacQueen, J. B., "Some Methods for Classification and Analysis of Multivariate Observations," Proceedings of the 5 th Berkeley Symposium on Mathematical Statistics and Probability, University of California Press, Berkeley, 1967, pp. 281-297.
+
+
+
+	
+		Aggregate Statistics of the National Airspace System
+		
+			JimmyKrozel
+		
+		
+			BobHoffman
+		
+		
+			StevePenny
+		
+		
+			TarynButler
+		
+		10.2514/6.2003-5710
+		AIAA-2003- 5710
+	
+	
+		AIAA Guidance, Navigation, and Control Conference and Exhibit
+		Austin, TX
+		
+			American Institute of Aeronautics and Astronautics
+			August 11-14, 2003. October 1, 2004
+			5
+		
+	
+	Aggregate Statistics of the National Airspace System
+	Krozel, J., Hoffman, B., Penny, S., and Butler, T., "Aggregate Statistics of the National Airspace System," AIAA-2003- 5710, Proceedings of AIAA Guidance, Navigation, and Control Conference, Austin, TX, August 11-14, 2003. 5 Federal Aviation Administration, "Order 7210.55C: Operational Data Reporting Requirements," U. S. Department of Transportation, October 1, 2004.
+
+
+
+	
+		The Future National Airspace System: Design Requirements Imposed by Weather Constraints
+		
+			JimmyKrozel
+		
+		
+			BrianCapozzi
+		
+		
+			TonyAndre
+		
+		
+			PhilSmith
+		
+		10.2514/6.2003-5769
+		AIAA-2003-5769
+	
+	
+		AIAA Guidance, Navigation, and Control Conference and Exhibit
+		Austin, TX
+		
+			American Institute of Aeronautics and Astronautics
+			August 11-14, 2003
+		
+	
+	Krozel, J., Capozzi, B., Andre, A. D., and Smith, P., "The Future National Airspace System: Design Requirements Imposed By Weather Constraints," AIAA-2003-5769, Proceedings of AIAA Guidance, Navigation, and Control Conference, Austin, TX, August 11-14, 2003.
+
+
+
+	
+		Validating the Airspace Concept Evaluation System for Different Weather Days
+		
+			ShannonZelinski
+		
+		
+			LarryMeyn
+		
+		10.2514/6.2006-6115
+	
+	
+		AIAA Modeling and Simulation Technologies Conference and Exhibit
+		Keystone, CO
+		
+			American Institute of Aeronautics and Astronautics
+			August 21-24, 2006
+		
+	
+	AIAA 2006-6115
+	Zelinski, S., and Meyn, L., "validating The Airspace Concept Evaluation System For Different Weather Days," AIAA 2006-6115, Proceedings of AIAA Modeling and Simulation Technologies Conference and Exhibit, Keystone, CO, August 21-24, 2006.
+
+
+
+	
+		Build 4 of the Airspace Concept Evaluation System
+		
+			LMeyn
+		
+	
+	
+		Proceedings of AIAA Modeling and Simulation Technologies Conference and Exhibit
+		AIAA Modeling and Simulation Technologies Conference and ExhibitKeystone, Colorado
+		
+			August 21-24, 2006
+		
+	
+	AIAA-2006-6110
+	Meyn, L., et al, "Build 4 of the Airspace Concept Evaluation System," AIAA-2006-6110, Proceedings of AIAA Modeling and Simulation Technologies Conference and Exhibit, Keystone, Colorado, August 21-24, 2006.
+
+
+
+	
+		Design of Center-TRACON Automation System
+		
+			HErzberger
+		
+		
+			TJDavis
+		
+		
+			SGreen
+		
+	
+	
+		AGARD Conference Proceedings 538: Guidance and Control Symposium on Machine Intelligence in Air Traffic Management
+		Berlin, Germany
+		
+			May 11-14, 1993
+		
+	
+	Erzberger, H., Davis, T. J., and Green, S., "Design of Center-TRACON Automation System," AGARD Conference Proceedings 538: Guidance and Control Symposium on Machine Intelligence in Air Traffic Management, Berlin, Germany, May 11-14, 1993.
+
+
+
+	
+		Data-Centric Air Traffic Management Decision Support Tool Model
+		
+			JamesMurphy
+		
+		
+			RonaldReisman
+		
+		
+			RobSavoye
+		
+		10.2514/6.2006-7830
+		AIAA-2006-7830
+	
+	
+		6th AIAA Aviation Technology, Integration and Operations Conference (ATIO)
+		Wichita, Kansas
+		
+			American Institute of Aeronautics and Astronautics
+			September 25-27, 2006
+		
+	
+	Murphy, J. R., Reisman, R., and Savoye, R., "A Data-Centric Air Traffic Management Decision Support Tool Model," AIAA-2006-7830, Proceedings of 6 th AIAA Aviation Technology, Integration and Operations Conference (ATIO), Wichita, Kansas, September 25-27, 2006.
+
+
+
+	
+		FACET: Future ATM Concepts Evaluation Tool
+		
+			KarlDBilimoria
+		
+		
+			BanavarSridhar
+		
+		
+			ShonRGrabbe
+		
+		
+			GanoBChatterji
+		
+		
+			KapilSSheth
+		
+		10.2514/atcq.9.1.1
+	
+	
+		Air Traffic Control Quarterly
+		Air Traffic Control Quarterly
+		1064-3818
+		2472-5757
+		
+			9
+			1
+			
+			2001
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	Bilimoria, K. D., Sridhar, B., Chatterji, G. B., Sheth, K. S., and Grabbe, S. R., "FACET: Future ATM Concepts Evaluation Tool," Air Traffic Control Quarterly, Vol. 9, No. 1, 2001, pp. 1-20. American Institute of Aeronautics and Astronautics
+
+
+				
+			
+		
+	

file125.txt

@@ -0,0 +1,358 @@
+
+	
+	
+		
+Introductionhis paper describes the automated scenario generation process recently developed and implemented in the Air Traffic Management (ATM) Testbed (ATMTB).The ATMTB was formerly known as the Shadow Mode Assessment with Realistic Technologies (SMART) for the National Airspace System (NAS) Testbed (SMART-NAS Testbed (SNTB)).Motivation for the development of this testbed at the National Aeronautics and Space Administration (NASA) is to enable benefit, impact, safety and cost assessments for accelerating the deployment of Concept and Technologies (C&T) in the NAS.Today, C&T introduction into the NAS takes decades.The primary reason for this is an inability to assess the operational impact of the interaction between the proposed C&T and operationally deployed systems in terms of NAS-wide safety, traffic flow efficiency, roles and workload of controllers and traffic managers, and impact on fleet operations.Transition of C&T to operations requires mathematical modeling and simulation, Human-in-the-Loop (HITL) testing and shadow-mode evaluation driven by operational data.Cautious, slow and incremental steps are typically taken towards deployment because of limitations in each of these steps.This includes HITLs limited to a few scenarios, pilots and controllers, and the inability to inject decisions derived from a shadow-mode system into the operations for impact and benefit assessment.Whereas interaction with the operational system during testing and stages of deployment is not permissible due to safety concerns, it is possible to create a simulation environment that closely mimics the NAS using the same operational systems/hardware for enabling such assessments.Driven by this objective, ATMTB is developing infrastructure to enable mathematical modeling, HITLs and testing with operational systems in a simulated environment.The primary motivation for automated scenario generation for HITL simulations is the difficulty of creating scenarios manually.For example, traffic scenarios for the Multi-Aircraft Control System (MACS) 1 , used frequently for HITL-based air traffic concept evaluations at NASA, are generated manually by first creating an initial scenario (seed-scenario) by selecting flight-plans from recorded air traffic data and then modifying it by repeatedly running it in MACS until the characteristics desired for meeting the objectives of the HITL test are achieved.This process is time consuming.Even creating a seed-scenario that results in successful MACS simulation is tedious because of missing and erroneous data.Because of these difficulties, researchers typically base their experimental evaluations on only a few days of data.The evaluation of a concept or technology's system-wide impacts in terms of cost and benefits with one or two days of data is of limited utility.Therefore, the second motivation for automated scenario generation is that these evaluations should instead be conducted with many days of data with distinct/desired characteristics, given the availability of archived data.In the past several years, because of the decreasing cost of storage, large volumes of aviation related data have been collected by several organizations including NASA and the Federal Aviation Administration (FAA).NASA has recently invested in cleaning up and improving the consistency of the archived data.The scenario generation capability has been significantly enhanced this year to download these data files directly from the storage location and generalized to create surface traffic scenarios for Airspace Target Generator (ATG) and flight scenarios for Airspace and Traffic Operations Simulation (ATOS) 2 in addition to scenarios for MACS.This capability has been used to generate MACS scenarios for Dynamic Routes for Arrivals in Weather (DRAW) 3 and Integrated Demand Management 4 HITLs, and ATG scenarios for Airspace Technology Demonstration (ATD-2) 5 .It is currently being used to generate MACS scenarios for Instrument Flight Rules (IFR) traffic, Visual Flight Rules (VFR) traffic and the expected Urban Air Mobility (UAM) traffic for HITL-based evaluations under the ATM-eXploration (ATM-X) project 6 to enable future UAM vehicles to operate in the NAS.The rest of the paper is organized as follows.Because the examples and the results in this paper are focused on MACS traffic scenarios for HITL-based investigations of operational feasibility of the Integrated Demand Management (IDM) concept, the IDM concept is briefly described in Section II.This discussion also highlights some of the difficulties associated with creating scenarios that represent realistic conditions.The manual scenario generation process is outlined in Section III.The automated scenario generation process is discussed in Section IV.Validation of the seed-scenario, comparison of the seed-scenario with the HITL-scenario, and comparison of the HITL-scenario input with the MACS simulation output are described in Section V.The seed-scenario was created using the automated scenario generation process whereas the HITL-scenario was created by manually altering the seed-scenario.Finally, the main findings are summarized in the Section VI.
+II. Integrated Demand Management HITL SetupIntegrated Demand Management (IDM) 7 is a Trajectory Based Operations (TBO) concept to collaboratively organize aircraft trajectories into well-managed flows that match traffic demand to the available capacity.The concept leverages FAA and NASA pre-departure, enroute and arrival technologies to achieve this objective.IDM uses Traffic Flow Management System (TFMS) tools to precondition traffic into the airspace domain of the Time-Based Flow Management (TBFM) system.If it was possible to predict future capacity and flight times accurately, the preconditioned traffic would arrive at the metering locations as intended; TBFM would only impose small delays required for meeting the runway spacing constraint.Unfortunately, incorrect capacity forecast, delayed departure from the airport, wind and weather introduce uncertainty to the arrival time forecast, which disrupts the schedule and sequence intended by preconditioning.TBFM then has to impose additional delays to adjust the schedule for complying with the capacity constraints at the meter fix and runway.Given that the uncertainty is higher and the cost of delay is lower when the aircraft are on the ground compared to when they are airborne and close to the TBFM freeze-horizon boundary, a proper balance between TFMS and TBFM delays is needed for reducing fuel consumption (by delaying as little as possible while airborne), maintaining the airline schedule and fully utilizing the available airport capacity.Several HITL and Automation-In-The-Loop experiments have been completed to investigate the operational feasibility of the IDM concept under realistic conditions.The testbed is currently being enhanced to support fasttime Monte-Carlo simulations for IDM concept evaluations. 8These experiments typically have the structure presented in Fig. 1.MACS simulates air traffic data based on the input traffic and weather/wind scenario files; it also provides a high-fidelity air traffic control simulation environment for controller and pilot interactions.In conjunction with MACS, an emulation of the Collaborative Trajectory Options Program (CTOP), called nCTOP (NASA CTOP), was constructed to perform the key functions of the TFMS version with CTOP capability.The nCTOP and MACS Planner Station blocks shown in Fig. 1 represent emulation of the TFMS with CTOP used at the Air Traffic Control System Command Center.Key functions of nCTOP includes setting capacity constraints at an FCA, automatically assigning delay and allocating trajectories to the pre-departures to balance the predicted arrival traffic demand at the FCA according to its capacity limit.Inputs therefore include the capacity scenario being simulated, and the schedule and Trajectory Options Sets (TOS) likely to be submitted by flight operators.Expect Departure Clearance Times (EDCT) and TOS allocation are output to all MACS stations through the MACS simulation manager.MACS stations for each pilot and controller communicate with all other MACS stations in the simulation, updating aircraft positions.A research version of the FAA's operational TBFM version 4.2.3 with NASA modifications is used to simulate the generation of arrival timelines; controllers are able to reschedule internal departures to fit into the overhead stream based on calculated Scheduled Times of Arrival (STA) at the metering locations such as meter fixes and runway threshold.The experimental setup in Fig. 1 illustrates an example of a concept that requires multiple scenarios; IDM requires capacity scenarios, scheduled traffic scenarios, likely airline TOS, detailed MACS traffic scenarios, and weather scenarios, including convective weather and wind.In addition, the scenarios derived from traffic and weather often require significant modification to meet the desired characteristics for the experiment, which in some instances can reduce experiment realism.For example, in the experiments described in Ref. 9, the baseline traffic scenario derived from recorded traffic from a single day -July 22, 2014 was modified based on feedback from subject matter experts to have the most representative characteristics of the nominal operations into Newark Liberty International (KEWR) during a clear weather day.This five-hour scenario included a total of 66 aircraft.Experiments were ultimately run investigating two wind severity levels-mild and heavy wind, and two traffic demand profiles with different distributions.However, in reality, under such wind conditions, airlines might have filed flight-plans differently compared to the ones in the traffic scenario.Availability of a technique for identification of days with the appropriate clear weather, wind conditions and traffic demand profile would have provided increased realism, as well as reduced the time required to generate the scenario.Future experiments will add significant complexity as convective weather is introduced at different locations 10 , making the generation of realistic scenarios even more challenging.Examples of such scenarios are the use of coded departure routes because of predicted convective weather activity downstream of the TBFM freeze-horizon, and the use of tactical rerouting (e.g., common tactical routes) due to unpredicted convective weather blocking an arrival gate.The IDM example discussed in this section illustrates some of the challenges for automated scenario generation.At the present time, only generation of seed traffic scenarios that run in MACS, ATOS and ATG are being considered.These seed-scenarios would have to be modified based on subject matter expert feedback and to meet additional experiment requirements not reflected in the seed-scenarios.Discussions in the following sections are limited to traffic scenario generation for MACS simulations.
+III. Manual Scenario Generation ProcessThe manual MACS scenario generation process consists of the nine steps summarized below.1) Identify the desired scenario characteristics based on the experimental objectives.a) Determine the general characteristics that serve the purpose of the study.b) Talk to the Subject Matter Experts (SME) to augment and refine the desired characteristics.2) Search for the day and the time-period.a) Select the days with the desired weather conditions.b) Check Aviation System Performance Metrics (ASPM) data for those days, selected in the previous step, and see how the traffic demand evolved for the desired runway configurations.If the runway configurations in the ASPM data do not match the desired runway configurations, a relatively easy scenario editing step is employed in Step 5 to specify the routes to the desired runways instead.c) Choose the time-period based on the desired scenario characteristics.3) Download the Center-TRACON Automation System (CTAS) data for the selected day from the storage location.TRACON is acronym for Terminal Radar Approach Control.4) Convert the downloaded CTAS data into the MACS scenario format using the TCSim Route Analyzer/Constructor (TRAC).5) Modify the scenario if needed by Step 2b and look for any obvious errors in the scenario editor.6) Play the scenario in TRAC and make a determination of its suitability for MACS simulation based on traffic evolution.7) Run an open-loop MACS simulation with the generated-scenario for the time-period, chosen in Step 2c, and analyze the resulting MACS outputs to determine the extent to which the simulation meets the scenario requirements.8) Augment the analyzed-scenario with additional data for meeting the remaining scenario requirements that could not be met in the earlier steps.This step might consist of adding flights, for example, from different flows, regions, hours and days to increase traffic volume.9) Repeat Steps 5 through 8 until all the scenario requirements are met.Step 1 of the manual scenario generation process will stay the same for the automated scenario generation process because the automated scenario generation process will also have to output a scenario in accordance to the desired scenario characteristics.In Step 2, researchers use a guess-and-try technique by first picking a few days that they guess might meet the scenario characteristics identified in Step 1 and then examining the ASPM data for those days.An exhaustive search of such days in a year for example would be difficult to accomplish following the current manual process.It might be feasible to automate this step by enabling search based on surface, enroute and terminal traffic and weather metrics from multiple sources organized in databases and in groups, where the groups could be based on unsupervised/supervised classification techniques employing big-data technologies with data driven metrics/metadata derived from NASA's ATM-data-warehouse.It could be designed to support complex queries such as "find all days in 2016 that are like 01/20/2016" and "find days in 2015 with severe weather within 300 miles from Newark airport and Ground Delay Program (GDP) in Chicago."This might become a significant capability in the future for accelerating concept evaluation and acceptance because it will provide a large set of scenarios representing different operational conditions instead of the few manually-created scenarios for concept evaluation and acceptance testing.Steps 3 and 4 of the manual scenario generation process are quick and accomplished using computer programs.Researchers have reported that the initial MACS scenario file output by TRAC tool from Step 4 requires a lot of manual data entry in Step 5 due to missing and erroneous data.Researchers often resort to looking at old scenario files and talk to SMEs to determine reasonable values to enter missing and to replace erroneous data.Also, the route from the entry location to the runway has to be created by manually copying the filed flight-plan into a column in the scenario file and then modifying them.The automated scenario generation process in the ATMTB creates this route as a sequence of waypoints from the entry point, location of which is derived from track-data, to the closest point ahead of the entry point on the filed flight-plan followed by the waypoints in flight-plan till the end of the Standard Terminal Arrival Route (STAR) and then waypoints along the approach procedure to the designated runway.Approach procedures are defined in the MACS adaptation data.Steps 6 and 7 will eventually be replaced by the verification step of the automated scenario generation process.Whereas it is difficult to completely automate Step 8, it might be possible to automate it partially by creating scenarios for different days and conditions, and then judiciously combining them with the seed-scenario to create a scenario that meets the needs of the experiment.
+IV. ATMTB Automated Scenario Generation ProcessATMTB infrastructure at its present stage of development can be described in terms of the following elements-(1) web-based frontend and backend, (2) Simulation Architect, (3) publish-subscribe messaging middleware, (4) Component Library, (5) simulation management, and (6) scenario generation.The web-based frontend and backend enable the user to interact with the ATMTB for tasks such as composing a simulation, running a simulation and retrieving output data.The Simulation Architect application launched from the web frontend provides a graphical user interface for enabling the user to drag-and-drop and connect predefined (user defined and testbed native) blocks for composing a simulation/scenario generation task.The Simulation Architect writes a set of instructions for simulation management based on block properties such as the component (executable) associated with a particular block, and the links between the blocks.Links specify the input and output relationships between the blocks, which defines the publisher and subscriber relationships in the simulation.Management of the distributed simulation is accomplished by Execution and Component Managers.Execution Manager interprets the instructions provided by the Simulation Architect to instruct the Component Managers to download components from the Component Library to the designated computers and to start them up.Once started, components interact with each other by publishing messages and subscribing to messages delivered using the messaging middleware.Unlike the other five elements, which are testbed infrastructure elements, the scenario generation capability is an application that runs on the testbed.The scenario generation capability was initially developed for creating traffic scenarios for MACS simulations.The automated scenario generation process in ATMTB is initiated by dragging and dropping blocks, specifying the block properties and linking the blocks graphically using the Simulation Architect.The Simulation Architect view for composing MACS scenario generation is shown in Fig. 2. The blocks labeled-Data Loader, Data Filters, and MACS Scenario Builder are parts of the scenario generation program.The preliminary step of scenario generation consists of the user picking a day (date) and specifying it as a property of the MACS Scenario Builder block for the scenario generation program to download the traffic data file from the storage location and read the associated traffic data during runtime.The type of traffic file to be downloaded is specified by selecting the appropriate Data Loader block; Fig. 2 shows the setup for loading ATAC (a particular format) data.The properties specified in Data Filter bocks and the "and"/"or" relations specified by chaining Data Filter bocks in the simulation builder provides instructions for the scenario generation program for reducing (down-selecting) the input traffic data.For example, three data filters can be chained together in series to tell the program to first select arrivals to Newark Liberty International (KEWR) based on the Arrival Airport property of the first filter block, then select aircraft landing on Runway 22L based on the Landing Runway property of the second filter block, and finally select aircraft landing between 17:00 UTC 6/6/2016 and 5:00 UTC 6/7/2016 based on Event Time property of the third filter block.Inclusion of the MACS Scenario Builder block tells the scenario generator to build a scenario for MACS simulation.Other blocks with inputs to the MACS Scenario Builder block instruct the scenario generation program to use the filtered data, aircraft performance models, adaptation data, wind data and initial conditions.The links between the blocks specify the data flow.The output of the Simulation Architect is a set of instructions for the Execution Manager that includes a configuration file for the generation component.The Execution Manager instructs the Component Manager to download the scenario generation program executable from the Component Library to a particular machine and to start it.The Component Manager also provides the configuration file, created by the Simulation Architect and provided by the Execution Manager, to the scenario generation program for generating the scenario.MACS scenarios are generated by processing recorded air traffic data archived in the ATM-data-warehouse in three different types of files-Reduced Record (RD), Event Data (EV) and Integrated Flight Format (IFF).RD files contain a single record for each flight, where the record contains information such as the reference time, unique key, aircraft ID, aircraft type, beacon code, airline, origin (airport or Fix-Radial-Distance (FRD)), takeoff/landing runway, destination airport, top-of-climb/top-of-descent time, runway threshold arrival/departure time, flight-plan (including route) data, and sector/center transition list.EV files contain multiple records for events related to each flight such as reference time, unique key, aircraft ID, aircraft type, event time, event type, object class, old name and new name.MACS scenario generation currently processes takeoff/landing and crossing events, which includes sector, center and TRACON crossings.Object class, old name and new name provide additional information related to the event.For example, complete information for a takeoff from San Francisco (SFO) airport would be available in the EV file as event type-takeoff, object class-airport, old name-"?"(not needed for takeoff event) and new name-SFO.Similarly, a center crossing event for a flight leaving Oakland Center (ZOA) and entering Los Angeles Center (ZLA) would be available as event type-crossing, object class-center, old name-ZOA and new name-ZLA.IFF files contain multiple records for each flight, where the records contain all flight-plans including amended flight-plans and track-data.Data associated with these records include reference time, unique key, aircraft ID, aircraft type, message type (for example, filed flight-plan and amended flight-plan), origin (airport or FRD), destination airport and filed altitude.Data contained in the RD and EV records are especially useful for filtering the traffic data for building the scenarios.The IFF data are useful for augmenting the traffic data derived from RD and EV records.Three types of filters are currently available.RD String Filters are used for selecting records from RD files by matching specified strings to those in the records.Supported filters include Aircraft Type, Airline, Arrival/Departure Airport, Aircraft ID, Center, Sector, and Landing/Takeoff Runway.Filter and List of Strings are properties of the RD String Filter block; the user selects the desired filter from the list of filters and provides a list of strings appropriate for the selected filter.For example, airport code such as KEWR is a string that is compatible with the Arrival/Departure Airport filter.Similarly, 22L is an appropriate string in the list of strings with the Landing Runway filter option.The RD Airport Proximity filter is used for selecting flights to/from airports either inside or outside the specified region by processing RD records.The user sets up the filter by selecting from a list of options related to the properties and inputting the values needed by the properties.Supported properties include Filter, Reference Location, Reference Distance and Airports Included.Options associated with the Filter property are Departure Airport and Arrival Airport; the Reference Location property expects an airport code like KEWR; the Reference Distance property expects distance in nautical miles; the Airports Included property expects values such as all inside, all outside and a list of specified airports codes like KEWR.Finally, Event Time Filter uses EV records to select flights.The Event Time Filter block has Event Type, Minimum Value, Maximum Value and Include/Exclude properties.Examples of Event Type are Landing, Takeoff, Top-of-Climb and Top-of-Descent.Minimum and Maximum Values are day (yearmonth-date) and UTC time (hour-minute-second)).The Include/Exclude property option specifies whether the flight events within the specified time interval are to be included or excluded.In addition to the selection of data specified using filter blocks on the Simulation Architect, Entry Track Method, Entry State Method, Aircraft Performance Model, Airspace Adaptation Database and Atmosphere Model have to be specified as shown in Fig. 2. Figure 3 shows the various inputs that have to be specified for MACS scenario generation and the choices associated with them.Three options for the Entry Track Method relevant to MACS scenario generation are: Distance, Start Time and Top-of-Descent.Target Airport ID and Distance from the Airport are the two parameters of the Distance block.Starting locations of the selected flights are chosen to be inside/outside the circular region defined by these two parameters.Start Time block enables the user to input the desired time past the simulation start time for selecting the starting position.For example, if the desired time is 30 minutes, the position of the flight at or just after when the simulation time is 30 minutes past the simulation start time would be chosen as the starting position.The Top-of-Descent block allows the user to specify a time with respect to top-of-
+V. Validation and Comparison of Automatically Generated and Manually Refined ScenariosThe discussion and the results in this section pertain to the seed-scenario, HITL-scenario and the MACS simulation output; Fig. 5 summarizes the procedure for creating them.The ATAC data are used by the automated scenario generation procedure, described in the previous section, to create the seed-scenario.This seed-scenario is then manually refined to create the HITL-scenario.Finally, traffic is simulated using MACS with HITL-scenario as input.Two sets of results are presented below.The first set compares the seed-scenario with the HITL-scenario, Blocks 3 and 5, and the second set compares the HITLscenario with the MACS simulation output, Blocks 5 and 7 in Fig. 5.The seed-scenario for MACS simulation of arrival traffic to KEWR spanning six-hours starting at 17:00 UTC was created by processing June 6, 2016 RD, EV and IFF files archived in the ATM-data-warehouse.The seedscenario has 299 flights with 274 landing on Runway 22L, six on 22R and one on 29.Arrival runway could not be determined for the remaining 18 aircraft.Two types of analysis were done to characterize the seed-scenario.The first type consisted of determining the number of flights associated with the same parameter value such as call-sign and beacon-code.Figure 6 and7 show the number of flights associated with the same callsign and aircraft-type, respectively.Table 1 summarizes these results for different parameters.For example, of the 290 unique call-signs, 9 call-signs were associated with more than one flight; Fig. 6 shows that each of the nine callsigns were associated with two aircraft.Of the 35 different aircraft types in the seed-scenario, 24 (see the second row of Table 1) were associated with several aircraft as shown in Fig. 7. Similarly, one destination airport, KEWR, was associated with every flight.Of the four landing runways-22L, 22R, 29 and "not-set", one aircraft landed on 29, 274 on 22L, six on 22R and 18 did not have an assigned runway (not-set category).Thus, one flight was  The two types of analysis proved to be very useful for determining errors in the scenario.For example, the entry point IAS histogram in Fig. 8 shows that the scenario generation program determined the IAS of an aircraft to be 712 knots.The Mach number for the passenger aircraft associated with this flight was determined to be 1.7, which is wrong.Whereas checks were built into the scenario generation program, the checks are not always successful because of data quality issues.In this particular instance, several successive actual track-data reports used for determining the entry state were erroneous.Figure 8 also shows that 78 aircraft had the correct entry point IAS of zero because they were on the ground at the simulation start time.The cruise altitude histogram showed that seven flights had a cruise altitude of zero, which is incorrect.Results suggest that these types of analyses should be included as an extension to the automated scenario generation process to remove flights with improper parameters from the seed-scenario.In addition to detecting data quality issues, an important aspect of validation is determining the reasonableness of the scenario.For example, it is not desirable for several flights to have the same call-sign in the HITL-scenario.There are two possible ways of addressing this issue.One is to create new call-   number column of the MACS scenario file, where Mach number is determined using BADA model speeds if the cruise altitude is above the Mach transition altitude.This implies that the researcher should run the MACS simulation with wind data.If the researcher uses the file without wind data, MACS would simulate flights with unrealistic groundspeed.Analyses for generating the results for the paper suggest that if realistic landing rate is desired in the scenario and the researcher wishes to run the scenario without wind data for example, average cruise groundspeed should be output in the cruise speed/Mach column of the scenario file.Figure 9 shows the actual landing rate at KEWR, and the predicted landing rate using Eq. ( 1) with average cruise groundspeed.Landing rate is determined as the number of flights in the hourly window, continuously shifted temporally at a fiveminute interval.The figure suggests that MACS scenario with average cruise groundspeed would result in a scenario that would reasonably replicate the actual landing rate.The two types of analyses done for the seed-scenario were repeated for the HITL-scenario to determine the differences between them.The manually refined HITLscenario that was used for the IDM HITL in March 2018 was created by the researcher by selecting flights from the seed-scenario and altering some of the values such as cruise speeds and entry time to achieve the desired landing rate.To have the demand exceed arrival capacity of 40 aircraft/hour, entry times of flights in the seed-scenario were altered to squeeze six-hours of arrival traffic into five-hours for creating the HITL-scenario.The HITLscenario has 191 flights, a subset of flights in the seedscenario, with all landing on Runway 22L.Other than three flights, all the flights in the HITL-scenario are in the seed-scenario.All flights from the seed-scenario within a 40 nautical-mile circular region around KEWR were not selected for the HITLscenario; some flights were rejected if their entry time was less than 30 minutes past 17:00 UTC.Flights were also removed in an attempt to maintain the ratio of the number of internal flights to the total number of flights in the HITL-scenario to the 23% seen in the seed-scenario, where the internal flights are those that originated within the 400 nautical-mile circular region surrounding KEWR.The ratio of the internal to the total flights in the HITL-scenario was found to be 30%.Results summarized for the seed-scenario in Table 1 are provided for the HITLscenario in Table 2.This table shows that the flights in the HITL-scenario had a unique call-sign, and that they landed on the same runway (Runway 22L).The ratios of "Once to Unique" and "Repeated to Unique" in Table 1 and 2 expressed as percentage are shown side-by-side in Table 3.This table shows that most ratios seen in the seed-scenario are maintained in the HITL-scenario except for the entry point sector-ID.Compared   to seed-scenario with 70 entry point sector-IDs, the HITL-scenario had three sector-IDs: ZDC-01, ZOB-01 and ZBW-01, which were assigned to 65, 74 and 52 flights, respectively.Tailoring of the HITL-scenario to achieve the objective of higher traffic demand with respect to airport arrival rate of 40 aircraft per hour, which was realized by squeezing six-hours of traffic into five-hours, is apparent in Fig. 10. Figure 10 shows the actual and the predicted landing rate graphs for the HITL-scenario.The actual landing rate graph is based on the actual landing time of 191 aircraft in the HITL-scenario whereas the predicted landing rate graph is based on Eq. (1).Comparing Figs. 9 and 10 it is seen that several flights arriving during the early part of the scenario were removed from the seedscenario to create a gradually increasing traffic demand in the HITL-scenario.The increase in traffic demand achieved in the HITLscenario can also be achieved by an algorithm as follows.Let the desired arrival rate be n aircraft/hour.The desired temporal separation, t  , between successive aircraft is then 60/ n minutes.Thus, ( 1) ( )LL t i t i t    (2)where () L ti is the landing time of the leading aircraft and ( 1)L ti  is the landing time of the following aircraft.Solution of the recursion Eq. ( 2) is ( ) (1) ( 1)LL t i t i t    (3)where (1) L t is the landing time of the first aircraft and 1 i  .Combining Eq. ( 2) with (1), the entry times can be determined as,. () ( ) (1) ( 1) () R EL Avg li t i t i t Vi     (4)The final step of the validation process is comparison of the MACS simulated traffic with that intended by the scenario.Figure 11 shows the comparison of the predicted landing rate with the MACS simulated traffic landing rate Figure 10.HITL-scenario KEWR landing rate.using the HITL-scenario.Analysis showed that the predicted landing rate graph is sensitive to the cruise speed.As expected, faster cruise speeds shift the graph to the left and slower to the right along the abscissa.The difference between the two graphs seen in Fig. 11 is due conversion of Mach to cruise speed (true airspeed) and the aircraft performance models employed in MACS.requires true airspeed to be specified below Mach transition altitude and Mach above it.Mach numbers specified in the HITL-scenario were converted to true airspeed using standard atmosphere for predicting the landing rate.Using June 6, 2016 RUC data for this conversion could have resulted in a slightly different outcome.An additional source of error is that 18 aircraft in the MACS simulation came close to landing but did not actually land, they continued flying past the runway.To create a substantial scenario validation capability, the analyses described in the paper will need to be extended.One such example is the ability to determine the deviation of the MACS simulated track-data with respect to the flight-plan specified in the input scenario data.This could help identify errors in the flight-plan, missing waypoints in the MACS adaptation database, and MACS trajectory modeling errors.
+VI. ConclusionsThe automated scenario generation process recently developed and implemented in the Air Traffic Management Testbed being developed at the NASA Ames Research Center was described.The earlier manual scenario generation process for generating Multi-Aircraft Control System scenarios for use in the Human-in-the-Loop experiments was described to motivate automated scenario generation.Two scenarios were analyzed: (1) the seed-scenario generated using the automated scenario generation method and (2) the Human-in-the-Loop-scenario created by a researcher starting from the seed-scenario.Results summarized in tables show that many of the characteristics seen in the seed-scenario are preserved in the Human-in-the-Loop-scenario.Two types of analyses were described for comparing the seed and the Human-in-the-Loop scenarios.The first type analyzed duplicate parameters associated with flights such as call-sign, beacon-code and entry point sector-ID; the second type examined the distributions of route length, cruise speed, cruise altitude, actual landing time, predicted landing time, entry time, and entry point speed and altitude.Results obtained suggest these analyses are useful for determining data quality issues and for eliminating flights with unreasonable parameter values from the seed-scenario.Landing rate based on Multi-Aircraft Control System simulated traffic using the Human-in-the-Loop-scenario were compared with the expected landing rate based on the route length and average cruise speed of flights in the Humanin-the-Loop-scenario. Causes for the differences seen in the landing rates were identified.Close examination of the Human-in-the-Loop-scenario revealed that many of the desired characteristics such as flights having unique callsigns and airport arrival rate demand exceeding the airport arrival rate capacity can also be achieved in the seedscenario by enhancing the automated scenario generation process.A method for altering the entry time of flights to get the desired landing rate was described as an example of such enhancement.and confirm the findings.Hyo-Sang provided the HITL-scenario, a copy of the MACS software that he had used for the HITL and the adaptation data required for running that version of MACS.Hyo-Sang also described the manual scenario generation process for MACS simulations that had been used prior to the automated scenario generation process described in this paper.Authors are grateful to Dr. Min Xue, Dr. Antony Evans, Shannon Zelinski and Dr. Banavar Sridhar for reviewing the paper and providing feedback.Contributions of the other ATM Testbed team members-John Robinson, James Murphy, Alan Lee, Chok Fung (Jack) Lai, Phu (Phil) Huynh and Huu Huynh to the idea, design and development of the scenario generation capability is gratefully acknowledged.Figure 1 .1Figure 1.Example IDM experimental setup.
+Figure 2 .2Figure 2. Simulation Architect view for composing MACS scenario generation.
+descent for selecting the initial position of the flight.A value of -5 minutes for example would result in the selection of the position five minutes (or slightly more because track-data might not be available exactly at 5 minutes) prior to the time the flight reaches the top-of-descent point.At the current stage of development, there is a single option associated with each of the other inputs needed for generating MACS scenarios.The only option available for the Entry State Method is From Track.Inclusion of the From Track block the scenario generator to use actual track data and the Mach transition altitude, determined using Base of Aircraft Data (BADA) 11 aircraft performance model and the specified atmospheric model, to determine the state of the flight such as altitude, heading, calibrated airspeed and Mach number at the entry location.The only option for Aircraft Performance Model is BADA Model block, and for Airspace Adaptation Database is National Flight Data Center (NFDC) Database block.Two options for the Atmosphere Model are Rapid Refresh block and Standard Atmosphere block.The steps for MACS scenario generation starting from loading and filtering the traffic data to output of scenario data in a file are summarized in Fig. 4. The first step consists of loading RD, EV and IFF files from ATM-data-warehouse and filtering traffic data according to the filters specified on the Simulation Architect, and creating the flight data structure.The second step consists of assigning a BADA aircraft model in the flight data structure based on aircraft type and BADA Synonym List, and sorting the flight-plans of each flight by time.BADA Synonym List enables mapping of aircraft types that do not exist in the BADA database to the ones that exist in the database.The next step consists of finding the entry track data of the flights based on the simulation start time and the Entry Track Method specified on the Simulation Architect.Entry track data consist of time, latitude and longitude, altitude, groundspeed, course heading, Rate of Climb or Descent (ROCD) and sector ID of the entry point.The last flight-plan prior to entry track time is determined in the
+Figure 3 .3Figure 3. Inputs and associated options for MACS scenario generation.
+Figure 4 .4Figure 4. Summary of MACS scenario generation steps.
+Figure 5 .5Figure 5. Summary of scenario and MACS output data generation steps.
+tis the predicted landing time, E t is the entry time (takeoff time for aircraft on the ground), R l is the route length and .Avg V is the average cruise groundspeed, which is determined by averaging the actual cruise speed derived from track-data within the top-of-climb and top-of-descent interval.Predicted landing rate comparison with the actual landing rate is useful for sanity check.
+Figure 6.Flights with the same call-sign.
+Figure 7 .7Figure 7. Flights with the same aircraft-type.
+Figure 8 .8Figure 8. Entry point IAS.
+Figure 9 .9Figure 9. Seed-scenario KEWR landing rate.
+Figure 11 .11Figure 11.MACS simulated using HITL-scenario versus HITL-scenario KEWR landing.rate.
+Table 1 .1Summary of seed-scenario results.#ParameterOnce RepeatedUnique1.Call-sign28192902.Aircraft-type1124353.Destination airport0114.Landing runway1345.MACS flight-plan148471956.ATC flight-plan174412157.Beacon-code256212778.Departure airports50681189.Entry point altitude736113410. Entry point IAS775813511. Entry point sector-ID47237012. Aircraft weight52429
+Table 2 .2Summary of HITL-scenario results.#ParameterOnce RepeatedUnique1.Call-sign19101912.Aircraft-type1020303.Destination airport0114.Landing runway0115.MACS flight-plan64411056.ATC flight-plan80401207.Beacon-code18151868.Departure airports4150919.Entry point altitude35468110. Entry point IAS23143711. Entry point sector-ID03312. Aircraft weight31619
+Table 3 .3Comparison of seed-scenario with HITL-scenario.Seed-scenarioHITL-scenario
+		
+		
+
+			
+AcknowledgementsThe authors thank Dr. Antony Evans and Dr. Hyo-Sang Yoo.This paper would not have been possible without their help.Tony described the IDM concept and ran MACS with the HITL-scenario to identify some issues
+			
+
+			
+
+				
+
+
+	
+		Human-In-the-Loop Evaluation of NextGen Concepts in the Airspace Operations Laboratory
+		
+			ThomasPrevot
+		
+		
+			PaulLee
+		
+		
+			ToddCallantine
+		
+		
+			JoeyMercer
+		
+		
+			JeffreyHomola
+		
+		
+			NancySmith
+		
+		
+			EverettPalmer
+		
+		10.2514/6.2010-7609
+	
+	
+		AIAA Modeling and Simulation Technologies Conference
+		Toronto, Ontario, Canada
+		
+			American Institute of Aeronautics and Astronautics
+			August 2-5, 2010
+		
+	
+	Prevot, T., et. al., "Human-in-the-Loop Evaluation of NextGen Concepts in the Airspace Operations Laboratory," AIAA 2010-7609, AIAA Modeling and Simulation Technologies Conference, Toronto, Ontario, Canada, August 2-5, 2010.
+
+
+
+	
+		A Multi-Operator Simulation for Investigation of Distributed Air Traffic Management Concepts
+		
+			MarkPeters
+		
+		
+			MarkBallin
+		
+		
+			JSSakosky
+		
+		10.2514/6.2002-4596
+	
+	
+		AIAA Modeling and Simulation Technologies Conference and Exhibit
+		Monterey, California
+		
+			American Institute of Aeronautics and Astronautics
+			August 5-8, 2002
+		
+	
+	Peters, M. E., Ballin, M. G., and Sakosky, J. S., "A Multi-Operator Simulation for Investigation of Distributed Air Traffic Management Concepts," AIAA Modeling and Simulation Technologies Conference and Exhibit, Monterey, California, August 5-8, 2002.
+
+
+
+	
+		Dynamic Arrival Routes: A Trajectory-Based Weather Avoidance System for Merging Arrivals and Metering
+		
+			ChesterGong
+		
+		
+			DaveMcnally
+		
+		10.2514/6.2015-3394
+	
+	
+		15th AIAA Aviation Technology, Integration, and Operations Conference
+		Dallas, Texas
+		
+			American Institute of Aeronautics and Astronautics
+			June 22-26, 2015
+		
+	
+	Gong, C., McNally, D., and Lee, C. H., "Dynamic Arrival Routes: A Trajectory-Based Weather Avoidance System for Merging Arrivals and Metering," AIAA 2015-3394, 15 th AIAA Aviation Technology, Integration, and Operations Conference, Dallas, Texas, June 22-26, 2015.
+
+
+
+	
+		Integrated Demand Management: Coordinating Strategic and Tactical Flow Scheduling Operations
+		
+			NancyMSmith
+		
+		
+			ConnieBrasil
+		
+		
+			PaulULee
+		
+		
+			NathanBuckley
+		
+		
+			ConradGabriel
+		
+		
+			ChristophPMohlenbrink
+		
+		
+			FaisalOmar
+		
+		
+			BonnyParke
+		
+		
+			ConstantineSperidakos
+		
+		
+			Hyo-SangYoo
+		
+		10.2514/6.2016-4221
+	
+	
+		16th AIAA Aviation Technology, Integration, and Operations Conference
+		Washington, DC
+		
+			American Institute of Aeronautics and Astronautics
+			2016
+		
+	
+	Smith, N. M., et. al., "Integrated Demand Management: Coordinating Strategic and Tactical Flow Scheduling Operations," 16 th AIAA Aviation Technology, Integration, and Operations Conference, Washington, DC, 2016.
+
+
+
+	
+		Airspace Technology Demonstration 2 (ATD-2) Technology Description Document
+		
+			AGing
+		
+		Memorandum: NASA/TM-2018-219767
+		
+	
+	
+		21076-1320
+		Hanover, MD
+		
+			NASA Center for AeroSpace Information
+			7115
+		
+	
+	NASA Technical
+	Standard Drive 5301. cited 5/10/2018
+	Ging, A., et. al., "Airspace Technology Demonstration 2 (ATD-2) Technology Description Document," NASA Technical Memorandum: NASA/TM-2018-219767, NASA Center for AeroSpace Information, 7115 Standard Drive 5301, Hanover, MD 21076-1320. URL: https://www.aviationsystemsdivision.arc.nasa.gov/publications/2018/NASA-TM-2018-219767.pdf [cited 5/10/2018].
+
+
+
+	
+		ATM-X: Air Traffic Management -eXploration
+		
+			WNChan
+		
+		
+	
+	
+		Partnership Workshop
+		Moffett Field, California
+		
+			April, 2018
+		
+	
+	cited 5/10/2018
+	Chan, W. N., "ATM-X: Air Traffic Management -eXploration," Partnership Workshop, NASA Ames Research Center, Moffett Field, California, April, 2018. URL: https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20180002413.pdf. [cited 5/10/2018].
+
+
+
+	
+		Evaluation of Integrated Demand Management looking into Strategic & Tactical Flow Management
+		
+			CMoehlenbrink
+		
+		
+	
+	
+		Europe Air Traffic Management Research and Development Seminar
+		
+			June 27-30, 2017
+			Seattle, Washington
+		
+	
+	12 th USA/. cited 5/10/2018
+	7 Moehlenbrink, C., et. al., "Evaluation of Integrated Demand Management looking into Strategic & Tactical Flow Management," 12 th USA/Europe Air Traffic Management Research and Development Seminar, Seattle, Washington, June 27-30, 2017. URL: http://www.atmseminarus.org/seminarContent/seminar12/papers/12th_ATM_RD_Seminar_paper_51.pdf [cited 5/10/2018].
+
+
+
+	
+		Using an Automated Air Traffic Simulation Capability for a Parametric Study in Traffic Flow Management
+		
+			HeatherArneson
+		
+		
+			AntonyDEvans
+		
+		
+			DeepakKulkarni
+		
+		
+			PaulULee
+		
+		
+			JinhuaLi
+		
+		
+			MeiYWei
+		
+		10.2514/6.2018-3665
+	
+	
+		2018 Aviation Technology, Integration, and Operations Conference
+		Atlanta, GA
+		
+			American Institute of Aeronautics and Astronautics
+			June 24-28, 2018
+		
+	
+	8 Arneson, H., Evans, A. D., Kulkarni, D., Lee, P., Li, J., Wei, M. Y., "Using an Automated Air Traffic Simulation Capability for a Parametric Study in Traffic Flow Management," 18th AIAA Aviation Technology, Integration, and Operations Conference, Atlanta, GA, June 24-28, 2018.
+
+
+
+	
+		Required time of arrival as a control mechanism to mitigate uncertainty in arrival traffic demand management
+		
+			Hyo-SangYoo
+		
+		
+			ChristophMohlenbrink
+		
+		
+			ConnieBrasil
+		
+		
+			NathanBuckley
+		
+		
+			AlGlobus
+		
+		
+			NancyMSmith
+		
+		
+			PaulULee
+		
+		10.1109/dasc.2016.7778013
+	
+	
+		2016 IEEE/AIAA 35th Digital Avionics Systems Conference (DASC)
+		Sacramento, California
+		
+			IEEE
+			2016
+		
+	
+	Yoo, H., et. al., "Required Time of Arrival as a Control Mechanism to Mitigate Uncertainty in Arrival Traffic Management," 35 th IEEE Digital Avionics Systems Conference, Sacramento, California, 2016.
+
+
+
+	
+		Impact of Different Trajectory Option Set Participation Levels within an Air Traffic Management Collaborative Trajectory Option Program
+		
+			Hyo-SangYoo
+		
+		
+			ConnieBrasil
+		
+		
+			NancyMSmith
+		
+		
+			NathanBuckley
+		
+		
+			GitaHodell
+		
+		
+			ScottKalush
+		
+		
+			PaulULee
+		
+		10.2514/6.2018-3040
+	
+	
+		2018 Aviation Technology, Integration, and Operations Conference
+		Atlanta, GA
+		
+			American Institute of Aeronautics and Astronautics
+			June 24-28, 2018
+		
+	
+	Yoo, H., et. al., "Impact of Different Trajectory Option Set Participation Levels within an Air Traffic Management Collaborative Trajectory Option Program," 18 th AIAA Aviation Technology, Integration, and Operations Conference, Atlanta, GA, June 24-28, 2018.
+
+
+
+	
+		User Manual for the Base of Aircraft Data (BADA) Revision 3.6
+		
+			ExperimentalEurocontrol
+		
+		
+			Centre
+		
+		No. 10/04
+		
+			July, 2004
+			Eurocontrol Experimental Centre Publications Office, B.P
+			15
+			Bretigny-sur-orge, France
+		
+	
+	EEC Note
+	Eurocontrol Experimental Centre, "User Manual for the Base of Aircraft Data (BADA) Revision 3.6," EEC Note No. 10/04, Eurocontrol Experimental Centre Publications Office, B.P. 15, 91222 -Bretigny-sur-orge, France, July, 2004.
+
+
+				
+			
+		
+	

file126.txt

@@ -0,0 +1,377 @@
+
+	
+	
+		
+Introductionhe automated scenario generation capability in the Air Traffic Management (ATM) Testbed (ATMTB) has been used to generate Multi-Aircraft Control System (MACS) scenarios for Human-in-the-Loop (HITL) evaluations.MACS is a distributed simulation system with multiple pseudo-pilot and air traffic controller stations that is frequently used for National Aeronautics and Space Administration (NASA)'s evaluations of air traffic management concepts. 1utomated scenario generation has been used for creating MACS scenarios for Dynamic Routes for Arrivals in Weather (DRAW) 2 and Integrated Demand Management 3 HITL experiments.It has also been used for creating Airspace Target Generator (ATG) scenarios, used in realistic airport surface traffic simulation, for the Airspace Technology Demonstration (ATD-2) 4 subproject.Automated scenario generation is currently being used to generate MACS scenarios for Instrument Flight Rules (IFR) traffic, Visual Flight Rules (VFR) traffic and the expected Urban Air Mobility (UAM) traffic for evaluations under the ATM-eXploration (ATM-X) project 5 to enable simulation of future UAM vehicles to operate in the National Airspace System (NAS).This paper describes a two-step automated process for creating MACS traffic scenarios according to the desired scenario characteristics.The first step of the two-step procedure described in this paper employs the ATMTB automated scenario generation process first described in Ref. 6.The second step, introduced in this paper, enhances the scenario, created by the first step, by eliminating flights with unreasonable values associated with parameters such as cruise speed and cruise altitude, and selecting/adjusting flights based on desired scenario characteristics specified by the user such as route length, entry time, landing rate and ratio of number of flights inside to outside the terminal area (for example with respect to the Time-Based Flow Management freeze horizon).Prior to the automated scenario generation capability in ATMTB, traffic scenarios for MACS were generated manually by first creating an initial scenario (seed-scenario) by selecting flight-plans from recorded air traffic data
+II. BackgroundThe research focus for IDM 7 Trajectory Based Operations (TBO) concepts is improving the efficiency and predictability of air traffic operations.Many of the air traffic management tools and technologies in the enroute, terminal-area and surface domains developed both by the Federal Aviation Administration (FAA) and NASA that are used for operations in the National Airspace System have been difficult to use in the congested Northeast region airspace due to a mix of high traffic-volume, weather conditions, the proximity of major airports in the New York Metroplex and in neighboring centers, airspace geometry, and operational procedures for separating the flows in and out of major airports.TBO concepts seek to collaboratively organize aircraft trajectories into well-managed flows that match traffic demand to the available capacity by initially leveraging FAA and NASA pre-departure, enroute, arrivaldeparture and surface technologies. 8he charter of the IDM project is to explore ways of integrating near-term to mid-term NextGen traffic management systems to improve efficiency in situations when the traffic demand exceeds capacity of resources such as airports and airspace.In the IDM concept, FAA's Traffic Flow Management System (TFMS) and the Time-Based Flow Management (TBFM) system are used. 7In the current system, TFMS is used strategically for determining departure delays at the airports of origin in response to constraints at destination airports and flow constrained areas (FCA).TBFM is a terminal-area traffic management system that tactically assigns a scheduled time of arrival (STA) to arrivals at the metering locations such as meter-fixes, meter-arcs and runway threshold based on capacity constraints at those locations.TFMS uses a Ration-By-Schedule algorithm that is based on the STA filed by the aircraft operator while TBFM computes an estimated time of arrival (ETA) using track and flight-plan data for its STA assignment.IDM seeks to establish coordination between strategic TFMS and tactical TBFM decisions for reducing delays by using TFMS to precondition traffic into the airspace domain of the TBFM system.Unfortunately, incorrect capacity forecast, delayed departure from the airport, wind and weather introduce uncertainty to the arrival time forecast, which disrupts the schedule and sequence intended by preconditioning.TBFM then has to impose additional delays to adjust the schedule for complying with the capacity constraints at the metering locations.Given that the uncertainty is higher and the cost of delay is lower when the aircraft are on the ground compared to when they are airborne and close to the TBFM freeze-horizon boundary, a proper balance between TFMS and TBFM delays is needed for reducing fuel consumption (by delaying as little as possible while airborne), maintaining the airline schedule and fully utilizing the available airport capacity.Several HITL and Automation-In-The-Loop experiments have been completed to investigate the operational feasibility of the IDM concept under realistic conditions. 9,10 ecently, fast-time Monte-Carlo simulations are also being developed for IDM concept evaluations. 11In these experiments, MACS simulates air traffic data based on the input traffic and weather/wind scenario files; it also provides a high-fidelity air traffic control simulation environment for controller and pilot interactions.The progress made on the IDM project will be continued for developing a concept of operations and accompanying system architecture that evaluates the integration of FAA's systems and NASA's Airspace Technology Demonstration (ATD) ground-based and airborne systems for a future service-oriented airspace system. 8This concept of operations needs to include the operation of new entrants such as supersonic aircraft, space launch vehicles, high-altitude long endurance platforms, Unmanned Aerial Systems and Urban Air Mobility vehicles in addition to the traditional airspace users like airlines and general aviation aircraft.The concept of operations developed is expected to be evaluated in a HITL simulation in the future using Northeast region scenarios developed with service provider and user inputs.
+III. Traffic Scenario SelectionThis section describes the analysis of runway configurations at JFK, EWR, LGA and TEB with the objective of identifying days with high-volume of arrival and departure traffic for creating MACS traffic scenarios for TBO studies.Because traffic flow patterns on the airport surface and in the terminal airspace depend on the runway configuration (runways used for arrivals and runways used for departures), hourly arrival and departure data for every day in 2017 were obtained for JFK, EWR, LGA and TEB airports from the FAA's Aviation System Performance Metrics (ASPM) database using the information in the Throughput Analysis Standard Report.These data were processed to determine the total number of hours each configuration was in use and the total number of operations (sum of arrivals and departures) conducted in each configuration during 2017.The top five configurations based on the percentage of hours in use are summarized in Table 1.The second column in the table shows the arrival runways such as 31L and 31R and departure runways such as 31L separated by a vertical bar (see the first row of the second column in Table 1).The third column shows the total number of hours the particular configuration was in use during the year.The fourth, fifth and the sixth columns present the total number of arrivals, departures and their sum, for the corresponding configurations, respectively.Finally, the seventh and the eighth columns list the percentages based on the total number of hours the airport was in operation and the total number of operations during the year.These percentages were computed by removing data corresponding to when the airport was closed or the configuration information was absent.Next, the configurations at JFK, EWR, LGA and TEB were considered together at every hour of every day in 2017 to determine the most frequently used combinations of configurations.These combinations are summarized in Table 2. Comparing these tables, one observes that the top five configurations at JFK, EWR, LGA and TEB in Table 1 are used 74%, 84%, 86% and 96% of the time (based on hours), respectively, while the top five combinations of the configurations in Table 2 are only used 17% of the time.Of the 687 unique combinations of the configurations observed in the 2017 data, ten were used  2% of the time, 12 were used 1% of the time and the remaining 665 were used less than 1% of the time.The top ten and the top 22 combinations were used 28% and 40% of the time, respectively.It also turned out that the top five configurations listed in Table 2 consist of combinations of the top five most frequently used configurations at the four airports listed in Table 1.The configuration in the second row in Table 2 is composed of the most frequently used configuration at JFK, EWR, LGA and TEB in Table 1.Finally, the sum of the operations at JFK, EWR, LGA and TEB were computed for every hour of every day in 2017 and the configuration data were sorted in non-increasing order of the sum of operations to identify dates and times with large number of operations.These results are summarized in Table 3.While archived traffic data do exist  3).This interval was also chosen because it had balanced operations (ratio of arrivals to departures close to one) with 865 arrivals and 784 departures.The MACS scenario was generated using the 5/23/2017 air traffic data archived in the ATM-data-warehouse to ensure that the traffic simulation would result in most flights arriving at JFK, EWR, LGA and TEB within this six-hour period.The MACS scenario generation is described next.
+IV.Step
+1: Automated Scenario Generation ProcessThe automated scenario generation process in ATMTB is initiated by dragging and dropping predefined blocks (user defined and testbed native), specifying the block properties and linking the blocks graphically using the Simulation Architect, where the Simulation Architect application is launched from the web frontend as described in Ref. 6.While the details are available in Ref. 6, the description below is included as background for the second step of the two-step scenario generation process.MACS scenarios are generated by processing recorded air traffic data archived in the ATM-data-warehouse, which is a platform for collecting, archiving, processing, querying and retrieving ATM data.Processed data derived from FAA's System-Wide Information Management (SWIM) data are available in the ATM-data-warehouse in three different types of files-Reduced Record (RD), Event Data (EV) and Integrated Flight Format (IFF).RD files contain a single record for each flight, where the record contains information such as the reference time, unique key, aircraft ID, aircraft type, beacon code, airline, origin (airport or Fix-Radial-Distance (FRD)), takeoff/landing runway, destination airport, top-of-climb/top-of-descent time, runway threshold arrival/departure time, flight-plan (including route) data, and sector/center transition list.EV files contain multiple records for events related to each flight such as reference time, unique key, aircraft ID, aircraft type, event time and event type.MACS scenario generation currently processes takeoff/landing and crossing events, which includes sector, center and TRACON crossings.IFF files contain multiple records for each flight, where the records contain all flight-plans including amended flight-plans and trackdata.Figure 2 shows the various inputs that have to be specified for MACS scenario generation and the choices associated with them.Data contained in the RD and EV records are especially useful for filtering the traffic data for building the scenarios.The IFF data are useful for augmenting the traffic data derived from RD and EV records.Three types of filters are currently available.RD String Filters are used for selecting records from RD files by matching specified strings to those in the records.The user selects the desired filter from the list of filters and provides a list of strings appropriate for the selected filter.For example, airport code such as KEWR is a string that is compatible with the Arrival/Departure Airport filter.Similarly, 22L is an appropriate string in the list of strings with the Landing Runway filter option.The RD Airport Proximity filter is used for selecting flights to/from airports either inside or outside the specified region by processing RD records.The filter is set up by selecting from a list of options related to the properties like Reference Location and Reference Distance and inputting the values needed by the properties.or a named fix, along the flight-plan for connecting the entry point to the flight-plan, and builds the MACS route for the flight starting at the entry point and ending at the landing runway.Entry state data are determined in the sixth-step using trackdata, which is specified by selecting the From Track block, the only available Entry State Method, on the Simulation Architect.Entry state data consist of true heading, calibrated airspeed, Mach, flight state (overflight, arrival or departure), and in-Mach or in-CAS mode at the entry point.MACS requires a target waypoint with speed and altitude constraints to be specified.The target waypoint is specified based on the flight state at the entry point.For flights in takeoff and climb phase at the entry point, the first waypoint after top-of-climb is chosen to be the target waypoint.If the flight is in cruise phase at the entry point, the next waypoint is chosen to be the target waypoint; if the next waypoint is beyond top-of-descent, the next waypoint with speed and altitude constraints on the approach route is chosen as the target waypoint; else, the airport is chosen as the target waypoint.If the flight is in descent phase at the entry point, the next waypoint with speed and altitude constraints on the approach route is chosen as the target waypoint.If the approach route is missing, the airport is chosen as the target waypoint.Data for several comment fields in the MACS scenario file are generated in the seventh-step.These data are useful for debugging and analysis.Values for all the data fields specified in the header of the version of MACS being used are assigned in the eighth-step based on the computations done in the earlier steps.The scenario data are output in a file in the last-step shown in Fig. 3.
+V.Step 2: Scenario RefinementThe automated scenario generation process described in the previous section was employed to create a MACS scenario file for arrivals to JFK, EWR, LGA and TEB using 5/23/2017 traffic data archived in ATM data warehouse.All flights, including the ones with flight plans and the ones without flight plans, landing during the 18 through 23 UTC interval were considered for inclusion in the scenario.For flights without flight plans, the track data were processed to create flight-plans by specifying their route as a sequence of latitude-longitude pairs from the starting location to the destination airport.Cruise altitude and cruise speed were assigned based on the maximum altitude and maximum groundspeed seen in the track-data of these flights.The simulation start-time for creating the MACS scenario was chosen to be 12 UTC, which is six-hours prior to 18 UTC, to ensure that all flights in the scenario are able to land within the 18 UTC to 23 UTC time interval.The automated scenario generation process with these traffic data and scenario parameters resulted in the MACS scenario file with 808 flights, 57 fewer arrivals compared to 865 reported in ASPM.Starting with the scenario file with 808 flights, the scenario refinement steps in Fig. 4 were employed to improve the data quality and to adjust the scenario for meeting experiment requirements.Flights with route-length of less than 20 nautical-miles eliminated in the first step reduced the number of flights by 17.In the second step, none of the flights were removed by filtering based on cruise speed because the minimum speed of 127 knots and the maximum speed of 571 knots are reasonable.Next, in step 3, filtering based on cruise altitude of more than 600 feet eliminated 22 additional flights whose cruise altitude was zero, most likely because of missing altitude information.VFR aircraft without a transponder with altitude reporting capability are not required to provide altitude reports automatically to air traffic control.The possibility of assigning a reasonable cruise altitude based on the performance characteristics of the type of aircraft and the length of route could be investigated in the future.In step 4, the entry time filter was implemented to remove flights with an entry time of 30 minutes past the simulation start-time as suggested in Ref. 6.This step did not filter any flights because the earliest entry time in this dataset was one-hour and 28-minutes.The internal to external flights ratio filter (step 5) is designed to eliminate a number of shorter and longer flights from the flights in the scenario file to achieve the desired ratio.Flights are categorized as internal flights if the length of route is less than a prescribed threshold and external otherwise.A threshold value of 400 nautical-miles was used in this study.Of the 769 remaining flights at the end of stepfour in Fig. 4, 238 were categorized as internal flights and 531 as external flights with the resulting internal to external flights ratio of 0.45.Let, the desired ratio be r , 1x be the number of external flights in the dataset, 2x be the number of internal flights in the dataset, 1sx be the number of selected external flights and 2sx be the number of selected internal flights.The procedure for selecting 1sx and 2sxsuch that 2 1 s s x r x  (1)is given by: and ifss xx x x x r rx      (3)These two solutions can be written together as follows using the Iverson's notation:            2 1 1 2 1 1 2 2 1 2 1 1 2 2 1 1 s s x x rx x x rx x r x rx x rx rx x x                (4)where the logical expressions within the square-brackets mean a value of one or zero depending on whether they are true or false.Table 4 shows the number of internal and external flights for different values of r .Observe that a large value of r like 300 results in only internal flights to be selected; with zero external flights selected, the desired internal to external ratio is  .Finally, the internal and external fights are selected by first sorting the lists of internal flights and external flights in non-increasing order of route length, and then picking the required number starting from the top of the two lists.Alternatively, the required numbers can be selected randomly from the two unordered lists of internal and external flights.Step 5 for selecting internal and external flights based on their desired ratio, though implemented, was not applied; all 769 flights in the scenario were accepted.Figure 5 shows the hourly arrival traffic count considering arrivals to i is temporally separated from the leading aircraft 1 i  by more than t  , set the scheduled arrival time of the following aircraft to its originally proposed time of arrival; if not separated, add t  to the scheduled arrival time of the leading aircraft and assign it to the following aircraft as the scheduled time of arrival.Thus, this scheduler only delays aircraft.The resulting delay is given as( ) ( ( ) ( 1)) ( )( 1)p s p s i t t i t i t i t i t             (7)where the logical expression inside the square-brackets means a value of one when true and zero otherwise.Scheduling results were generated by imposing an hourly arrival capacity constraint of 100 aircraft.Figure 6 shows the original unconstrained hourly arrival traffic counts as a function of time (shown earlier in Fig. 5) and the traffic counts resulting from scheduling traffic using Eq.(6) to meet the specified capacity constraint.The average delay was found to be 49 minutes, maximum delay was one-hour and 44 minutes, and total delay was 634 hours.To achieve the desired arrival schedule of traffic simulated by MACS, the entry time of the flights have to be adjusted.Because the flight time is given as:  Figures 7 and8 show the histograms of the original entry times and the new scheduled entry times.Observe the reduction in the number of flights in the bins and the spreading of the flights and entry times to beyond 12 hours past 12 UTC in Fig. 8.f p E t t t  (8)The scheduling procedure described in this section is useful both for IDM HITL and fast-time MACS based simulations.This procedure can be used for allocating ground delay in response to airport capacity constraint forecast to achieve strategic traffic flow management objectives.The same procedure can then be used to allocate airborne delay to flights in the terminal area in response to actual airport capacity constraint to achieve tactical traffic flow management objectives.A slightly modified version of the scheduler in Eq. ( 6), ( ) ( 1)ss t i t i t    (10)where the delay, t  , could be changing as a function of time, can be used for increasing the arrival traffic demand beyond the capacity of the airport as was needed for the manually modified HITL scenario in Ref. 6; the HITL scenario in Ref. 6 achieved demand exceeding capacity by squeezing six-hours of traffic into five hours.
+VI. ConclusionsThe two-step procedure for automated scenario generation for Multi-Aircraft Control System based traffic simulation was described.The first step utilized the scenario generation process currently being used by the Air Traffic Management Testbed in development at NASA Ames Research Center.The second step, which implemented refinements to the scenario output from the first step, for meeting the objectives of the Human-in-the-Loop experiments and fast-time simulations, was also described.Flights were filtered in the second step based on route length, cruise speed, cruise altitude, entry time and the ratio of internal to external flights.The procedure for selecting internal and external flights was described.Finally, first-come first-served schedulers were described for curtailing arrival traffic demand to meet the airport arrival capacity constraints, and to increase arrival traffic demand over the airport capacity to meet Human-in-the-Loop experiment objectives.To determine the most frequently used runway configurations and the ones used during the busiest periods in terms of the number of operations (sum of arrivals and departures), runway configurations used during every hour of every day in 2017 and the associated numbers of arrivals and departure were obtained from the FAA's Aviation System Performance Metrics database.Results of analysis were presented in tables to summarize top five most frequently used configurations at John F. Kennedy, Newark Liberty, LaGuardia and Teterboro airports both individually and together.This analysis led to the identification of 5/23/2017 as a busy traffic day on which to base the scenario generation.Flights with and without flight plans in the 5/23/2017 traffic data were processed using the two-step procedure and the scheduling procedure to generate the results.These results show that the automated procedures discussed in the paper can be used to generate traffic scenarios that meet the requirements of Human-in-the-Loop experiments and fast-time simulations for evaluation of air traffic concepts.The automated process can replace the tedious manual scenario generation process; it is less error prone and makes it possible to generate large number of scenarios needed for Monte-Carlo evaluations, which is very difficult to achieve with the manual process.Fig. 11Fig. 1 Hourly sum of traffic counts at JFK, EWR, LGA and TEB on 5/23/2017 based on ASPM data.
+Fig. 22Fig. 2 Inputs and associated options for MACS scenario generation.
+Fig. 3 .3Fig. 3. Summary of MACS scenario generation steps.
+Fig. 4 .4Fig. 4 Scenario refinement steps.
+Fig. 55Fig. 5 Actual and predicted hourly arrival traffic counts at JFK, EWR, LGA and TEB taken together.
+Fig. 7 Fig. 6 .Fig. 8768Fig. 7 Histogram of the original entry times.
+Table 1 Top five most frequently used configurations at JFK, EWR, LGA and TEB1AirportConfigurationHoursArrivalsDepartures Operations % Hours% OperationsJFK31L, 31R | 31L3,817102,88995,443198,3324344JFK13L | 13R1,02026,72928,29755,0261112JFK13L, 22L | 13R74124,97822,64347,621810JFK22L, 22R | 22R69015,81715,90331,72077JFK4L, 4R | 4L49310,35410,09520,44954EWR 22L | 22R3,58294,37497,409191,7834044EWR 4R | 4L2,91472,91676,997149,9133334EWR 11, 22L | 22R49618,85616,27035,12658EWR 22R | 22R3342,6731,7864,45931EWR 4L | 4L2872,5011,8914,39231LGA22 | 132,04146,92746,22293,1492425LGA31 | 41,51936,90636,39473,3001820LGA4 | 131,38532,57633,40665,9821618LGA22 | 311,21028,71528,70657,4211416LGA31 | 311,17521,09821,29442,3921411TEB19 | 244,05039,00237,75276,7544748TEB6 | 13,00125,57525,66551,2403532TEB19, 24 | 246447,4037,94315,34679TEB1, 6 | 15835,0665,26410,33066TEB24 | 241251,6541,7383,39212
+Table 3 Top five hours with most operations at JFK, EWR, LGA and TEB taken together3ConfigurationArrivals Departures Operations Local Time UTC DateJFK-13L | 13R; EWR-22L | 22R; LGA-22 | 13; TEB-19 | 2413018331318229/13/2017JFK-13L, 22L | 13R; EWR-4R, 11 | 4L; LGA-22 | 13; TEB-6 | 115215430618226/7/2017JFK-31L, 31R | 31L; EWR-22L | 22R; LGA-22 | 13; TEB-19 | 2415814430218225/23/2017JFK-13L, 22L | 13R; EWR-11,22L | 22R; LGA-22 | 13; TEB-191461553011722 11/21/2017| 24JFK-13L, 22L | 13R; EWR-11,22L | 22R; LGA-22 | 13; TEB-1914715129818228/10/2017| 24
+Table 2 Top five most frequently used combinations of JFK, EWR, LGA and TEB configurations2ConfigurationHours Arrivals Departures Operations % Hours % OperationsJFK-31L, 31R | 31L; EWR-4R | 4L; LGA-31 | 4; TEB-6 | 141038,99838,82277,82055JFK-31L, 31R | 31L; EWR-22L| 22R; LGA-22 | 13; TEB-19 |29025,23726,83152,0683324JFK-31L, 31R | 31L; EWR-22L| 22R; LGA-22 | 31; TEB-19 |27927,21826,88854,1063424JFK-31L, 31R | 31L; EWR-4R | 4L; LGA-31 | 4; TEB-1, 6 | 126425,90126,77952,68033JFK-31L, 31R | 31L; EWR-4R | 4L; LGA-4 | 13; TEB-6 | 126424,72725,93350,66033
+Table 4 Number of internal and external flights based on r values r4InternalExternal005310.251325310.52384760.7523831712382383002380t is the original entry time, the new scheduled entry time, sE t , can be determined as:
+		
+		
+
+			
+AcknowledgementsAuthors are grateful to Dr. Min Xue, Dr. Antony Evans, Confesor Santiago, William Chan and Katharine Lee for reviewing the paper and providing feedback.Discussions with Dr. Robert Windhorst are gratefully acknowledged.
+			
+
+			
+
+				
+
+
+	
+		Human-In-the-Loop Evaluation of NextGen Concepts in the Airspace Operations Laboratory
+		
+			ThomasPrevot
+		
+		
+			PaulLee
+		
+		
+			ToddCallantine
+		
+		
+			JoeyMercer
+		
+		
+			JeffreyHomola
+		
+		
+			NancySmith
+		
+		
+			EverettPalmer
+		
+		10.2514/6.2010-7609
+	
+	
+		AIAA Modeling and Simulation Technologies Conference
+		Toronto, Ontario, Canada
+		
+			American Institute of Aeronautics and Astronautics
+			August 2-5, 2010
+		
+	
+	Prevot, T., et. al., "Human-in-the-Loop Evaluation of NextGen Concepts in the Airspace Operations Laboratory," AIAA 2010-7609, AIAA Modeling and Simulation Technologies Conference, Toronto, Ontario, Canada, August 2-5, 2010.
+
+
+
+	
+		Dynamic Arrival Routes: A Trajectory-Based Weather Avoidance System for Merging Arrivals and Metering
+		
+			ChesterGong
+		
+		
+			DaveMcnally
+		
+		10.2514/6.2015-3394
+	
+	
+		15th AIAA Aviation Technology, Integration, and Operations Conference
+		Dallas, Texas
+		
+			American Institute of Aeronautics and Astronautics
+			June 22-26, 2015
+		
+	
+	Gong, C., McNally, D., and Lee, C. H., "Dynamic Arrival Routes: A Trajectory-Based Weather Avoidance System for Merging Arrivals and Metering," AIAA 2015-3394, 15 th AIAA Aviation Technology, Integration, and Operations Conference, Dallas, Texas, June 22-26, 2015.
+
+
+
+	
+		Integrated Demand Management: Coordinating Strategic and Tactical Flow Scheduling Operations
+		
+			NancyMSmith
+		
+		
+			ConnieBrasil
+		
+		
+			PaulULee
+		
+		
+			NathanBuckley
+		
+		
+			ConradGabriel
+		
+		
+			ChristophPMohlenbrink
+		
+		
+			FaisalOmar
+		
+		
+			BonnyParke
+		
+		
+			ConstantineSperidakos
+		
+		
+			Hyo-SangYoo
+		
+		10.2514/6.2016-4221
+	
+	
+		16th AIAA Aviation Technology, Integration, and Operations Conference
+		Washington, DC
+		
+			American Institute of Aeronautics and Astronautics
+			2016
+		
+	
+	Smith, N. M., et. al., "Integrated Demand Management: Coordinating Strategic and Tactical Flow Scheduling Operations," 16 th AIAA Aviation Technology, Integration, and Operations Conference, Washington, DC, 2016.
+
+
+
+	
+		Airspace Technology Demonstration 2 (ATD-2) Technology Description Document
+		
+			AGing
+		
+		Memorandum: NASA/TM-2018-219767
+		
+	
+	
+		21076-1320
+		Hanover, MD
+		
+			NASA Center for AeroSpace Information
+			7115
+		
+	
+	NASA Technical
+	Standard Drive 5301. cited 5/10/2018
+	Ging, A., et. al., "Airspace Technology Demonstration 2 (ATD-2) Technology Description Document," NASA Technical Memorandum: NASA/TM-2018-219767, NASA Center for AeroSpace Information, 7115 Standard Drive 5301, Hanover, MD 21076-1320. URL: https://www.aviationsystemsdivision.arc.nasa.gov/publications/2018/NASA-TM-2018-219767.pdf [cited 5/10/2018].
+
+
+
+	
+		ATM-X: Air Traffic Management -eXploration
+		
+			WNChan
+		
+		
+	
+	
+		Partnership Workshop
+		Moffett Field, California
+		
+			April, 2018
+		
+	
+	cited 5/10/2018
+	Chan, W. N., "ATM-X: Air Traffic Management -eXploration," Partnership Workshop, NASA Ames Research Center, Moffett Field, California, April, 2018. URL: https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20180002413.pdf. [cited 5/10/2018].
+
+
+
+	
+		Automated Scenario Generation for Human-in-the-Loop Simulations
+		
+			GanoBrotoChatterji
+		
+		
+			KeePalopo
+		
+		
+			YunZheng
+		
+		
+			JimmyNguyen
+		
+		10.2514/6.2018-3751
+	
+	
+		2018 Modeling and Simulation Technologies Conference
+		Atlanta, GA
+		
+			American Institute of Aeronautics and Astronautics
+			June 25-29, 2018
+		
+	
+	6 Chatterji, G. B., et. al., "Automated Scenario Generation for Human-in-the-Loop Simulations," AIAA Modeling and Simulation Technologies Conference, Atlanta, GA, June 25-29, 2018.
+
+
+
+	
+		Evaluation of Integrated Demand Management looking into Strategic & Tactical Flow Management
+		
+			CMoehlenbrink
+		
+		
+	
+	
+		Europe Air Traffic Management Research and Development Seminar
+		
+			June 27-30, 2017
+			Seattle, Washington
+		
+	
+	12 th USA/. cited 5/10/2018
+	Moehlenbrink, C., et. al., "Evaluation of Integrated Demand Management looking into Strategic & Tactical Flow Management," 12 th USA/Europe Air Traffic Management Research and Development Seminar, Seattle, Washington, June 27-30, 2017. URL: http://www.atmseminarus.org/seminarContent/seminar12/papers/12th_ATM_RD_Seminar_paper_51.pdf [cited 5/10/2018].
+
+
+
+	
+		Overview of NASA's Air Traffic Management -eXploration (ATM-X) Project
+		
+			WNChan
+		
+	
+	
+		AIAA Aviation Technology, Integration, and Operations Conference
+		Atlanta, GA
+		
+			June 25-29, 2018
+		
+	
+	Chan, W. N., et. al., "Overview of NASA's Air Traffic Management -eXploration (ATM-X) Project," AIAA Aviation Technology, Integration, and Operations Conference, Atlanta, GA, June 25-29, 2018.
+
+
+
+	
+		Required time of arrival as a control mechanism to mitigate uncertainty in arrival traffic demand management
+		
+			Hyo-SangYoo
+		
+		
+			ChristophMohlenbrink
+		
+		
+			ConnieBrasil
+		
+		
+			NathanBuckley
+		
+		
+			AlGlobus
+		
+		
+			NancyMSmith
+		
+		
+			PaulULee
+		
+		10.1109/dasc.2016.7778013
+	
+	
+		2016 IEEE/AIAA 35th Digital Avionics Systems Conference (DASC)
+		Sacramento, California
+		
+			IEEE
+			2016
+		
+	
+	Yoo, H., et. al., "Required Time of Arrival as a Control Mechanism to Mitigate Uncertainty in Arrival Traffic Management," 35 th IEEE Digital Avionics Systems Conference, Sacramento, California, 2016.
+
+
+
+	
+		Integrated Demand Management (IDM) - Minimizing Unanticipated Excessive Departure Delay while Ensuring Fairness from a Traffic Management Initiative
+		
+			Hyo-SangYoo
+		
+		
+			ConnieBrasil
+		
+		
+			NancyMSmith
+		
+		
+			PaulULee
+		
+		
+			ChristophMohlenbrink
+		
+		
+			NathanBuckley
+		
+		
+			AlGlobus
+		
+		
+			GitaHodell
+		
+		10.2514/6.2017-4100
+	
+	
+		17th AIAA Aviation Technology, Integration, and Operations Conference
+		
+			American Institute of Aeronautics and Astronautics
+			June 2017
+			
+		
+	
+	Yoo, H., et al., "Integrated Demand Management: Minimizing Unanticipated Excessive Departure Delay while Ensuring Fairness from a Traffic Management Initiative," 17th AIAA Aviation Technology, Integration, and Operations Conference, 5-9 June 2017.
+
+
+
+	
+		Using an Automated Air Traffic Simulation Capability for a Parametric Study in Traffic Flow Management
+		
+			HeatherArneson
+		
+		
+			AntonyDEvans
+		
+		
+			DeepakKulkarni
+		
+		
+			PaulULee
+		
+		
+			JinhuaLi
+		
+		
+			MeiYWei
+		
+		10.2514/6.2018-3665
+	
+	
+		2018 Aviation Technology, Integration, and Operations Conference
+		Atlanta, GA
+		
+			American Institute of Aeronautics and Astronautics
+			June 24-28, 2018
+		
+	
+	Arneson, H., Evans, A. D., Kulkarni, D., Lee, P., Li, J., Wei, M. Y., "Using an Automated Air Traffic Simulation Capability for a Parametric Study in Traffic Flow Management," 18th AIAA Aviation Technology, Integration, and Operations Conference, Atlanta, GA, June 24-28, 2018.
+
+
+
+	
+		User Manual for the Base of Aircraft Data (BADA) Revision 3.6
+		
+			ExperimentalEurocontrol
+		
+		
+			Centre
+		
+		No. 10/04
+		
+			July, 2004
+			Eurocontrol Experimental Centre Publications Office, B.P
+			15
+			Bretigny-sur-orge, France
+		
+	
+	EEC Note
+	Eurocontrol Experimental Centre, "User Manual for the Base of Aircraft Data (BADA) Revision 3.6," EEC Note No. 10/04, Eurocontrol Experimental Centre Publications Office, B.P. 15, 91222 -Bretigny-sur-orge, France, July, 2004.
+
+
+				
+			
+		
+	

file127.txt

@@ -0,0 +1,276 @@
+
+	
+	
+		
+II. Sensitivity Study MethodThis section describes the sensitivity analysis method used for determining the affect that departure and arrival capacity reduction at one major airport has on the departure and arrival delays at the other 33 major airports in the continental United States.The 34 airports considered in this study are tracked in the Operational Evolution Plan (OEP) of the Federal Aviation Administration (FAA) and are referred to as OEP airports.The sensitivity determination method consists of conducting an ACES simulation with the 34 airport departure and arrival capacities set at their most common settings for establishing the baseline delays.Then, performing a series of ACES runs with airport departure and arrival capacities reduced at each airport one at a time, while maintaining baseline capacity value at the other airports, to determine the change in arrival and departure delays at each of the 34 airports.The Airspace Concept Evaluation System (ACES) was used to do the simulations.ACES is a comprehensive computational model of the national airspace system consisting of air traffic control and traffic flow management models of air route traffic control centers, terminal radar approach controls (TRACON), airports and the air traffic control system command center (ATCSCC). 2It simulates flight trajectories through the enroute-phase of flights, where the enroute-phase for piston-props is 6,000 feet, for turboprops is 8,000 feet and for jet aircraft is 10,000 feet.A queuing model simulates the surface movement and flight through the terminal airspace.Thus, with continuous aircraft dynamics and discrete air traffic control and traffic flow management events, ACES is a hybrid-system.The traffic flow management and air traffic control models in ACES use airport and sector capacity thresholds for delaying flights while they are on the ground and during their enroute phase to ensure that these capacity thresholds are not exceeded.Some of the ACES outputs are arrival and departure counts at airports, traffic counts in sectors and air traffic system performance metrics including arrival, departure, enroute and total delays.Validation studies in Refs. 3 and 4 have shown that ACES generates realistic delays and airport operational metrics similar to those observed in the real-world.Due to these capabilities, ACES was chosen as the system for conducting the airport departure and arrival rate sensitivity study discussed in this paper.
+III. Simulation Inputs and OutputsInput for ACES simulations consists of scenario files containing capacity data (airport arrival and departure capacities, and sector capacities), traffic data (scheduled departure times and flight-plans), and adaptation data (sector/center geometric data).These inputs are described below.Delay metrics, the outputs of ACES, are defined in this section.
+A. Airport CapacitiesTo determine airport departure and arrival capacities, four-months of data spanning the period from March 1, 2006 through June 30, 2006 reported in the FAA's Aviation System Performance Metrics (ASPM) database were collected.This database can be accessed via the web site: http://www.apo.data.faa.gov/.Airport capacity data for a particular airport can be obtained by selecting the Analysis tab and choosing Airport, Weather and Hourly radio buttons on the graphical user interface.Table 1 shows the airport capacity data for Hartsfield-Jackson Atlanta International airport during each hour of March 17, 2006.The first column shows the local hour and the second column lists the landing and takeoff conditions at that hour.Instrument approach condition is indicated by IA and visual approach condition by VA.The airport departure rate (ADR), which is the number of takeoffs per hour, is tabulated in the third column.The airport arrival rate (AAR), which is the number of landings per hour, is listed in the fourth column of the table.Finally, the total capacity of the airport, which is the sum of the ADR and AAR, is given in the last column of the table.In addition to the items in Table 1, the airport capacity data contain the actual number of arrivals and departures during the hour, cloud-ceiling, visibility, temperature, windspeed, wind-angle and arrival and departure runway configurations.
+Baseline CapacitiesThe data of the type listed in Table 1 were analyzed via scripts written in the Matlab language 5 to determine the most frequently used total capacities, along with the associated arrival and departure capacities, for each of the 74 ASPM airports including the 34 OEP airports.Honolulu International airport, which is one of the 35 OEP airports, was excluded from analysis because this study is devoted to airports within the continental United States.After obtaining the most frequently assigned total capacity -mode capacity from the entire dataset, instances with total capacities equal to the selected mode capacity were placed in a subset.Departure and arrival capacities were then selected from this subset based on the minimum of the cost function given in Eq. (1):2 ) ( AAR ADR C C J ! = (1) ADR Cis the airport departure rate and AAR C is the airport arrival rate.Observe that a minimum value of the function is obtained when the ADR is equal to the AAR.Table 2 lists the selected ADR and AAR values corresponding to the mode value of total capacities.The first and the seventh columns list the International Civil Aviation Organization codes for the airports.The second and the eighth columns indicate whether the airport is included in the OEP or not.Mode values of the total capacities are given in columns five and eleven.The frequency of occurrence of the mode value of the total capacity for each of the 74 airports is given as a percentage of the total of 2928 (24 hours !122 days) possible instances in columns six and twelve.The ADR and AAR values listed in this table were used in the ACES simulation for generating the baseline delay values.
+Reduced CapacitiesMatlab scripts were also used to identify instances where total capacities, Total C , were close to 50% of the baseline capacities, Mode C , listed in Table 2.The desire was to identify instances in real data when ADR and AAR were severely reduced.The ADR and AAR values corresponding to 50% capacities were obtained based on the minimum of Eq. (2):2 2 ) 5 . 0 ( ) ( Total Mode AAR ADR C C C C J ! + ! = (2)These ADR and AAR values for the 34 OEP airports are listed in Table 3.This second set consists of the reduced airport departure and arrival capacities that were used in ACES simulations for comparisons against the baseline capacities listed in Table 2.Note that the reduced total capacities are not exactly 50% of the mode capacity; they are as close to 50% as possible based on the actual four-months of airport capacity data that were analyzed.For example, the reduced total capacity of Cincinnati/Northern Kentucky International Airport (KCVG) in Table 3 is 69% of the baseline total capacity of 156 aircraft/hour in Table 2.
+B. Flight-Plans and Adaptation DataFlight-plans for the simulations were derived from the Aircraft Situation Display to Industry (ASDI) data, which is provided via the FAA's Enhanced Traffic Management System (ETMS) 6 , spanning the period from zero Coordinated Universal Time (UTC) on 17 March 2006 to zero UTC on 19 March 2006.These days were selected because 1) they were within the March 1, 2006 to June 30, 2006 time period and 2) they had experienced high traffic-volume, low weather impact and low delays.There were 48,258 departures on the 17 th (a Friday) and 35,394 departures on the 18 th (a Saturday) according to the Centers: Summary of Domestic Operations Report in the FAA's   Air Traffic Operations Network (OPSNET) database. 7Delay data obtained from the OPSNET database for these days are provided in Table 4.The second row of the table lists the total number of aircraft delayed by fifteenminutes or more.The third and the fourth rows show the number of aircraft delayed due to weather and due to traffic-volume.Total delay is given in the fifth row.Average delay given in the sixth row is obtained as the ratio of the total delay to the total number of aircraft delayed by fifteen-minutes or more.It can be verified that these two days are low delay days by comparing the total time delay values in Table 4 with those in
+C. Flight Schedule and ConnectivityThe flight connectivity data, data conditioning steps and delay metrics are described in this section.
+Flight Connectivity DataFlight connectivity data relating the same physical aircraft to two or more flights segments were obtained from the Bureau of Transportation Statistics (BTS) for the two days.Airline flightnumbers were used as tail-numbers for flights not found in the BTS data.The airline flight-numbers, aircraft tailnumbers and the associated flight-plans for all the flights were then included in the Flight Data Set (FDS) file.The subsequent step consists of assigning a departure time to the flights in the FDS file.Scheduled departure times derived from the BTS data are assigned to the flights in the FDS file found in the BTS data.For flights that are not in the BTS data, proposed departure times from flight-plan messages in the ASDI data are assigned as scheduled departure times.In instances when the route of flight is available but the departure time is not, average taxi times reported in the FAA's Aviation System Performance Metrics (ASPM) database are subtracted from the departure message times reported in the ASDI data to estimate the gate departure times.Scheduled departure times are then set to these gate departure times.After assigning a scheduled departure time for every flight, an ACES simulation is run without airport and sector capacity constraints to compute the unconstrained arrival time of each flight at its destination airport.These computed arrival times are then used as scheduled arrival times at the destination airports of the flights.
+Data ConditioningData conditioning steps are needed to compensate for missing and incomplete data.Although the data conditioning steps taken introduce some errors in the simulation, they help keep most flights in the simulation.Errors are due to discrepancies between the airline flight schedule and the simulated flight schedule Although departure schedules are provided as ACES input, arrival schedules for the flights are created during the configuration step of ACES.These computed arrival times need to be earlier than the scheduled departure times of the next segment of the flights.Data in the initial FDS file are therefore processed further to ensure that flight connectivity is preserved and that the arrival and departure schedules linked to the same physical aircraft account for the turn-around-time.Turn-around-time is defined as the time required for unloading the aircraft after arrival at the gate and preparing it for departure.Turn-around-time was assumed to be 40minutes irrespective of the size of the aircraft.The procedure for checking flight connectivity and turnaround-times is summarized in Fig. 1.The process is begun by running an ACES simulation with the initial FDS file and storing the results in the output database.The output database is examined to retrieve flights with a common tail-number.These flights are sorted in time and then a check is performed to determine if the destination airport of the previous flight segment is the same as the origin of the next segment.If the check fails, a new tailnumber is assigned to the subsequent flight segments.For example, consider the four flight segments in Table 5.Since the first segment of the flight ends at Los Angles International (KLAX) and the next segment begins at KLAX, these two segments are proper.The third segment starts at Chicago O'Hare International (KORD) which indicates that the flight connectivity between the second and the third leg is broken.A new tail-number, N12345-1, is assigned in the FDS file to associate this flight with a different aircraft.Tail-numbers of the subsequent segments are also altered.This means that the tail-number of the fourth segment in Table 5 is also altered to N12345-1 because it shares its airport of origin with the airport of destination of segment three.Next, the scheduled arrival and departure times of the flight segments are examined to determine if there is adequate turn-around-time between the segments.If it is determined that the condition described by Eq. ( 3) is not met, the scheduled departure time is altered to meet the condition.The amount of change in the departure time also appears in the scheduled time of arrival of this flight segment at the next airport.Since the unimpeded flight time between a pair of origin-destination airports is a constant, a change in departure schedule alters the arrival schedule by the same amount.Once the schedule of a flight segment is altered, schedules of subsequent flight segments are also altered to ensure that the turn-around-time requirement is met.The process summarized in Fig. 1 was applied to the initial FDS file that contained data for 98,674 flights operating out of 2,669 U. S. and foreign airports that were operated during the 48-hour period from March 17 th to the 18 th .Flight schedules and tail-numbers were altered for 37,638 flights to create the modified FDS file.
+D. Selection of Time Periods for Capacity ReductionSince the system-wide impact is a function of the time of day when ADR or AAR is reduced, peak-demand times were identified for each airport.A three-hour period around the peak demand time was identified as the time for ADR and AAR reduction at each airport.These times are provided in Table 6.The second and seventh columns list the two dates -3/17/2006 and 3/18/2006 associated with start-times and end-times for reduction of the ADR and AAR values.
+E. Delay MetricsThe delay metrics described below are ACES outputs that have been used for the study described in this paper.Scheduled times are employed in the simulation to provide the datum for computation of delays.Delays associated with the departure and arrival, which are defined below, are computed as those in Ref. 3. Scheduled takeoff time, stt t , is defined as:utot sgdt stt t t t + = ,(4)The gate arrival delay, gad t , is obtained as:sgat agat gad t t t ! = . (9)Substituting Eqs. ( 7) and ( 8) in Eq. ( 9) and using the definition in Eq. ( 6), it is seen that) ( ) ( utit atit uft aft dd gad t t t t t t ! + ! + = . (10)Equation (10) shows that the departure delay is accounted as part of the arrival delay.Arrival delay can be reduced by absorbing a part of the departure delay in flight.These metrics were computed with the baseline and reduced airport capacities to study the system-wide impact of capacity reduction at the 34 OEP airports.Results of this study are discussed in the next section.
+IV. ResultsResults obtained via ACES simulations with baseline capacities are described in Subsection A and those obtained using reduced capacities are discussed in Subsection B.
+A. Baseline Capacity ResultsA simulation was conducted with the conditioned FDS file, baseline sector capacities and baseline airport departure and arrival capacities listed in Table 2.Aircraft-counts in each sector resulting from the baseline ACES simulation were retrieved from the output database and added together to compute the total number of aircraft in the continental United States above 10,000 feet altitude at one-minute intervals.This time history of aircraft count was then compared with the time history of the actual number of flights, above 10,000 feet altitude.Actual flights for those days, recorded in the ASDI data, were processed using NASA's Future ATM Concepts Evaluation Tool (FACET). 9The two time histories are shown in Fig. 2. Observe that the ACES simulation starts with all aircraft on the ground, whereas in the actual air traffic system there are always flights that are airborne.Figure 2 shows that the simulated traffic catches up with the actual traffic around four UTC.The general trend of the simulated traffic is similar to the actual traffic for the twenty-four hours between eight UTC on 17 March 2006 and eight UTC on 18 March 2006 (location marked 32 UTC in Fig. 2).Differences between the time histories are both due to issues with the actual flight data and with the simulation.Several issues related to the quality of ASDI data are described in Ref. 10.These issues make it difficult to exactly determine how many flights are in the airspace at a given instant of time.Flight-plan amendments, cancellations and pop-up flights are not included in the simulation.Flights with track information but missing flight-plans in ASDI data are not included in the simulation.Additionally, the trajectory flown by the real aircraft can be different than the one synthesized in the simulation.During the simulation, aircraft were delayed on the ground and in the air to ensure that the airport and sector capacities were not exceeded.Figure 3 shows the baseline ADR value of 96, scheduled takeoff demand and the achieved takeoff rate at the Hartsfield-Jackson Atlanta airport as a function of time.The time along the abscissa is with respect to 17 March 2006, 0:00 UTC.The dashed line shows the baseline ADR value.The scheduled demand is shown with a solid-line marked with crosses (x) and the achieved departure rate, measured as the number of aircraft that departed in one-hour time period, is shown with another solid-line marked with circles (o).Observe that the scheduled demand was greater than the ADR value, whereas the achieved ('actual') departure rate is close to the ADR value.Comparing the scheduled demand and the achieved rate graphs in Fig. 3 between the locations marked as 28 UTC and 32 UTC, it is seen that the excess demand is modulated by shifting the flights to later times.Actual departure rates beyond 44:00 UTC should be ignored because departed flights that did not reach their destination airports prior to termination of the simulation were not counted.The arrival rate was also controlled in ACES to guarantee that the baseline AAR capacities are not exceeded.Figure 4 shows the baseline AAR value of 96, scheduled arrival demand and the achieved arrival rate at the Hartsfield-Jackson Atlanta airport.Observe that the arrival rate constraint was also met by delaying flights, which is reflected in the duration of the achieved arrival rate being close to the AAR value.It should be noted that most of the arrival delays are realized prior to departure at the departure airport and minimally in the airborne phase.In this sense, delays are mostly realized (imposed) at airports of origin both for ADR constraints at airports of origin and AAR constraints at the airports of destination.This is also the way most of the delays occur in the real air traffic system.For example, controlled departure times are issued at airports of origin during a Ground Delay Program at a destination airport.The values of dd t and gad t for each of the 34 OEP airports were obtained from the ACES baseline simulation.Table 7 lists these values along with the number of aircraft that departed from and arrived at each airport and the number of aircraft that landed at each airport during the twenty-four hour period spanning from eight UTC on 17 March 2006 to eight UTC on 18 March 2006 (location marked as 32 UTC in Fig. 2).Columns one and six list the airports.Departure-counts are listed in columns two and seven, and the total departure delays in minutes obtained by summing the departure delays of aircraft delayed by 15-minutes or more are given in columns three and eight.This 15minutes delay metric is commonly used by the FAA for assessing the performance of the air traffic system.Arrival-counts are provided in columns four and nine, and the total arrival delays in minutes obtained as the sum of arrival delays of aircraft delayed by 15-minutes or more are listed in columns five and ten.It should be noted that the delays in Table 7 cannot be compared with the OPSNET delays given in Table 4 because of their definitions.Delays in ACES are compared against schedule, whereas delays in OPSNET are compared with respect to the time when pilot requests permission to depart.In ACES, once a flight incurs departure delay, it can continue to incur departure delays as it arrives and departs from other airports.In the real system, it is possible that departure delay is only accounted once.Delays would not accrue in subsequent flight segments, if the air traffic controller permits the flight to depart soon after departure request is made by the pilot.Departure delay per flight is obtained as the ratio of the total departure delays to the departure-counts and the arrival delay per flight is obtained as the ratio of the total arrival delays to the arrival-counts.These ratios, obtained using the data in Table 7, are shown in Fig. 5.This figure shows that Hartsfield-Jackson Atlanta International (KATL) flights experience the most departure and arrival delays.One of the reasons is apparent from Fig. 3, which shows that the ratio of peak departure demand to departure capacity is 1.7.For comparison, Chicago O'Hare (KORD), which has similar ADR and AAR values as Atlanta, has a peak departure demand to capacity ratio of 1.1.Flights departing from George Bush Intercontinental/Houston Airport (KIAH) and flights arriving at Fort Lauderdale/Hollywood International (KFLL) also experience significant delays.The large difference between   2).
+B. Reduced Capacity ResultsOne-hundred-and-two ACES simulations were conducted with reduced ADR and AAR capacities listed in Table 3 for the time-durations given in Table 6.The baseline ADR and AAR values for the non-OEP airports listed in Table 2 were kept for all the simulations, only the values for OEP airports were altered for the sensitivity study.The first set of 34 ACES simulations were conducted by changing the ADR value for each OEP airport one at a time, while keeping the baseline ADR values for the other airports.The baseline AAR values for all OEP airports were kept for this set of simulations.Figure 6 shows an example of the achieved departure rate in response to ADR reduction at the Hartsfield-Jackson Atlanta airport from the baseline rate of 96 aircraft per hour to 48 aircraft per hour during 13:00 UTC through 16:00 UTC.The dashed-line in the graph shows the ADR value and the solid-line marked with circles shows the achieved rate.The scheduled departure demand is shown by the solid-line marked with crosses.The impact of the ADR reduction at Atlanta airport on the departure and arrival delays at the other airports is shown in Fig. 7.This figure shows the percentage increase or decrease in the delay values compared to the baseline delay values at those airports, which are given in columns three, five, eight and ten of Table 7.A key observation is that as departure delays at Atlanta increase, arrival delays at the other airports also increase.This is an expected result based on Eq. ( 10); it is interesting that the departure delays at some of the airports increase.This effect is explained by the fact that departures from these airports are connected to arrivals from Atlanta.The same physical aircraft arriving from Atlanta is flown out of these airports to other out-bound destinations.An arrival delay associated with these flights shows up as departure delay for the connected out-bound flights.Finally, one observes that the arrival delays also increase at Atlanta although the AAR values were not changed.This again is due to delayed departures from airports that depart aircraft for Atlanta.The impact of ADR reduction at each of the 34 OEP airports on the departure delays at other OEP airports is summarized in Table 8.The first column of this table lists the airport whose ADR was reduced and the header row indicates the impacted airport.A value of 89 in the first element of the first row states that total departure delay of flights delayed by 15-minutes or more at Atlanta increased by 89% compared to the baseline departure delay value of 151,897 minutes (see Table 7) due to reduced ADR at Atlanta.Similarly, the second element of the second row shows that the departure delay increased by 57% at Boston Logan airport due to reduced ADR at the Boston Logan airport.Note that the percentage values in the table have been rounded.Viewing Table 8 as a matrix, it is seen that the diagonal elements have a higher value compared to the off-diagonal terms.This is an expected result because ADR reduction at the airport directly affects departures from that airport.Closer examination reveals that for some of the airports, departure delays do not increase significantly with reduced ADR values.It was determined that for these airports, the departure demand is either lower or only slightly greater than the reduced ADR values.The only two airports -KMEM and KPHX for which the reduced ADR values were found to be same as the baseline values in the four months of operational data (see Tables 2 and3), additional departure delays were not expected.The effect of ADR reduction on the arrival delays is shown in Table 9. Viewing the data in Table 9 as a matrix, it is seen that the values of diagonal elements are low, which indicates that reduced ADR at an airport does not significantly increase arrival delays at that airport.Some airports -Atlanta (KATL), Houston (KIAH), John F. Kennedy (KJFK), San Francisco (KSFO) and Salt Lake City (KSLC) did not follow this trend.Reduced ADR at these airports had the effect of increasing arrival delays at these airports.Examining the rows of Table 9, it is seen that the off-diagonal terms are large for some airports.For example, the value of 41 in the first row under the KDFW heading means that total delays of flights arriving at Dallas/Fort Worth (KDFW) that were delayed by fifteenminutes or more increased by 41% compared to the baseline value in Table 7 due to reduced ADR at Atlanta.Increase in arrival delay should be expected because the delay caused by ADR reduction at the airports of origin can be expected to be propagated to the airports of destination.The significance of increase in delay should be judged by comparing the baseline delay value for the airport against the baseline delay values of other airports, which are given in Table 7.For example, an increase of 50% in delays at Atlanta is considerably more significant compared to the same increase at Salt Lake City.The next set of 34 ACES simulations were conducted with reduced AAR values at each of the 34 OEP airports.Baseline ADR values were kept for all the airports.Figure 8 shows the impact of AAR reduction at Atlanta on the other OEP airports.The bar-graphs show that the total arrival delay due to flights arriving late by fifteen-minutes or more increases by more than 90% compared to the baseline arrival delay in Table 7.The figure shows that departure delays at several airports increase due to reduced AAR at Atlanta.This is to be expected because the arrival constraint at Atlanta is met by delaying the out-bound flights to Atlanta at their airports of origin.Arrival delay at Atlanta also contributes to departure delay at Atlanta due to in-bound out-bound flight connectivity.This departure delay then propagates as arrival delay at other airports.In some instances, the departure and arrival delays are reduced slightly at other airports.This is essentially due to shifting of the departure and arrival times of the affected flights to times of lower demand at these airports.The results shown in both Figs.7 and 8 demonstrate that the impact of capacity reduction at one airport on the delays at another airport is complicated because of network (flight-connectivity) effects.Mathematical modeling of these effects is difficult, and therefore, a simulation capability like ACES is required for such an analysis.The impact of reduction of AAR at each airport on the departure delays at 34 OEP airports is summarized in Table 10.Data trends in this table are similar to those seen in Fig. 8.It should be noted that departure delays at La Guardia (KLGA), Minneapolis-Saint Paul (KMSP), Chicago O'Hare (KORD), San Francisco (KSFO) and Salt Lake City (KSLC) increase significantly due to their own reduced AAR rates.The sensitivity of arrival delays at the 34 OEP airports to reduced AAR at other airports is summarized in Table 11.This table shows that the reducing AAR at the airports, increases arrival delays significantly.Delays increase by more than 100% at Cleveland-Hopkins (KCLE), Charlotte/Douglas (KCLT), Newark Liberty (KEWR), Washington Dulles (KIAD), John F Kennedy (KJFK), La Guardia (KLGA), Minneapolis-Saint Paul (KMSP), Chicago O'Hare (KORD), Philadelphia (KPHL), Phoenix Sky Harbor (KPHX) and Salt Lake City (KSLC).The off-diagonal terms show that arrival delays also increase considerably at some airports due to AAR reduction at other airports.Instances are also seen where arrival delays decrease by a small amount.0 0 0 2 1 KPDX 0 0 0 -1 -1 0 1 -1 -2 2 1 -2 0 1 -3 4 KPHL 0 0 1 1 0 0 0 0 0 0 0 3 0 0 0 1 KPHX 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 KPIT 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 KSAN 0 0 -1 -2 1 2 0 -1 1 1 0 -1 -1 0 1 2 -1 KSEA 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 KSFO 0 0 0 0 0 -1 0 6 2 -1 0 0 0 0 8 11 KSLC 0 0 -2 -2 -1 -1 0 4 2 1 0 -3 -1 0 1 3 0 KSTL 0 -1 -1 0 0 -1 0 0 0 0 0 -1 0 0 0 0 KTPA 0 0 1 0 0 0 0 -10 KDEN 3 -1 1 -1 0 0 1 5 -2 -1 2 4 4 -3 1 3 KDFW 0 0 0 0 0 0 0 1 0 0 0 1 1 1 0 0 KDTW 1 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 KEWR 0 1 0 0 1 1 1 0 0 0 0 0 -1 0 0 0 KFLL -1 1 0 0 0 0 0 0 -1 0 0 0 0 0 0 1 KIAD 0 0 0 0 0 0 0 -3 0 0 0 0 -2 1 0 -1 KIAH 0 0 2 0 1 2 0 -1 0 1 2 2 -2 3 2 3 KJFK 0 1 0 0 0 1 0 -3 2 0 0 0 -1 -1 0 1 KLAS 0 0 1 0 0 0 0 2 0 1 0 15 1 4 1 0 KLAX 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 KLGA 95 2 1 0 2 2 -1 -2 2 0 3 0 -1 1 0 4 KMCO 0 2 -2 0 0 1 0 -410 0 -1 0 1 0 -1 1 0 0 0 0 0 0 -1 0 KCLE 0 0 1 1 0 0 0 0 0 1 0 0 0 0 0 0 KCLT 0 0 1 1 2 1 1 1 1 1 1 0 0 0 1 1 KCVG 0 0 1 1 0 1 1 0 1 1 1 0 0 0 0 1 KDCA 0 0 0 0 0 0 0 0 0 1 0 0 0 0 -1 0 KDEN 0 0 -1 -1 2 1 5 4 2 2 2 -1 1 1 2 3 KDFW 0 1 0 2 0 0 0 3 1 1 1 0 0 0 0 2 KDTW 0 0 0 1 0 1 1 1 0 0 0 0 0 0 1 1 KEWR 0 2 2 3 5 2 3 0 2 4 3 2 2 2 1 2 KFLL 0 0 0 0 1 0 0 1 0 2 1 1 0 3 -2 1 KIAD 0 0 -1 2 1 0 0 0 0 1 0 0 0 1 0 2 KIAH 0 1 280 KSAN 0 0 0 2 2 0 -2 1 -1 1 -1 -1 1 1 2 1 KSEA 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 KSFO 0 1 1 1 0 3 0 9 2 1 0 0 1 5 8 10 KSLC 0 0 0 4 0 1 -1 4 -1 0 0 -2 1 2 4 2 KSTL 0 0 0 0 0 0 -1 1 0 1 1 -1 0 0 1 0 KTPA 0 0 2 0 1 0 1 1 0 1 1 1 0 1 00 0 0 -1 0 0 1 0 0 0 0 0 0 0 1 KCLE 0 0 1 0 0 0 1 0 0 0 1 0 0 0 0 0 KCLT 1 1 0 1 0 2 1 1 1 0 3 0 0 0 2 1 KCVG 0 0 0 0 0 1 0 1 0 0 0 1 0 1 0 0 KDCA 0 0 0 0 0 1 0 0 0 0 1 0 0 0 0 0 KDEN 5 -1 1 0 -1 2 3 0 0 1 1 3 4 3 1 -1 KDFW 0 0 1 1 0 1 2 1 1 1 1 3 2 2 0 0 KDTW 0 0 2 1 0 2 0 0 0 0 1 1 0 -1 0 0 KEWR 0 3 3 3 2 2 3 4 0 1 4 1 0 1 4 1 KFLL 0 1 1 1 0 0 1 0 0 0 2 1 0 1 1 2 KIAD 0 0 0 0 0 4 4 4 2 0 1 3 -1 0 1 0 KIAH 1 0 1 7 1 4 4 -1 2 2 4 5 -1 4 3 2 KJFK 0 3 0 1 2 3 2 4 0 1 1 4 1 2 4 4 KLAS 0 0 5 3 0 8 2 5 1 7 0 21 2 8 4 1 KLAX 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 KLGA 2 2 4 2 4 3 7 1 3 0 8 1 -1 1 -1 3 5 KMCO 0 0 1 1 0 1 1 0 1 0 0 0 0 0 -1 -1 KMDW 0 0 2 0 0 3 0 3 0 0 0 0 0 0 2 0 KMEM 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 KMIA 0 0 0 1 0 1 0 0 1 0 1 0 0 0 0 2 KMSP 1 140 0 0 KSAN 1 -1 -2 0 0 0 0 1 1 2 0 4 0 2 1 -1 KSEA 0 0 0 1 0 1 0 3 0 0 0 0 0 1 0 0 KSFO 0 0 0 0 0 2 3 4 1 3 0 8 17 11 1 0 KSLC 0 -1 1 -3 0 1 -1 5 -1 4 0 7 5 77 2 -1 KSTL -1 -1 0 0 -1 0 2 0 1 0 2 0 0 1 0 -1 KTPA 1 0 3 1 0 0 0 2 0 0 1 1 0 1 1 2The final set of 34 ACES simulations were conducted with both ADR and AAR reduced together at each airport, one at a time, while keeping baseline ADR and AAR values at the other airports to complete the sensitivity study.Tables 12 and13 summarize these results.Table 12 presents the impact on departure delays at the 34 OEP airports and Table 13 shows the impact on arrival delays.Both these tables show that the departure and arrival delays increase substantially at the airports were capacity is reduced.Other observations made in the previous tables remain the same for these tables too.The data presented in Tables 12 and13 show that the percentage change in departure and arrival delays at the affected airports due to both reduced ADR and AAR capacities is close to the maximum of delay change due to reduced ADR capacities or AAR capacities given in Tables 8 through 11.For example, departure delay increase of 93% at Dallas Fort Worth (KDFW) due to both ADR and AAR reduction at Hartsfield-Jackson Atlanta (KATL) (see Table 12) is closer to 85% increase in departure delays due AAR reduction at KATL (see Table 10) compared to 39% increase in departure delays due ADR reduction at KATL (see Table 8).These initial results show that it might not be possible to simply add the impact due to ADR capacity reduction to that due to AAR reduction to derive the combined impact of both ADR and AAR capacity reduction.The utility of the sensitivity data in Tables 8 through 13 for developing delay forecasting models remains to be seen.The results also provide insight into flight demand between pairs of airports.For example, the impact of capacity reduction at Atlanta on delays at Dallas Fort Worth (KDFW) is much more compared to those at San Francisco International (KSFO).This insight can also be gained by analyzing origin-destination pairs in the ACES FDS file.System-wide impact due to each airport is easily determined by first using the percent change in delays given in the rows of Tables 8 through 13 with the baseline delay values reported in Table 7 for determination of delay increase or decrease at each affected airport, and then adding these delay values.Figure 9 shows the increase in system-wide departure delays due to ADR and AAR reduction, obtained using the values in Table 12. Figure 10 depicts the impact on system-wide arrival delays obtained using values in Table 13.Both Figs. 9 and 10, show that capacity constraints at Atlanta, compared to constraints at other airports, have a significantly higher impact on the total system departure and arrival delays.One of the reasons is that there are significantly more flights with connected segments out of Atlanta compared to any other airport.Atlanta had 1,347 connected flights compared to 1,017 at Chicago, the airport with the next higher number of connected flights.In the real system the delays at            Atlanta might be considerably less because of the following reasons.Fifty-percent capacity for three hours with a peak demand capacity ratio of 3.3 (see Fig. 6), which means three times the demand, is extreme.When delays are this severe, flights are cancelled in the real system.Flights were not cancelled during the ACES simulations.0 3 -2 0 2 0 0 1 0 0 0 KBWI 0 1 0 -2 1 -1 0 1 -2 2 1 -2 -2 0 -1 0 KCLE 0 1 1 29 1 2 1 -1 1 2 1 0 2 0 -1 0 KCLT 0 0 -1 -1 28 2 -1 -2 4 2 1 -1 -1 0 0 1 KCVG 0 0 0 0 0 6 0 0 0 1 0 0 0 0 -1 0 KDCA 0 1 -1 1 0 0 6 0 1 1 0 0 0 0 -1 0 KDEN 0 0 1 0 -1 1 2 20 -2 0 1 1 2 0 1 4 KDFW 0 0 0 1 0 0 1 3 38 0 0 0 0 0 0 0 KDTW 0 0 0 1 0 2 1 -1 -1 10 0 -1 1 0 0 0 KEWR 0 11 4 4 7 4 1 -2 0 19 14 1 8 0 0 -1 KFLL 0 -1 -2 -2 2 -1 0 -1 3 3 1 20 -1 0 -1 2 KIAD 0 1 -2 4 2 3 0 0 3 3 6 0 14 0 0 1 KIAH 0 0 -1 1 1 1 3 -2 6 2 -1 -1 1 0 0 1 KJFK 0 12 -180 0 -1 -1 -1 0 0 0 0 0 0 0 0 0 0 KSAN 0 0 1 0 -1 -1 0 -2 0 0 0 0 2 0 2 1 KSEA 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 3 KSFO 0 0 -1 -1 -1 -1 -1 6 -2 -1 0 0 -2 0 -2 31 KSLC 0 -1 -2 0 -1 -3 1 11 5 -1 1 0 -1 0 0 7 KSTL 0 -1 -2 -2 0 -2 2 2 -2 1 0 -2 1 0 -2 -4 KTPA 0 0 -2 -1 2 0 -20 -4 0 -1 0 0 -1 1 1 KMDW 0 0 1 3 0 0 0 1 -1 -1 0 0 0 0 1 0 KMEM 1 0 2 -2 3 -1 1 0 0 -2 0 0 0 -3 3 0 KMIA 0 -1 1 0 0 -1 0 0 -1 0 0 -1 0 0 -1 0 KMSP 0 -1 0 1 0 0 -1 5 -2 6 0 0 0 1 0 1 KORD 0 0 2 1 -1 2 0 -1 2 9 -1 -1 0 -1 -4 -2 KPDX 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 KPHL 0 1 -1 0 1 -1 0 0 -3 0 1 0 0 1 -2 0 KPHX -1 -2 5 5 0 -2 -1 14 0 0 2 0 -1 -2 -3 13 4 KPIT 0 0 0 0 0 0 0 0 -1 0 0 0 0 0 -1 0 KSAN 0 0 0 0 1 0 0 -1 0 0 0 0 0 0 0 2 KSEA 0 0 0 0 0 0 0 0 -1 0 0 0 0 0 0 1 KSFO -1 0 2 -1 0 1 -3 2 1 -2 -3 -1 -1 0 4 17 KSLC 0 0 3 6 -1 0 1 -1 1 -3 -3 -2 -1 -1 -1 2 2 KSTL 0 -1 0 2 0 2 -3 -1 -6 1 -2 -2 0 0 -2 0 KTPA 0 -1 1 -1 -1 1 -2 2 -1 1 0 0 -1 0 0 -10 0 -1 0 -1 0 1 0 4 0 -1 0 -2 -1 0 KBWI 0 -1 -1 -6 -1 -1 0 0 -1 0 0 0 -1 1 1 0 -1 KCLE 0 0 1 0 0 1 -1 0 0 1 -1 -1 -1 0 1 0 0 KCLT -1 -1 0 2 -1 1 -1 0 1 0 9 1 0 0 -2 0 KCVG 0 0 0 0 0 0 0 0 0 0 0 1 0 0 -1 0 KDCA 0 0 0 0 0 0 -1 2 0 0 0 1 -1 0 -1 -1 -1 KDEN 4 -1 -1 4 0 0 2 2 2 0 5 3 -1 3 1 -1 KDFW 0 0 0 0 0 2 0 -2 0 0 0 1 -1 1 0 0 KDTW 0 0 0 -6 -1 1 1 1 -1 0 1 1 -1 -1 1 -1 KEWR 0 -1 0 -1 -1 -1 2 -1 0 0 1 3 -1 -2 1 1 0 KFLL 2 2 0 0 -1 1 -1 -1 2 0 5 1 0 1 1 1 KIAD 0 -1 0 1 1 -1 0 1 -1 0 3 3 -2 2 0 2 KIAH 1 -1 -2 -2 -1 2 -1 -1 -1 1 1 0 -3 2 -1 -2 KJFK -1 0 -1 0 -2 -2 1 -2 -1 1 -2 0 0 0 -2 1 KLAS 0 1 2 -2 1 0 0 1 1 1 3 10 -2 6 0 3 KLAX 0 0 0 0 0 0 0 0 0 1 0 2 0 1 0 0 KLGA 153 -2 9 1 -2 -2 5 -2 0 3 5 2 -2 1 0 0 KMCO 0 1 -1 1 0 0 -2 0 1 0 3 1 0 0 1 0 KMDW 0 1 48 0 0 0 0 2 0 0 -1 1 0 0 1 0 KMEM 2 -2 -1 65 0 -4 1 -1 0 2 4 2 -3 1 0 2 KMIA 0 -1 2 -1 16 0 0 0 1 0 0 1 -1 0 -1 1 KMSP 0 0 4 0 0 405 -2 1 0 0 4 1 -1 -1 -3 3 -1 KORD 0 -1 -1 2 -1 5 275 -2 0 0 3 12 -1 -1 -2 2 -1 KPDX 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 KPHL 0 1 3 -72 KMCO 0 1 0 -1 -1 1 -1 1 -1 1 0 1 -1 0 0 2 KMDW 0 0 1 3 0 0 0 1 0 5 0 0 0 0 0 0 KMEM 0 0 1 2 0 2 1 -1 6 4 -2 0 2 0 -1 2 KMIA 0 -1 -1 0 0 -1 -1 -1 0 0 -1 -2 -1 0 0 -1 KMSP 0 1 0 5 0100 0 -1 -1 -1 0 0 0 0 0 0 0 0 0 0 KSAN 0 0 -1 -2 2 2 1 -3 1 1 0 0 -1 0 0 -2 KSEA 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 3 KSFO 0 -1 -2 -3 -1 0 -1 13 0 -1 0 -1 -2 0 6 34 KSLC -1 -1 -3 -1 -4 -1 0 10 7 0 -1 -1 -2 0 -1 8 KSTL 0 -1 -3 -3 -1 -2 1 2 -2 1 0 -2 -1 0 -2 -2 KTPA 0 1 -1 -12 0 -1 0 -1 1 0 KBWI 1 -2 1 -2 -1 3 -1 2 -2 0 1 -1 2 -1 1 5 KCLE 1 0 6 0 0 0 0 1 -1 0 2 0 0 -1 0 -1 KCLT 2 0 2 0 4 6 5 -1 1 0 2 1 0 -2 -3 2 1 KCVG 0 1 1 0 0 0 1 -3 0 0 0 1 0 0 1 0 KDCA 1 1 0 0 -1 1 2 -3 1 0 1 -1 0 -1 1 3 KDEN 2 -1 1 0 0 3 5 2 -1 3 -5 17 -3 1 1 4 KDFW 0 0 0 1 0 0 4 -4 0 0 1 1 -1 0 -1 3 0 KDTW 1 -1 1 -3 -1 2 1 0 -1 0 1 2 1 -2 1 -1 KEWR 2 4 1 1 5 4 17 -5 1 0 23 -1 0 -3 0 3 KFLL 0 5 2 -1 -3 1 -1 -2 1 0 -1 1 2 -1 -1 1 15 KIAD 3 0 1 2 -2 0 1 0 0 0 5 0 0 0 13 4 KIAH 2 -2 4 2 -1 7 0 -1 -1 1 -2 3 -4 2 6 3 KJFK 2 2 -2 -1 1 4 2 -3 1 -1 9 -1Figure 2 .2Figure 2. ACES simulated aircraft counts and actual aircraft counts comparison.
+Figure 3 .3Figure 3. Departure rate achieved at Hartsfield-Jackson Atlanta airport with baseline airport and sector capacities.
+Figure 4 .4Figure 4. Arrival rate achieved at Hartsfield-Jackson Atlanta airport with baseline airport and sector capacities.
+Figure 5 .5Figure 5. Baseline departure and arrival delays at the 34 OEP airports.
+Figure 6 .6Figure 6.Departure rate achieved at Hartsfield-Jackson Atlanta airport with reduced ADR.
+Figure 9 .9Figure 9. System-wide impact of ADR and AAR reduction on departure delays.
+Table
+Table 2 .2Baseline capacities for the 74 ASPM airports.. Hartsfield-Jackson Atlanta International airportcapacity on March 17, 2006.Local Hour Weather ADR AAR Total0IA96961IA96962IA96963IA96964IA96965IA96966IA96787IA96828VA96949VA969410VA969411VA969412VA969413VA969414VA969415VA969416VA969417VA969418VA969419VA969420VA969421VA969422VA969423VA9694
+Table 8 of8Ref. 8. Along with the flight-plan data, adaptation data and capacity data are required for ACES simulation.Sector and center geometry definitions in the January 1, 2005 adaptation data obtained from ETMS have been used to generate the results in this paper.Baseline sector capacity values are also derived from January 1, 2005 ETMS data tables.
+Table 4 .4OPSNET delay data.Date3/17/2006 3/18/2006 3/19/2006# Aircraft Delayed7831441Weather1661199Volume396129Total Delay (min.)22,05414,21070,119Average Delay (min.)28.1729.8548.66
+Table 3 .3Reduced capacities for the 34 OEP airports.Airport ADR AAR Total Airport ADR AAR TotalKATL484896 KLGA2550KBOS262652 KMCO3672KBWI282856 KMDW2448KCLE282856 KMEM56116KCLT303060 KMIA3262KCVG5157108 KMSP2652KDCA262652 KORD50100KDEN6262124 KPDX3264KDFW5659115 KPHL2654KDTW6048108 KPHX48108KEWR303060 KPIT4080KFLL181836 KSAN2856KIAD303262 KSEA2856KIAH484896 KSFO2752KJFK202040 KSLC4080KLAS303464 KSTL3264KLAX5357110 KTPA1939
+Table 5 .5Flight segments operated by the same physical aircraft.Segment Tail-Origin Destinationnumber1N12345 KSFO KLAX2N12345 KLAX KDEN3N12345 KORD KIAD4N12345 KIADKORD
+Table 6 .6Time periods for reduced ADR and AAR values at the 34 OEP airports.Airport Start-Start-End-End-Airport Start-Start-End-End-datetimedatetimedatetimedatetime(UTC)(UTC)(UTC)(UTC)KATL3/1713:00 03/17 16:00 KLGA03/17 19:00 03/1722:00KBOS3/1723:00 03/182:00 KMCO 03/17 20:00 03/1723:00KBWI3/1720:00 03/17 23:00 KMDW 03/17 23:00 03/182:00KCLE3/1723:00 03/182:00 KMEM 03/17 13:00 03/1716:00KCLT3/1723:00 03/182:00 KMIA03/17 23:00 03/182:00KCVG 3/1723:00 03/182:00 KMSP03/17 23:00 03/182:00KDCA 3/1723:00 03/182:00 KORD03/180:00 03/183:00KDEN 3/1716:00 03/17 19:00 KPDX03/17 14:00 03/1717:00KDFW 3/1723:00 03/182:00 KPHL03/17 23:00 03/182:00KDTW 3/1722:00 03/181:00 KPHX03/17 16:00 03/1719:00KEWR 3/1723:00 03/182:00 KPIT03/17 19:00 03/1722:00KFLL3/1721:00 03/180:00 KSAN03/17 15:00 03/1718:00KIAD3/1720:00 03/17 23:00 KSEA03/181:00 03/184:00KIAH3/1718:00 03/17 21:00 KSFO03/17 18:00 03/1721:00KJFK3/1721:00 03/180:00 KSLC03/17 16:00 03/1719:00KLAS3/1723:00 03/182:00 KSTL03/17 18:00 03/1721:00KLAX 3/183:00 03/186:00 KTPA03/17 21:00 03/180:00where sgdttatt=tagdt+tatot,(5)where agdttdd=tatt!tstt.(6)Scheduled gate arrival time, sgat t , is defined as:tsgat=tstt+tuft+tutit,(7)where stt t is the scheduled takeoff time (wheels-off time), uft t is the unimpeded flight time and utit t is the unimpededtaxi-in time. Actual gate arrival time, agat t , is similarly defined is terms of the actual takeoff time, att t , actual flighttime, aft t , and the actual taxi-in time, atit t , as:tagat=tatt+taft+tatit.t is the scheduled gate departure time and utot t is the unimpeded (assuming it is the only aircraft) taxi-out time.Recollect that the scheduled gate departure time is available in the FDS file and that the unimpeded taxi times for the airports are obtained from the ASPM database.The actual takeoff time, att t , is similarly defined as: t is the actual gate departure time and atot t is the actual taxi-out time.Actual times are not real ones but simulated times.Departure delay is then obtained as:
+Table 7 .7Baseline delay results for 34 OEP airports.Airport Dep.Dep. DelayArr.Arr. DelayAirport Dep.Dep. DelayArr.Arr. Delaycount! 15 min.count! 15 min.count! 15 min.count! 15min.KATL1,761151,897 1,787166,517 KLGA64714,71718,934KBOS63312,13956813,336 KMCO56017,17820,659KBWI41812,5584269,586 KMDW4544,1945,509KCLE4184,1704083,707 KMEM6137,9963,278KCLT73510,1047356,940 KMIA5897,13011,928KCVG7216,6536063,368 KMSP7493,9425,657KDCA42810,8764336,157 KORD1,55811,015 1,55215,074KDEN8753,1748775,233 KPDX3721,6192,506KDFW1,0354,508 1,0016,009 KPHL80221,18210,715KDTW7514,8607275,714 KPHX95328,81217,893KEWR76021,31671314,047 KPIT3573,5613,524KFLL44811,49948720,481 KSAN3372,4743,617KIAD5968,0256107,532 KSEA4992,1963,184KIAH93045,94284312,642 KSFO5928,74619,072KJFK57412,6205087,847 KSLC6992,9343,610KLAS93517,49682616,825 KSTL4202,7984,258KLAX1,0136,1768607,700 KTPA4065,7346,431departure arrival
+Table 8 .8Impact of ADR reduction at one airport on departure delays at other OEP airports.100.0090.0080.00Change in delay (%)30.00 40.00 50.00 60.00 70.0020.0010.000.00KATLKBOSKBWIKCLEKCLTKCVGKDCAKDENKDFWKDTWKEWRKFLLKIADKIAHKJFKKLASKLAXKLGAKMCOKMDWKMEMKMIAKMSPKORDKPDXKPHLKPHXKPITKSANKSEAKSFOKSLCKSTLKTPAAirportdeparture arrival Figure 7. Impact of ADR reduction at Hartsfield-Jackson Atlanta airport on delays at other OEP airports.
+Table 9 .9Impact of ADR reduction at one airport on arrival delays at other OEP airports.AirportKATLKBOSKBWIKCLEKCLTKCVGKDCAKDENKDFWKDTWKEWRKFLLKIADKIAHKJFKKLASKLAXKATL37101734263536274127106135514KBOS02311060022106-11KBWI
+Table 10 .10Impact of AAR reduction at one airport on departure delays at other OEP airports.100.0090.0080.0070.00Change in Delay (%)20.00 30.00 40.00 50.00 60.0010.000.00-10.00-20.00KATLKBOSKBWIKCLEKCLTKCVGKDCAKDENKDFWKDTWKEWRKFLLKIADKIAHKJFKKLASKLAXKLGAKMCOKMDWKMEMKMIAKMSPKORDKPDXKPHLKPHXKPITKSANKSEAKSFOKSLCKSTLKTPAAirportdeparture arrival Figure 8. Impact of AAR reduction at Hartsfield-Jackson Atlanta airport on delays at other OEP airports.
+Table 11 .11Impact of AAR reduction at one airport on arrival delays at other OEP airports.AirportKATLKBOSKBWIKCLEKCLTKCVGKDCAKDENKDFWKDTWKEWRKFLLKIADKIAHKJFKKLASKLAXKATL96-701238528-114277-122-65KBOS0471-1000200-1-103-10KBWI016220-10-310-20-1-10KCLE0011020200-110-10111KCLT001517701-1010-1-1-111KCVG00000530-1-20000010KDCA01000-1480-20-1-1000-1KDEN00121235811-100-1-24KDFW010000-118001000-10KDTW001-1000-1-2510-100-11KEWR022700-10-26115-102-12KFLL0-1103-330-1004310-22KIAD00142220030-10213KIAH0002-1-230-21-1-153020KJFK05183920-3-10-2012411KLAS001300-1-3100000985KLAX00000000000000028KLGA015-21112-334-1030-40-1-242KMCO0011100
+Table 12 .12Impact of ADR and AAR reduction at one airport on departure delays at other OEP airports.AirportKATLKBOSKBWIKCLEKCLTKCVGKDCAKDENKDFWKDTWKEWRKFLLKIADKIAHKJFKKLASKLAXKATL1009155617413527938792432513012KBOS05820104-302011010KBWI01111-100-211-2-1000KCLE011341210120020-11KCLT00006121-2421-1-1001KCVG000002800010010-10KDCA01-1100110110000-10KDEN001-1-13187112000-14KDFW0001001411100000-10KDTW0001021-1-1360-11000KEWR01264652-20195628000KFLL0-1-1-12-100330330001KIAD02-2534003460540-10KIAH01060323114-10130-11KJFK01418271105-121010002KLAS0000001244100007727KLAX00000003000000112KLGA023-11896295846-35601
+Table 13 .13Impact of ADR and AAR reduction at one airport on arrival delays at other OEP airports.2-2-17
+Airport Increase in total system delay (minutes) arrival delay 256,989 minutesFigure 10.System-wide impact of ADR and AAR reduction on arrival delays.800006000040000200000KATLKORDKLGAKPHLKMSPKEWRKLASKJFKKIAHKPHXKSFOKBOSKSLCKIADKCLTKDFWKDENKFLLKTPAKMDWKMIAKMEMKCLEKPDXKCVGKSEAKDCAKDTWKLAXKMCOKSANKPITKBWIKSTL800006000040000200000KATLKORDKLGAKMSPKPHXKLASKPHLKEWRKSFOKJFKKIAHKCLTKFLLKSLCKIADKBOSKDFWKTPAKDENKCLEKSEAKMEMKMDWKDCAKCVGKLAXKDTWKMIAKMCOKPDXKSANKBWIKSTLKPIT
+Airport Increase in total system delay (minutes) departure delay 215,807 minutes
+		
+		
+
+			
+AcknowledgmentsThe authors wish to thank Dr. Robert Windhorst of NASA Ames Research Center for his support of this study.We also thank the Raytheon Team for enhancing the flight-connectivity functionality in the Airspace Concept Evaluation System (ACES), without which this study would not have been possible.Finally, we thank Tom Romer of NASA Ames Research Center for suggesting additional ways of examining the data and critiquing our results.His comments have helped improve the paper.
+			
+
+			
+The results presented in this paper were generated with a single day of air traffic data.The trends seen in the results are expected to hold for days with similar characteristics as those of the day used for computing the results.If the demand patterns change, numerical values will change but the method described in the paper can be used to generate the new sensitivity matrices.The impact of capacity reduction was studied by altering capacities one airport at a time.On the typical day multiple airports are impacted.The impact of combinations of ADR and AAR capacity reductions at multiple airports has not been studied.
+V. SummaryThis paper described a method for sensitivity study in which the airport departure rate (ADR) and airport arrival rate (AAR) were reduced at each of the 34 major airports in the United States, one at a time, and the impact on the departure and arrival delays at these airports was assessed.To compute these delay values, the Airspace Concept Evaluation System (ACES) was used.One-hundred-and-three ACES simulations were conducted to complete the study.In the first set of 34 runs, only the ADR values were altered.The AAR values were kept at their baseline level.In the second set of 34 runs, the AAR values were changed.The ADR values were kept at their baseline level.Both the ADR and AAR values were reduced for the final set of 34 simulations.The results obtained show that ADR reduction at an airport directly increases the departure delay at that airport.This departure delay then appears as arrival delay at the other airports.It was observed that the departure delays at other airports increase indirectly due to flight-connectivity effects.Reduction of AAR was seen to increase the arrival delay at the affected airport.Passing back of this arrival delay causes the departure delay to increase at the airports sending flights to this affected airport.Flight-connectivity was responsible for causing departure delays at the affected airport.Data tables in the paper provide numerical values that quantify the degree of impact of capacity reduction at one major airport on another.			
+			
+
+				
+
+
+	
+		Airport Capacity and NAS-Wide Delay Benefits Assessment of Near Term Operational Concepts
+		
+			MonicaAlcabin
+		
+		
+			RobertSchwab
+		
+		
+			MichaelCoats
+		
+		
+			MatthewBerge
+		
+		
+			LauraKang
+		
+		10.2514/6.2006-7720
+		AIAA-2006-7720
+	
+	
+		6th AIAA Aviation Technology, Integration and Operations Conference (ATIO)
+		Wichita, KS
+		
+			American Institute of Aeronautics and Astronautics
+			September 25-27, 2006
+		
+	
+	Alcabin, M. S., et al, "Airport Capacity and NAS-Wide Delay Benefits Assessment of Near-Term Operational Concepts," AIAA-2006-7720, Proceedings of AIAA Aviation Technology, Integration and Operations Conference (ATIO), Wichita, KS, September 25-27, 2006.
+
+
+
+	
+		Build 4 of the Airspace Concept Evaluation System
+		
+			LMeyn
+		
+	
+	
+		Proceedings of AIAA Modeling and Simulation Technologies Conference and Exhibit
+		AIAA Modeling and Simulation Technologies Conference and ExhibitKeystone, Colorado
+		
+			August 21-24, 2006
+		
+	
+	AIAA-2006-6110
+	Meyn, L., et al, "Build 4 of the Airspace Concept Evaluation System," AIAA-2006-6110, Proceedings of AIAA Modeling and Simulation Technologies Conference and Exhibit, Keystone, Colorado, August 21-24, 2006.
+
+
+
+	
+		Validating The Airspace Concept Evaluation System Using Real World Data
+		
+			ShannonJZelinski
+		
+		10.2514/6.2005-6491
+	
+	
+		AIAA Modeling and Simulation Technologies Conference and Exhibit
+		San Francisco, CA
+		
+			American Institute of Aeronautics and Astronautics
+			August 15-18, 2005
+		
+	
+	Zelinski, S. J., "Validating The Airspace Concept Evaluation System Using Real World Data," AIAA 2005-6491, Proceedings of AIAA Modeling and Simulation Technologies Conference and Exhibit, San Francisco, CA, August 15-18, 2005.
+
+
+
+	
+		Validating the Airspace Concept Evaluation System for Different Weather Days
+		
+			ShannonZelinski
+		
+		
+			LarryMeyn
+		
+		10.2514/6.2006-6115
+		
+	
+	
+		AIAA Modeling and Simulation Technologies Conference and Exhibit
+		Keystone, CO
+		
+			American Institute of Aeronautics and Astronautics
+			August 21-24, 2006. 1 September 2007
+		
+	
+	AIAA 2006-6115
+	Zelinski, S. J., and Meyn, L., "Validating The Airspace Concept Evaluation System For Different Weather Days," AIAA 2006-6115, Proceedings of AIAA Modeling and Simulation Technologies Conference and Exhibit, Keystone, CO, August 21-24, 2006. 5 URL: http://www.mathworks.com/products/matlab/[cited 1 September 2007].
+
+
+
+	
+		U.S. International Air Travel Statistics. Volpe National Transportation Systems Center. Center for Transportation Information, DTS-44, Kendall Square, Cambridge, MA 02142
+		10.1177/004728759303100425
+	
+	
+		Journal of Travel Research
+		Journal of Travel Research
+		0047-2875
+		1552-6763
+		
+			31
+			4
+			
+			02142, July, 2002
+			SAGE Publications
+			Cambridge, MA
+		
+		
+			Volpe National Transportation Systems Center, U. S. Department of Transportation, Kendall Square
+		
+	
+	Volpe National Transportation Systems Center
+	6 Volpe National Transportation Systems Center, "Enhanced Traffic management System (ETMS) Functional Description," Version 7.4, Volpe National Transportation Systems Center, U. S. Department of Transportation, Kendall Square, Cambridge, MA 02142, July, 2002.
+
+
+
+	
+		Nationwide Personal Transportation Survey, 1995: [United States]
+		10.3886/icpsr03595.v1
+	
+	
+		U. S. Department of Transportation
+		
+			October 1, 2004
+			Inter-university Consortium for Political and Social Research (ICPSR)
+		
+	
+	Federal Aviation Administration
+	Federal Aviation Administration, "Order 7210.55C: Operational Data Reporting Requirements," U. S. Department of Transportation, October 1, 2004.
+
+
+
+	
+		Characterization of Days Based on Analysis of National Airspace System Performance Metrics
+		
+			GanoChatterji
+		
+		
+			BassamMusaffar
+		
+		10.2514/6.2007-6449
+	
+	
+		AIAA Guidance, Navigation and Control Conference and Exhibit
+		Hilton Head, SC
+		
+			American Institute of Aeronautics and Astronautics
+			August 20-23, 2007
+		
+	
+	Chatterji, G. B., and Musaffar, B., "Characterization of Days Based on Analysis of National Airspace System Performance Metrics," Proceedings of AIAA Guidance, Navigation, and Control Conference, Hilton Head, SC, August 20-23, 2007.
+
+
+
+	
+		FACET: Future ATM Concepts Evaluation Tool
+		
+			KarlDBilimoria
+		
+		
+			BanavarSridhar
+		
+		
+			ShonRGrabbe
+		
+		
+			GanoBChatterji
+		
+		
+			KapilSSheth
+		
+		10.2514/atcq.9.1.1
+	
+	
+		Air Traffic Control Quarterly
+		Air Traffic Control Quarterly
+		1064-3818
+		2472-5757
+		
+			9
+			1
+			
+			2001
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	Bilimoria, K. D., Sridhar, B., Chatterji, G. B., Sheth, K. S., and Grabbe, S. R., "FACET: Future ATM Concepts Evaluation Tool," Air Traffic Control Quarterly, Vol. 9, No. 1, 2001, pp. 1-20.
+
+
+
+	
+		Analysis of ETMS Data Quality for Traffic Flow Management Decisions
+		
+			GanoChatterji
+		
+		
+			BanavarSridhar
+		
+		
+			DouglasKim
+		
+		10.2514/6.2003-5626
+		AIAA- 2003-5626
+	
+	
+		AIAA Guidance, Navigation, and Control Conference and Exhibit
+		Austin, TX
+		
+			American Institute of Aeronautics and Astronautics
+			August 11-14, 2003
+		
+	
+	Chatterji, G. B., Sridhar, S., Kim, D., "Analysis of ETMS Data Quality for Traffic Flow Management Decisions," AIAA- 2003-5626, Proceedings of AIAA Guidance, Navigation, and Control Conference, Austin, TX, August 11-14, 2003.
+
+
+				
+			
+		
+	

file128.txt

@@ -0,0 +1,320 @@
+
+	
+	
+		
+I. Introductionhis paper is motivated by the need for predicting wheels-off time at airports where advanced automation systems, such as the Surface Decision Support System (SDSS), that depend on surveillance information derived from Airport Surface Detection Equipment Model-X (ASDE-X) type systems will not be available.At airports with SDSS type systems, aircraft surface movement will be scheduled, which will reduce runway entry time and wheelsoff time uncertainty.Additionally, it will be possible to generate better estimates by using surveillance information.For example, the taxiway entry time of an aircraft waiting at the spot can be predicted more accurately by predicting the trajectories of aircraft in the movement areas starting from the current locations from surveillance data.At less equipped airports, wheels-off time will have to be estimated under the current day conditions with uncertainties associated with aircraft movement in the ramp-area and in the movement-area.Wheels-off time predictions are needed for coordinating departure release times with downstream facilities to meet flow management restrictions imposed by them.These restrictions are imposed for ensuring adequate separation between aircraft, creating orderly flows in terminal areas, and protecting sectors and airports from being overwhelmed by demand.Two of the commonly used restrictions that could benefit from wheels-off time estimation are Call For Release (CFR) and Expect Departure Clearance Time (EDCT).Accuracy requirement for CFR is actual wheels-off time within twominutes early to one-minute late window with respect to the estimated wheels-off time.The requirement for EDCT is plus-minus five minutes.In the first part of the paper, the initial set of 77 major U. S. airports in the Aviation System Performance Metrics (ASPM) database is reduced to the set of 29 airports after eliminating ASDE-X, small-hub and non-hub airports.This reduced set contains 27 medium-hub airports and two large-hub airports.Nine of these airports are being considered for surface surveillance systems.Currently, single radar-based Low Cost Ground Surveillance (LCGS) is
+II. Selection of Non-ASDE-X AirportsIn this section, the sources of taxi-out delay, taxi-out time, number of commercial operations, traffic flow management initiative caused delay counts, passenger enplanement counts, type of hub airport and type of surface surveillance equipment at the airport used for determining the non-ASDE-X airports in this study are described.Most of the data were derived from the Federal Aviation Administration's (FAA) Aviation System Performance Metrics (ASPM) and Operations Network (OPSNET) databases containing historical traffic counts and delay statistics.Data were also derived from the Terminal Area Forecast (TAF) Summary for fiscal years 2011-2040 report.The 77 airports included in the ASPM and OPSNET databases are listed in Table A-1 in the Appendix.These airports are referred to by their airport code in the rest of the document.The first report obtained from the ASPM database is the "Analysis By Airport Report (compared to flight plan)" for calendar year 2011.This report includes average taxi-out time and average taxi-out delay for each of the 77 major U. S. airports.Taxi-out time is the difference between the actual wheels-off time and the actual gate-out time.Taxi-out delay is difference between the taxi-out time and the unimpeded taxi-out time.Taxi-out times and delays are in minutes.Only data from itinerant flights to or from the ASPM 77 airports or operated by one of the ASPM 29 carriers are included.Itinerant flights are those that land at an airport, arriving from outside the airport area, or depart an airport and leave the airport area.Data on flights by ASPM carriers to international and domestic non-ASPM airports are also kept in the ASPM database.General aviation and military flights are excluded.Table A-2 in the Appendix lists the average taxi-out delay and the taxi-out delay ratio derived from this report.The average taxiout delay ratio is defined as the ratio of average taxi-out delay to average taxi-out time.The second report, "OPSNET: Airport Operations: Standard Report" for calendar year 2011 is obtained from the OPSNET database.This same report can also be obtained from the FAA's Air Traffic Activity System (ATADS) database.This report contains air-carrier, air-taxi, general aviation and military itinerant operations, and civil and military local operations.Local operations are performed by aircraft that remain in the local traffic pattern, execute simulated instrument approaches or low passes at the airport, and fly to or from the airport and a designated practice area within a 20-mile radius of the tower.Airport operations include all arrivals and departures at the airport; overflights are not included.OPSNET defines air-carrier as aircraft with a seating capacity of more than 60 seats or a maximum payload capacity of more than 18,000 pounds carrying passengers or cargo for hire or compensation and air-taxi as aircraft with a seating capacity of 60 or less or a maximum payload capacity of 18,000 pounds carrying passengers or cargo for hire or compensation.Air-carrier includes U. S. and foreign flagged carriers.Table A-2 in the Appendix lists the sum of itinerant air-carrier and air-taxi operations (commercial operations as defined by TAF report) for each of the 77 major U. S. airports based on the 2011 report.The third report, "OPSNET: Delays: Delay Types Report" was obtained from OPSNET database to get TMIfrom delay counts for calendar year 2011.These counts are also listed in Table A-2 in the Appendix.OPSNET defines TMI-from delays as traffic management initiative delays from a national or local traffic management initiative as experienced by aircraft departing from the selected facilities.These initiatives include departure spacing (DSP), enroute spacing (ESP), arrival spacing (ASP), miles-in-trail (MIT), minutes-in-trail (MINIT), Expect Departure Clearance Time (EDCT), Ground Stop (GS), and Severe Weather Avoidance Plan (SWAP).It should be understood that these delays are charged to facilities asking for the delays and not to the ones providing the delays.For example, 11,132 aircraft were delayed at Atlanta (see ATL TMI-from delay counts in Table A-2 in the Appendix) to comply with traffic management initiatives of other facilities.From this study's perspective, TMIfrom delay counts (number of aircraft experiencing TMI-from delays) indicates instances when improved wheels-off time estimate at the departure airport could help meet metering and spacing constraints imposed by others.The fourth report derived from OPSNET database, "OPSNET: Delays: EDCT/GS/TMI By Cause Report" provided the TMI-to weather delay counts and TMI-to volume delay counts.TMI-to delays is defined as delays resulting from a national or local traffic management initiative reported in OPSNET and charged to the facility that is the originating cause of the restriction.These initiatives also include DSP, ESP, ASP, MIT, MINIT, EDCT, GS, and SWAP.These delays may be experienced by aircraft at another facility, but are charged to the causal facility.For example, 5,561 aircraft were delayed at other airports due to weather caused restrictions imposed by Atlanta as listed in Table A-2 in the Appendix.TMI-to volume delays are caused by restrictions imposed for moderating the traffic demand.In addition to weather and volume delay counts and minutes, the report contains delays due to equipment, runway and other causes.Passenger enplanements data listed in Table A-2 in the Appendix were obtained from the main FAA website in the "Passenger Boarding and All-Cargo Data" Section (see Ref. 4).The main source of enplanement statistics is the U. S. Department of Transportation (DOT).Scheduled and nonscheduled U. S. certificated aircarriers, commuter air-carriers, and small certificated aircarriers submit data to DOT on Form 41 Schedule T-100.Foreign flag air carriers submit data to DOT on Form 41 Schedule T-100(f).In addition, an annual survey of airtaxi/commercial operators, who report their nonscheduled activity on FAA Form 1800-31, is conducted by the FAA.As one would expect, passenger enplanement is highly correlated to the number of commercial operations.The value of the correlation coefficient (also known as Pearson's correlation coefficient and Pearson's productmoment correlation coefficient) was found to be 96.7% and the p-value was found to be zero to six decimal places, which indicates highly significant correlation.Figure 1 shows the plot of passenger enplanements as a function of commercial operations based on the data in Table A-2 in the Appendix.The dotted line indicates a linear fit between enplanements and commercial operations.The coefficient of determination (R 2 value) was found to be 0.935.A closer look at enplanement data and commercial operations data in Table A-2 in the Appendix indicates some data inconsistencies.For example, 3,634 air-carrier and air-taxi operations were conducted at Oxnard, CA (OXR), but the number of enplanements is just 3. Similarly, numbers look very low for GYY, TEB, and VNY.The following response was received from the FAA when asked for an explanation for the discrepancy.The enplanement numbers posted by the FAA include all revenue passengers that boarded a flight conducted by a large certificated, commuter, or foreign air-carrier.They also include the enplanements for on demand air-taxi operators.However, these small operators are not required to report their passenger enplanements to the FAA.So while FAA has some data, it can vary from year to year based on whether the operators voluntarily report their passenger activity.This impacts the non-primary airports like TEB and VNY that serve corporate and business flights which may have a significant number of Part 135 on demand operations.FAA facility pay level data listed in Table A-2 in the Appendix were obtained from the "OPSNET: Facility Information: Detail Report" derived from the OPSNET database.Reference 5 provides a complete description of the formula for pay setting.The facility pay level varies between "null" and 12. Null level is indicated by a zero in the table.Since facility pay level is based on a formula that considers both the number of operations and complexity factors such as, tower with or without radar, performance characteristics of aircraft using the airport, runway and taxiway layout, proximity to other airports, military operations and terrain, it can be used as a metric for categorizing airports.Correlation between the number of operations and the levels given in Table A-2 was found to be 68.6%.The correlation improved to 78% with 0 levels excluded.In both these instances, the correlations were determined to be significant with p-value of zero up to six decimal places.It is thus seen that number of operations dominates the formula for pay setting.ASPM and OPSNET databases provide operational statistics on Core airports, Operational Evolution Partnership (OEP 35) airports, 45 airports tracked in OPSNET (OPSNET 45) and the 77 airports tracked in ASPM.Core airports are the 30 busiest commercial U. S. airports that serve as hubs for airline operations at major metropolitan areas.OEP 35 airports are commercial U.S. airports with significant activity.These airports serve major metropolitan areas and also serve as hubs for airline operations.More than 70 percent of passengers move through these airports.The Venn-diagram in Fig. 2 shows the airport codes of the Core, OEP 35, OPSNET 45 and ASPM 77 airports.The rectangular box shows all the 77 ASPM airports.The 30 Core airports are enclosed in the innermost circle with thick solid boundary.All the Core airports and five additional airports enclosed in the circle with dotted line boundary form the OEP 35 airport set.OPSNET 45 airports are enclosed in the largest circle with a thin solid boundary.Observe that OPSNET 45 set contains all the Core and OEP 35 airports except Honolulu International airport (HNL).The 31 airports outside the circles only belong to ASPM 77.The airports in Fig. 2 were identified as large-hub airports, medium-hub airports, small-hub airports and non-hub airports based on their designation in Ref. 6.A large-hub airport is defined as an airport with 1% or more of total U. S. passenger enplanements.A medium-hub airport is defined as an airport with 0.25% to 0.99% of total U. S. passenger enplanements.An airport with 0.05% to 0.249% of total U. S. passenger enplanements is categorized as a small-hub airport.Finally, an airport with less than 0.05% of total U. S. passenger enplanements is termed a non-hub airport.All Core airports other than Memphis International airport (MEM) are large-hub airports.These 29 airport codes inside the smallest circle are shown in red without a superscript.The 34 medium-hub and 6 non-hub airports are indicated in black and magenta, and by "m", and "*" superscripts, respectively.The airport codes of 8 small-hub airports are in blue and underlined.Figure 3 shows the airport codes of ASDE-X airports, Low Cost Ground Surveillance (LCGS) airports and Airport Surface Surveillance Capability (ASSC) airports.List of names of airports with ASDE-X, LCGS and ASSC were obtained from Refs.7-9.LCGS is based on single surface movement radar concept.FAA is currently evaluating LCGS at Spokane (GEG), Manchester (MHT), San Jose (SJC), Reno (RNO) and Long Beach (LGB).Airport codes of LCGS airports are indicated in blue with a "*" superscript.As opposed to ASDE-X that uses radar, multilateration and Automatic Dependent Surveillance-Broadcast (ADS-B), ASSC derives data from just multilateration and ADS-B.FAA expects ASSC to begin tracking transponder-equipped aircraft and ADS-B equipped ground vehicles by 2017 at Portland (PDX), Anchorage (ANC), Kansas City (MCI), New Orleans (MSY), Pittsburgh (PIT), San Francisco (SFO), Cincinnati (CVG), Cleveland (CLE) and Andrews Air Force Base.ASSC airport codes in green are underlined.Airports without a surface Figure 3. Airports with ASDE-X, LCGS and ASSC surface surveillance systems, and airports without surface surveillance systems.surveillance system are indicated with airport codes in black with an "n" superscript.The remaining 35 airports in red are ASDE-X airports.After removing all the ASDE-X airports, small-hub and non-hub airports from Fig. 3, the remaining 29 airports shown in Fig. 4 are considered as suitable candidates for further analysis.San Francisco and Tampa are the only two large-hub airports remaining in this set.The other 27 are medium-hub airports.Nine airports in this set will either have LCGS or ASSC systems for surface surveillance.Since surveillance data can be used for comparing estimates against reality, these airports could be considered for near term development and testing of wheels-off time estimation methods.To identify airports that could benefit from wheelsoff time estimation, the important metrics are, 1) taxi-out delay, 2) TMI-from delay counts and 3) number of commercial operations.Without a suitable means for accounting for taxiway delays due to interactions between arriving and departing aircraft, one could resort to estimating wheels-off time based on the single value of taxi-time for the airport from ASPM database.This however, would lead to larger wheels-off time prediction errors at airports with larger taxi-out delays.Thus, airports with larger taxi-out delays can be expected to benefit more by being able to reduce larger wheels-off time prediction errors by employing a wheels-off time estimation method.Airports with TMI-from delays have to ensure that the affected aircraft depart at times coordinated with downstream facilities so that the restrictions imposed by them are met.Airports with higher TMI-from delay counts have to depart more aircraft on time; therefore, wheels-off time estimate can be expected to have a greater impact at these airports.Finally, improved predictability of taxi-time and wheelsoff time has the potential of improving planning and scheduling for greater surface movement efficiency at busier airports, the ones with large number of commercial operations.In addition to estimating taxi-out time, gate departure time needs to be known or estimated for wheels-off time prediction since wheels-off time is obtained by adding the taxi-out time to the gate departure time.Unfortunately, gate departure time is difficult to estimate.If airlines are unable to provide gate departure times prior to actual departure, the only choices are scheduled departure time from the Official Airline Guide (OAG) or proposed departure time from filed flight-plans.These times, however, are not accurate.It is also difficult to estimate gate departure delay by observing airport state data.The study in Ref. 3 found the correlation between gate departure delay and metrics derived from airport state data such as, number of aircraft on the surface, airport departure rate, wind and visibility to be quite low.Like Ref. 3, this study also assumes that gate departure time is known.To group the 29 airports shown in Fig. 4 based on the values of taxi-out delay, TMI-from delay counts and number of commercial operations, and the other metrics listed in Table A-2 in the Appendix, the multiple-metric K-Means classifier described in Ref. 10 is used.The K-Means method partitions data into specified number of groups such that the means associated with the groups are as widely separated as possible.Data elements are then labeled based on their closeness to the group means for reducing the variance.Group means are then re-computed based on the elements assigned to the groups.The process of assignment of elements to the groups and computation of group means is continued till convergence is achieved, that is, group means do not change with successive iterations.The  1.The minimum, mean, standard deviation and maximum values for each group are listed in the last four columns.Airport codes in each group are arranged in the non-increasing order.For example, ONT has the maximum taxi-out delay in Group 1 and OGG has the minimum taxi-out delay in Group 1.The three groups can be considered to be low, medium and high taxi-out delay groups.Groups obtained with TMI-from delay counts and with number of commercial operations are summarized in Tables 2 and3.Observe from Table 2 that Raleigh-Durham (RDU) has to comply with more departure restrictions compared to the other airports.San Francisco (SFO), which has the most taxi-out delays and number of commercial operations, is a member of Group 2 in Table 2 with TMI-from delay counts of 1,848 in the year 2011 (see Table A-2 in the Appendix).Grouping based on taxi-out delay ratio and FAA level are summarized in Tables 4 and5.Comparing Table 1 to Table 4, it is seen that ten airports (ONT, MSY, BUR, SJC, OAK, DAL RDU, SAT, TUS and ABQ) in Table 1 move one level higher in Table 4.These airports have higher average taxi-out delays compared to their nominal taxi-out times.Comparing Table 3   two levels higher, and only two airports-SJU and SFO moved one level down.This reconfirms the finding that the number of operations is significantly correlated to the FAA pay level.Thus, FAA pay levels can be used in lieu of number of commercial operations.Finally, grouping based on TMI-to weather delay counts and TMI-to volume delay counts are given in Tables 6 and7, respectively.Only airports that delayed at least 10 aircraft in the year 2011 were considered for grouping in Tables 6 and7.This reduced the set of 29 airports to just 8 based on TMI-to weather delay counts.San Francisco, the sole member of Group 3 in Table 6, is known to be severely affected by visibility.In 2011, it was responsible for causing over 18,000 aircraft bound for SFO to be delayed elsewhere.Only three airports-CLE, SFO and DAL in Table 7 caused aircraft bound for those airports to be delayed elsewhere due to traffic volume.It is reasonable to expect that departures would be affected when arrivals are impacted by weather and traffic volume.Wheels-off time estimates at these airports would have to consider weather and traffic volume conditions.The grouping results discussed above in Tables 1 through 7 considered a single metric for classification.Groups can also be formed by first creating a composite ID for each airport based on single metric classifications and then placing all the airports with the same ID in a group.For example, CVG is a member of Group 1 based on taxi-out delay, Group 2 based on TMI-from delays and Group 2 based on number of commercial operations, therefore its composite ID is (1,2,2).Similarly, the composite ID of IND is (1,2,2).Thus, CVG and IND belong to the same group based on their composite ID.This method is described in Ref. 10. Table 8 lists the airport grouping with composite ID constructed based on Tables 1, 2 and3.Following this procedure, a member of group with Group ID (3,3,3) would be expected to benefit the most from wheels-off time estimation.Mean values of taxi-out delay, TMI-from delay counts and number of commercial operations for each group of airports are listed in the columns with headings-Mean 1, Mean 2 and Mean 3, respectively.Airport codes of airports that will receive ASSC for surface surveillance are shown in green boldface type and underlined.Airport code of San Jose, where LCGS is being tested, is shown in blue boldface type with "*" superscript.Table 8 shows that other than groups with IDs (2, 1, 1), (2,2,1) and (2,3,2), there is at least one airport in the group that will have a surveillance system.These airports should be initially targeted for developing and testing wheels-off time estimation methods.Track data (position as a function time) from surveillance can be used to determine gate, gate departure time, spot crossing time, ramp-area path and taxiway path needed for developing wheels-off time estimation methods.The estimated wheels-off time can be compared with the actual wheels-off time also using actual track data.The nine groups in Table 8 can be reduced further by merging smaller adjacent groups.Table 9 presents such a grouping by first giving preference to number of commercial operations and then to TMIfrom delay counts using the Group IDs.The first group consisting of Group IDS (1, 1, 1), (2, 1, 1) and (2, 2, 1) has MSY and SJC, two airports that will have surface surveillance systems.The second group of airports consisting of Group IDs (1, 1, 2), (2,1,2), (1,2,2), (2,2,2) and (2,3,2) has six airports-ANC, PDX, CVG, CLE, MCI and PIT that will receive ASSC.One of these airports can be chosen to represent the second set.The third group consisting of Group ID (3,2,3) has SFO as its sole member, an airport that will have ASSC.Thus, SJC with LCGS, CLE with ASSC and SFO with ASSC are good choices for representing the three groups in Table 9.
+III. Wheels-off Time Estimation MethodIn this section, the procedure for estimating wheel-off time at the 29 non-ASDE-X airports in Table 9 is described.It is assumed that airport geometry is available.Taxiways and runways are usually represented by polygons, where the locations of the vertices of polygons are specified by Cartesian coordinates with respect to a frame of reference.Locations of gates and spots are also specified with respect to the same frame of reference.Gate and spot locations and the polygons can be processed to create the node-link graph of the airport.An example of a node-link graph is shown in Fig. 5. Reference 11 describes the procedure for creating the node-link graph using taxiway and runway polygons.Any physical path for going from one location to another on the airport surface is represented by a sequence of nodes and links in this node-link graph.A location on the node-link graph is thus equivalent to a location on the physical airport surface.Since polygons, which are area elements, are represented by links, which are line elements, traversing along the links can be thought of as traversing along the taxiway centerline.Given this node-link graph representation, the first step of the proposed wheels-off time prediction consists of representing the taxi clearance issued by the controller as a path in the node-link graph.Taxiway clearance is specified as an ordered list of taxiway segments that the aircraft is required to follow after pushback from the gate to runway.Mapping from taxiway segments to polygons and from polygons to links is used to determine the path in the node-link graph.In this early phase of development, aircraft position data acquired by ASDE-X during surface movement at Dallas-Fort Worth airport are derived from recorded SDSS logs to identify path from gate to runway in the node-link graph.Polygon containing the aircraft position is identified and then polygon to link mapping is used to identify the corresponding link.The sequence of links then determines the path.The next step of the wheels-off time prediction consists of integrating the aircraft equations of motion along the path in the node-link graph.Starting with the gate location and the gate pushback time, this process generates a time history of positions along the path.This is the classical procedure of open-loop trajectory prediction.If this was the only aircraft moving on the surface, open-loop prediction would be reasonable.In reality however, aircraft moving on the surface interact with each other as the arrivals taxi-in towards their gates and departures taxi-out towards the runways.Aircraft have to stop at intersections to let other aircraft pass.Similarly, they have to often stop and wait for the active runway to be clear prior to crossing it.Separation rules also have to be followed.Departures also have to queue and wait prior to entering the runway so that there is adequate separation with respect to the prior aircraft that took off from the same runway.These rules have been programmed in the Surface Operation Simulator and Scheduler (SOSS) that is being developed at NASA Ames Research Center.SOSS uses kinematic models of different types of aircraft and the node-link graph to simulate surface traffic.While SOSS has been designed to work with schedulers for optimizing surface operations, SOSS has been used without a scheduler in this study.Routes from gate to runway, gate departure times and aircraft types are input to SOSS to simulate surface traffic for generating the results discussed in the section below.
+IV. ResultsThe SOSS-model-based method, described in Section III, and a data-driven method are evaluated.The first method consists of using SOSS.This means that ramp-area, movement-area and queue-area speeds specified for each aircraft type are used in the prediction.The second method is a data-driven method in which average taxi-times derived from several days of actual data are added to the gate-out time or the spot crossing time for predictions.Nineteen hours, 5:00 am to 12:00 midnight, of each day of Dallas-Fort Worth surface traffic data derived from 8/8/2011 to 8/13/2011 SDSS logs were processed to create inputs for the two methods and for validating the predictions.These days had good weather.There were a total of 5,208 arrivals and 5,256 departures.Table 10  the number of arrivals and departures, and the flow configuration on each day.DFW is operated in the southflow configuration 70% of the time. 3On the 11th, DFW switched from south-flow configuration to north-flow configuration at 2:00 p.m. local time and then back to southflow configuration at 5:00 p.m. DFW switched from north-flow configuration to south-flow configuration at 8:30 a.m. on the 13 th .The SOSS-based method and the data-driven method were used to generate estimates of queue-area entry time, runway entry time, wheels-off time, queue-area entry sequence and runway entry sequence.These times and sequences were compared against actual values derived from SDSS logs.Queue-area entry, runway entry and wheels-off time are defined using Fig. 6 as an example.Figure 6 shows the locations of queue-area entry nodes, hold nodes and departure nodes related to runway 17R.Observe that the queue-area entry nodes are placed such that the queue is set up in the correct order.For example, an aircraft on taxiway J could enter the queue-area earlier if the entry-node were placed closer to the intersection of taxiways J and Y than an aircraft that enters upstream on taxiway J (near the queue-entry node on taxiway J depicted in the figure) and still be behind the upstream aircraft.To maintain the correct order of entry into the queue, the queue-entry node is placed at the last entry node along the taxiway in the queue-area.In SOSS simulation, aircraft movement is simulated from gate to spot, from spot to runway hold node, from hold node to departure node and from departure node to wheels-off.For the example in Fig. 6, queues are formed along taxiways EF, EG and EH and taxiways J, K and L as aircraft wait to reach the runway hold nodes.SOSS computes spot crossing time, runway hold node arrival time, wait time at the hold node, departure node arrival time, wait time at the departure node and wheels-off time.Time from departure node to wheels-off is specified for different types of aircraft.Queue-entry time is determined as the time when SOSS simulated aircraft position is at or just prior to the queue-area entry node.Runway entry time is determined by adding the wait time at the hold node to the hold node arrival time.Actual hold node arrival time, runway entry time and wheels-off time are obtained by processing the track data obtained from SDSS logs with airport geometry data.Results obtained using the two methods are discussed below.
+SOSS-Model-BasedFigure 7 shows the histogram of the difference between the actual queue-area entrance time of the departures and the estimated queue-area entry times derived from SOSS simulation of 19 hours of 8/11/2011 Dallas-Fort Worth surface traffic consisting of 897 arrivals and 915 departures.August 11 was challenging because of airport configuration change from south-flow to north-flow and then back to south-flow.During the south-flow to northflow change, departures for 35L left their gates and queued in the queue-area and waited for a long time for departures and arrivals in the previous south-flow configuration to clear the runways.Queue-area entry time results shown in Fig. 7 were computed with respect to spot crossing time.This means that the SOSS simulation used spot position and time as initial conditions.The histogram in Fig. 7 shows the maximum error to be 21 minutes.It was determined that about 80% of the actual departures arrived within two-minutes early to one-minute window with respect to SOSS-based prediction of queue-area entry time.The two-minutes early to one-minute late window is   Figure 10.Cumulative absolute value of wheelsoff time estimation error.used as the wheels-off time requirement in Ref. 12 for Precision Departure Capability for Call For Release.Later on results are presented with respect to gate departure time.Gate-based results were found to be worse than the spotbased results.Figure 8 shows the cumulative absolute estimation error.For example, absolute value of queue-area entry time error is less than two-minutes for 87% and less than five-minutes for 97% of the departures.The main source of queue-area entry time error is the aircraft speeds assumed in SOSS.Actual maximum speeds in SDSS logs were found to be much higher in several instances compared to the nominal speeds assumed for the aircraft type in SOSS.The maximum speed difference was found to be 16 knots compared to SOSS speed of 15 knots.Average difference was found to be 8 knots for August 11 data.These findings suggest that SOSS nominal speeds can be better tuned to improve the estimates.Figure 9 shows the histogram of the difference between the actual wheels-off time of the flights and the estimated wheels-off time derived from SOSS simulation with reference to spot crossing.Maximum wheels-off time error was found to be 24 minutes.47% of the departures were within the two-minutes early to one-minute late window and 59% were within the plus-minus two-minute window.86% were within the plus-minus five-minute window.The cumulative absolute wheels-off time estimation error is given in Fig. 10.Comparing Fig. 9 to Fig. 7 and Fig. 10 to Fig. 8, it may be seen that queue-area entry time estimation is much better than wheels-off time estimation.Part of the reason is that Dallas-Fort Worth has multiple queues in the queuing-area from which flights exit to enter the runway.Currently a simple logic of first-in first-out based on the queue-area entry time is being used to determine runway entry order.In the real operations, flights that need to comply with traffic flow management initiatives are given priority.Better prediction of runway entry time, which directly affects wheels-off time prediction, might require knowledge of flight priority.Queue-area entry time, runway entry time and wheels-off time estimation errors were computed with spot and gate as references for the six days listed in Table 10.The two-minute early to one-minute late compliance results are summarized in Table 11.The next set of results is for queue-area entry and runway entry sequences.To determine queue-area entry sequence for SOSS simulated aircraft and actual aircraft (based on SDSS track data), time of arrival at the queue entry nodes are sorted in increasing order for departures going to each runway.Sequence error is then computed as the difference between the actual aircraft and SOSS simulated aircraft positions in the sorted lists.This same procedure is repeated to determine runway entry sequence error.It should be noted that the runway entry sequence is same as the wheels-off sequence because only one aircraft is permitted to be on an active runway at a time.Table 12 shows the percentages of departures without sequence errors on the six days.Table 13 shows percentages with at most one sequence error.This means that the actual aircraft was either in the correct sequence or was just ahead or just behind the SOSS predicted sequence.These results show that sequence errors are reduced when spot crossing time is used as a reference for SOSS-based predictions.Maximum spot-based queue-area and runway entry sequence errors were found to be 13 and 12, respectively, on August 8 data.The highest gate-based queue-area and runway entry sequence errors were found to be 12 and 12, respectively, also on August 8 data.Results for the six days show that there is a significant loss of estimation accuracy from queue-area entry to runway entry.The loss is less from runway entry to wheels-off.Gate-based results are worse than spot-based results because of imprecise gate-out time information.Gate-out time is estimated based on the proximity of the SDSS reported position to the gate and SDSS reported speed, which indicates movement.
+SDSS-Based Average Taxi-TimeThis method assumes that historical taxi-time data from spot and gate to queue-area entry locations and wheelsoff are available.This method like the SOSS-based method, discussed in the previous section, does not assume that surface surveillance information is available in real-time.Historical information can be derived from the Out-Off-On-In (OOOI) data provided by airlines and Automatic Dependent Surveillance (ADS) position reports provided by ADS equipped aircraft.To compute the spot and gate to queue-area and wheels-off taxi-times, six days of surface data were processed to identify the unique 226 spot-runway and 522 gate-runway combinations.Next, the number of departures associated with spot-runway and gate-runway were counted.Analysis showed that 25% of the spotrunway combinations were used by a single aircraft, 36% were used by two or fewer aircraft and 61% were used by 10 or fewer aircraft.The maximum number of times the spot-runway combination was used was 301 times.Of the 522 gate-runway combinations, 31% were used by only one departure, 45% were used by two or fewer departures and 57% were used by 10 or fewer departures.The maximum number of times was 38.The six days of taxi-time data were averaged and assigned to each spot-runway and gate-runway pair.Queue-entry and wheels-off times were predicted by adding the spot crossing time or the gate departure time to the average taxi-times.The actual queueentry time and wheels-off time of the departures were compared to these predictions.Table 14 shows the percentages of departures that could be predicted within the two-minutes early to oneminute late compliance window.This table also shows that spotbased estimates are a bit better than gate-based estimates.Comparing the gate-based results in Table 11 to those in Table 14, it is seen that both queue-area entry time and wheelsoff time errors are less with this method.One of the reasons for better results is that the average taxitime is same as the actual taxi-time in instances of single departures associated with a spot-runway or gate-runway pair.The second reason is that the taxi-times based on actual track data include the influence of the actual path (not the idealized node-link path) and speed.The maximum queue-entry time and wheels-off time errors were 29 minutes and 26 minutes on August 8, respectively, when spot crossing time was used as the reference for estimation.Wheels-off time could be predicted within plus-minus five-minutes for at least 90% of the departures.This minimum value of 90% was obtained for August 11 data.Maximum queue-entry time error of 27 minutes and wheels-off time error of 24 minutes were obtained with gate-based predictions of August 8 departures.At a minimum, wheels-off time of 89% percent of departures in each of the six days could be predicted within plus-minus five-minutes.The percentage, 89%, was lowest for August 11 departures.August 13 gate-based wheels-off result of 55.6% in Table 14 are close to 59.2% obtained with the neural network in Ref. 3. Wheels-off time compliance of 53.5% of neural network predictions with respect to six days, August 7 through 12, of training data is comparable to the average compliance of 51.5% in Table 14.Queue-area entry and runway entry sequence error results are given in Tables 15 and16.Maximum queue-area entry and runway entry sequence errors were obtained with August 8 data.For spot-based estimates, these were 14 and 12 departures.For gate-based estimates, these errors were 11 and 10.The sequence errors for each of the six days were found to be very close to those obtained with the method described in the previous section.Comparing the results in Tables 15 and16 to Tables 12 and13, it is seen that the gate-based estimates are a bit better with the SDSS-based Average Taxi-time model compared to with the SOSS-based simulation method.These results suggest that the Average Taxi-Time model could be used for predictions at all airports without resorting to a more detailed simulation based approach.This method is also simple to implement, it does not require airport geometry and the Gate-out and wheels-off data reported by airlines can be used to determine the average gate to wheels-off taxi-time needed by this model.The method is also computationally efficient because an addition operation is required for estimating wheels-off time and sorting is required to estimate the runway entry sequence.
+V. Conclusions and Future WorkIn the first part of this paper, 29 airports were identified for development and testing of wheels-off estimation methods after removing airports with Airport Surface Detection Equipment Model-X (ASDE-X), small-hub airports and non-hub airports from the set of 77 major U. S. airports tracked in the Federal Aviation Administration's Aviation System Performance Metrics database.These 29 airports were classified into three groups using a K-Means procedure based on taxi-out delay, traffic management delay counts and number of commercial operations.Within these three groups, San Jose International, Cleveland-Hopkins International and San Francisco International are recommended for further development and validation of wheels-off time estimation methods.San Jose has the least number of commercial operations and San Francisco has the most.In the second part of the paper, a simulation based method and a data-driven method, which uses historical taxi-time information, for estimating queue-area entry time, runway entry time and wheels-off time were described.Queue-area entry time, runway entry time and wheelsoff time estimates were generated with reference to spot crossing time and gate departure time for six days of August 2011 Dallas-Fort Worth surface traffic data.These estimates were compared with the actual values determined by processing actual ASDE-X based aircraft position data.The main findings are as follows.Spot-based estimates are better compared to gate-based estimates.The data-driven method produces better gate-based estimates compared to the Surface Operation Simulator and Scheduler (SOSS) based method assuming model speeds.If surface surveillance data are unavailable, the data-driven method could be used for estimating queue-area entry time, runway entry time and wheels-off time.This method could also be used for estimating queue-area and runway entry sequence.These conclusions are expected to be equally applicable to airports such as San Jose, Cleveland and San Francisco.The next step consists of creating a geometry and node-link model for the San Jose airport with the data received from the City of San Jose.Several days of San Jose airport surface surveillance data acquired via the Low Cost Ground Surveillance (LCGS) system will be processed to create the parameters and inputs needed by the datadriven model and by the SOSS-based simulation.The estimates generated by these methods will then be compared with LCGS derived values.Figure 1 .1Figure 1.Correlation between commercial operations and passenger enplanements.
+Figure 2 .2Figure 2. ASPM 77 airports categorized as largehub, medium-hub, small-hub and nonhub airports.
+Figure 4 .4Figure 4. Airports remaining after removing ASDE-X airports, and small-hub and non-hub airports.
+Figure 5 .5Figure 5. Node-link graph of Dalla-Fort Worth airport.
+Figure 8 .8Figure 8. Cumulative absolute value of the queuearea entry time estimation error.
+Figure 7 .7Figure 7. Queue-area entry time estimation error.
+Figure 9 .9Figure 9. Wheels-off time estimation error.Figure10.Cumulative absolute value of wheels-
+Table 3 .3Grouping based on number of commercial operations.Group IDAirportsMin.MeanStd.Max.57,64890,43018,572 116,4142PDX, ANC, CLE, TPA, CVG, RDU, BNA, IND, MCI, SJU, OAK, PIT127,723 153,339 22,103 190,1083SFO386,941 386,9410386,9411DAL, SAT, AUS, SJC, MSY, SMF, OGG, ABQ, BUF, PBI, OMA, JAX, ONT, RSW, BUR, TUS
+Table 1 .1Grouping based on average taxi-out delays in minutes.Group IDAirportsMin.Mean Std.Max.1ONT, MSY, BUR, IND, AUS, SJU, CVG, SJC, OAK, DAL, ANC, OGG1.51.90.22.12RDU, PIT, CLE, SAT, TUS, ABQ, BNA, RSW, TPA, BUF, PBI, PDX, JAX, MCI, OMA, SMF2.22.50.33.23SFO4.44.40.04.4
+Table 4 .4Grouping based on average taxi-out delay ratio in percentage.three groups obtained based on taxi-out delay are summarized in TableGroup IDAirportsMin.MeanStd. Max.1AUS, OGG, IND, SJU, ANC, CVG13.915.61.517.32ONT, CLE, PIT, SMF, BNA, TPA, PDX, RSW, MSY, OAK, BUF, MCI, DAL, OMA, BUR, PBI, JAX, SJC17.919.91.322.03SFO, ABQ, SAT, RDU, TUS23.324.21.526.8
+Table 2 .2to Table5, it is seen that 13 airports (ABQ, AUS, JAX, MSY, PBI, BUF, DAL, RSW, TUS, CVG, TPA, CLE and PIT) moved one level up in Table5, one airport-SAT jumped Grouping based on TMI-from delay counts.Group IDAirportsMin.MeanStd. Max.1MSY, PDX, AUS, SAT, OMA, DAL, SMF, BUR, SJC, OAK, TUS, ABQ, ONT, SJU, ANC, OGG104592849052CVG, PIT, CLE, IND, SFO, BUF, TPA, MCI, BNA, PBI, JAX, RSW1,1721,772412 2,5313RDU3,5573,55703,557Downloaded by NASA AMES RESEARCH CENTER on August 14, 2013 | http://arc.aiaa.org|DOI: 10.2514/6.2013-4274Copyright© 2013 by the American Institute of Aeronautics and Astronautics, Inc.The U.S. Government has a royalty-free license to exercise all rights under the copyright claimed herein for Governmental purposes.
+Table 5 .5Grouping based on FAA level.Group IDAirportsMin.MeanStd.Max.1BUR, OGG, SJC, SJU, OMA, ONT, SMF66.60.572ABQ, AUS, BNA, IND, JAX, MCI, MSY, PBI, RDU, SFO, ANC, BUF, DAL, OAK, PDX, RSW, TUS88.60.593CVG, TPA, CLE, PIT, SAT1010.40.611
+Table 6 .6Grouping based on TMI-to weather delay counts.Group IDAirportsMin.MeanStd.Max.1CLE, IND, TPA, RDU, CVG, SAT1849491462DAL30030003003SFO18,679 18,6790 18,679Table 7. Grouping based on TMI-tovolume delay counts.Group IDAirports Counts1CLE112SFO243DAL173
+Table 9 .9Grouping based on preferences to number of commercial operations and TMI-from delay counts.Group IDAirportsMean 1Mean 2Mean 31, 1, 1AUS, OGG, BUR, DAL, MSY, ONT, SJC*1.946897,4062, 1, 1OMA, SMF, ABQ, SAT, TUS2.450390,6712, 2, 1BUF, JAX, PBI, RSW2.31,42577,9201, 1, 2ANC, SJU, OAK1.8221 151,6432, 1, 2PDX2.3887 190,1081, 2, 2CVG, IND22,209 149,2392, 2, 2BNA, CLE, MCI, PIT, TPA2.51,859 150,4312, 3, 2RDU3.23,557 144,3993, 2, 3SFO4.41,848 386,941
+Table 8 .8Grouping based on taxi-out delay, TMI-from delay counts and number of commercial operations.Group IDAirportsMean 1Mean 2Mean 31, 1, 1AUS, OGG, BUR, DAL, MSY, ONT, SJC*1.946897,4061, 1, 2ANC, SJU, OAK1.8221 151,6431, 2, 2CVG, IND2.02,209 149,2392, 1, 1OMA, SMF, ABQ, SAT, TUS2.450390,6712, 1, 2PDX2.3887 190,1082, 2, 1BUF, JAX, PBI, RSW2.31,42577,9202, 2, 2BNA, CLE, MCI, PIT, TPA2.51,859 150,4312, 3, 2RDU3.23,557 144,3993, 2, 3SFO4.41,848 386,941
+Table 10 .10lists Selected days.Date# Arrivals # DeparturesDayFlow8/8/2011868881MondaySouth8/9/2011870860TuesdaySouth8/10/2011900902Wednesday South8/11/2011897915ThursdaySouth, North, South8/12/2011888892FridaySouth8/13/2011785806SaturdayNorth, SouthFigure 6.Examples of queue-area entry nodes, hold nodes and departure nodes.
+Table 11 .11Compliance within two-minutes early to one-minute late.DateSpot-based Queue (%) Runway (%) Wheels-off (%) Queue (%) Runway (%) Wheels-off (%) Gate-based885.864.058.260.047.146.3983.753.750.363.344.540.91084.758.553.361.047.044.91180.152.647.059.542.438.61287.358.051.863.049.046.11384.660.454.262.451.447.5Table 12. No sequence error.DateSpot-based Queue (%) Runway (%) Queue (%) Runway (%) Gate-based870.462.049.547.1965.951.752.744.41072.959.148.845.01169.154.851.445.21272.262.653.647.91372.564.552.048.3
+Table 13 .13At most one sequence error.DateSpot-based Queue (%) Runway (%) Queue (%) Runway (%) Gate-based894.690.785.983.2993.383.585.577.21094.690.784.181.21193.086.385.980.01292.989.084.981.41395.291.485.081.8
+Table 14 .14Compliance within two-minutes early to one-minute late.DateSpot-based Queue (%) Wheels-off (%) Queue (%) Wheels-off (%) Gate-based886.956.577.750.6986.250.077.748.11086.652.578.451.61181.754.674.552.11286.853.377.250.91386.157.477.855.6
+Table 16 .16At most one sequence error.DateSpot-based Queue (%) Runway (%) Queue (%) Runway (%) Gate-based896.091.490.387.1993.484.588.381.51095.690.691.985.51193.983.290.781.11294.189.590.586.91395.489.790.286.0
+Table 15 .15No sequence error.models if gate departure time based wheels-off time and runway entry sequence predictions are desired.node-linkDateSpot-based Queue (%) Runway (%) Queue (%) Runway (%) Gate-based871.764.963.152.4970.054.260.051.01071.758.461.550.81168.352.263.047.31274.660.459.452.71373.458.661.851.9
+Table A -A1. 77 ASPM airports.#Airport CodeAirport NameLocation1ABQAlbuquerque International SunportAlbuquerque, New Mexico2ANCTed Stevens Anchorage InternationalAnchorage, Alaska3ATLHartsfield -Jackson Atlanta InternationalAtlanta, Georgia4AUSAustin-Bergstrom InternationalAustin, Texas5BDLBradley InternationalWindsor Locks, Connecticut6BHMBirmingham-Shuttlesworth InternationalBirmingham, Alabama7BNANashville InternationalNashville, Tennessee8BOSGeneral Edward Lawrence Logan InternationalBoston, Massachusetts9BUFBuffalo Niagara InternationalBuffalo, New York10BURBob HopeBurbank, California11BWIBaltimore/Washington International Thurgood Marshall Baltimore, Maryland12CLECleveland-Hopkins InternationalCleveland, Ohio13CLTCharlotte/Douglas InternationalCharlotte, North Carolina14CVGCincinnati/Northern Kentucky InternationalCovington, Kentucky15DALDallas Love FieldDallas, Texas16DAYJames M Cox Dayton InternationalDayton, Ohio17DCARonald Reagan Washington NationalWashington, District of Columbia18DENDenver InternationalDenver, Colorado19DFWDallas/Fort Worth InternationalDallas-Fort Worth, Texas20DTWDetroit Metropolitan Wayne CountyDetroit, Michigan21EWRNewark Liberty InternationalNewark, New Jersey22FLLFort Lauderdale/Hollywood InternationalFort Lauderdale, Florida23GYYGary/Chicago InternationalGary, Indiana24HNLHonolulu InternationalHonolulu, Hawaii25HOUWilliam P HobbyHouston, Texas26HPNWestchester CountyWhite Plains, New York27IADWashington Dulles InternationalWashington, District of Columbia28IAHGeorge Bush Intercontinental/HoustonHouston, Texas29INDIndianapolis InternationalIndianapolis, Indiana30ISPLong Island Mac ArthurNew York, New York31JAXJacksonville InternationalJacksonville, Florida32JFKJohn F Kennedy InternationalNew York, New York33LASMc Carran InternationalLas Vegas, Nevada34LAXLos Angeles InternationalLos Angeles, California35LGALa GuardiaNew York, New York36LGBLong Beach (Daugherty Field)Long Beach, California36MCIKansas City InternationalKansas City, Missouri38MCOOrlando InternationalOrlando, Florida39MDWChicago Midway InternationalChicago, Illinois40MEMMemphis InternationalMemphis, Tennessee41MHTManchesterManchester, New Hampshire42MIAMiami InternationalMiami, Florida43MKEGeneral Mitchell InternationalMilwaukee, Wisconsin44MSPMinneapolis-St Paul International/Wold-ChamberlainMinneapolis, Minnesota
+Table A -A1. 77 ASPM airports (Contd.).Table A-2. Airport metrics.# # Airport Airport Code Code Taxi-out Avg.Taxi-out Avg.from TMI-Airport Name Air-carrier ItinerantWeather TMI-toVolume TMI-toLocation EnplanementsLevel45MSY DelayLouis Armstrong New Orleans International Delay Delay and Air-taxi DelayDelayNew Orleans, Louisiana46OAK (min.)Metropolitan Oakland International Ratio (%) Counts Ops.CountsCountsOakland, California147ABQOGG2.51Kahului 23.929399,7990Kahului, Hawaii 0 2,768,4359248ANCOMA1.66Eppley Airfield 14.0132186,6989Omaha, Nebraska 2 2,354,9878349ATLONT7.52Ontario International 37.0 11,132916,8245,561Ontario, California 1,271 44,414,12112450AUSORD1.97Chicago O'Hare International 17.3 858 113,1110Chicago, Illinois 0 4,436,6619551BDLOXR2.67Oxnard 20.81,22986,8380Oxnard, California 0 2,772,3157652BHMPBI2.34Palm Beach International 19.8 36358,6940West Palm Beach, Florida 0 1,429,2828753BNAPDX2.43Portland International 21.1 1,576142,2472Portland, Oregon 0 4,673,0479854BOSPHL5.2Philadelphia International 28.7 5,734355,6078,964Philadelphia, Pennsylvania 80 14,180,730 10955BUFPHX2.38Phoenix Sky Harbor International 19.7 1,764 80,6450Phoenix, Arizona 0 2,582,59781056BURPIT2.07Pittsburgh International 18.9 41267,7260Pittsburgh, Pennsylvania 0 2,144,91571157BWIPSP3.35Palm Springs International 26.7 4,715258,540561Palm Springs, California 97 11,067,31991258CLEPVD2.74Theodore Francis Green State 21.9 2,203 179,382146Providence, Rhode Island 11 4,401,033101359CLTRDU5.52Raleigh-Durham International 30.9 4,308 513,8021,468Raleigh/Durham, North Carolina 1,558 19,022,535 121460CVGRFD1.93Chicago/Rockford International 13.9 2,531 157,36720Chicago/Rockford, Illinois 1 3,422,466 111561DALRSW1.78Southwest Florida International 19.2 458 116,414300Fort Myers, Florida 173 3,852,886862 16 DAYSAN2.57San Diego International 19.8 82148,2170San Diego, California 0 1,247,33381763DCASAT4.3San Antonio International 26.7 7,565278,757895San Antonio, Texas 95 9,053,004101864DENSDF3.71Louisville International -Standiford Field 26.5 3,302 630,9691,702Louisville, Kentucky 7 25,667,4991265 19 DFWSEA3.26Seattle-Tacoma International 22.7 5,296 640,5411,592Seattle, Washington 5 27,518,3581266 20 DTWSFO3.25San Francisco International 18.5 4,149436,534997San Francisco, California 303 15,716,8651167 21 EWRSJC8.49Norman Y. Mineta San Jose International 40.4 6,476 402,988 26,201San Jose, California 1,419 16,814,092102268FLLSJU4.16Luis Munoz Marin International 26.2 4,397 227,06125San Juan, Puerto Rico 20 11,332,466969 23 GYYSLC0.09Salt Lake City International 0.8 101,5740Salt Lake City, Utah 0 1,42002470HNLSMF1.96Sacramento International 15.5 29206,4460Sacramento, California 0 8,689,6991171 25 HOUSNA1.8John Wayne-Orange County 19.9 593139,280112Santa Ana, California 238 4,753,55482672HPNSTL2.39Lambert-St Louis International 18.8 1,946 64,60159St Louis, Missouri 171 972,38572773IADSWF4.04Stewart International 25.2 5,766314,384899Newburgh, New York 350 11,044,383112874IAHTEB4.84Teterboro 29.34,031516,708797Teterboro, New Jersey 166 19,306,660122975INDTPA2.02Tampa International 16.6 1,886141,11149Tampa, Florida 0 3,670,39693076ISPTUS1.51Tucson International 17.5 13619,8280Tucson, Arizona 0 781,39673177JAXVNY2.22Van Nuys 18.01,19177,3150Van Nuys, California 0 2,700,514932JFK8.532.79,296405,9768,89697523,664,8321033LAS3.7926.62,700484,19441613119,872,6171134LAX3.9626.64,503583,16780832130,528,7371135LGA11.6747.3 10,716364,14019,8453,19511,989,2271036LGB2.2316.729736,839001,512,212836MCI2.2119.61,603136,398005,011,000938 MCO3.6727.33,831300,075952917,250,4151139 MDW2.9927.11,855209,789661429,134,576840 MEM2.1414.11,883291,700550114,344,2131041 MHT2.0318.21,14551,376001,342,308542MIA3.4321.22,623375,20924712218,342,15812
+			Downloaded by NASA AMES RESEARCH CENTER on August 14, 2013 | http://arc.aiaa.org| DOI: 10.2514/6.2013-4274Copyright© 2013 by the American Institute of Aeronautics and Astronautics, Inc.The U.S. Government has a royalty-free license to exercise all rights under the copyright claimed herein for Governmental purposes.
+		
+		
+
+			
+AcknowledgementsThe authors thank Dr. Tatsuya Kotegawa, Dr. Waqar Malik, William Chan and Dr. Banavar Sridhar for their careful review and suggestions for improving the paper.
+			
+
+			
+AppendixThe 77 major U. S. airports in the ASPM database are listed in Table .A-1. Table A-			
+			
+
+				
+
+
+	
+		NASA Contract Number: NNA12AA14C, Mosaic ATM, Inc., 801 Sycolin Road, SE, Suite 306
+		
+			AtmMosaic
+		
+		
+			Inc
+		
+	
+	
+		Airport Surface Traffic Management Requirements Resulting from Variations in Airport Characteristics: Report on Airport Survey
+		Leesburg, VA
+		
+			20175. March 21, 2012
+			5686
+		
+	
+	Mosaic ATM, Inc., "Airport Surface Traffic Management Requirements Resulting from Variations in Airport Characteristics: Report on Airport Survey," NASA Contract Number: NNA12AA14C, Mosaic ATM, Inc., 801 Sycolin Road, SE, Suite 306, Leesburg, VA 20175-5686, March 21, 2012.
+
+
+
+	
+		Performance Evaluation of SARDA: An Individual Aircraft-based Advisory Concept for Surface Management
+		
+			YoonJung
+		
+		
+			TyHoang
+		
+		
+			MiwaHayashi
+		
+		
+			WaqarMalik
+		
+		
+			LeonardTobias
+		
+		
+			GautamGupta
+		
+		10.2514/atcq.22.3.195
+	
+	
+		Air Traffic Control Quarterly
+		Air Traffic Control Quarterly
+		1064-3818
+		2472-5757
+		
+			22
+			3
+			
+			2011
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+		
+			USA/Europe Air Traffic Management Research and Development Seminar
+		
+	
+	Jung, Y., Hoang, T., Montoya, J., Gupta, G., Malik, W., Tobias, L., and Wang, H., "Performance Evaluation of a Surface Traffic Management Tool for Dallas/Fort Worth International Airport," 9th USA/Europe Air Traffic Management Research and Development Seminar, 2011, pp. 1-10.
+
+
+
+	
+		Wheels-Off Time Prediction Using Surface Traffic Metrics
+		
+			GanoChatterji
+		
+		
+			YunZheng
+		
+		10.2514/6.2012-5699
+		
+	
+	
+		12th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference and 14th AIAA/ISSMO Multidisciplinary Analysis and Optimization Conference
+		Indianapolis, Indiana
+		
+			American Institute of Aeronautics and Astronautics
+			Sep. 17-19, 2012
+		
+	
+	Federal Aviation Administration. cited: 2/25/2013
+	Chatterji, G. B., and Zheng, Y., "Wheels-Off Time Prediction Using Surface Traffic Metrics," AIAA-2012-5699, Proceedings of 12th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference, Indianapolis, Indiana, Sep. 17- 19, 2012. 4 Federal Aviation Administration, URL: http://www.faa.gov/airports/planning_capacity/passenger_allcargo_stats/passenger/ [cited: 2/25/2013].
+
+
+
+	
+		Health hazard evaluation report: HHE-81-042-832, Federal Aviation Administration, New York Air Route Traffic Control Center, Ronkonkoma, New York.
+		10.26616/nioshhhe81042832
+		cited: 2/25/2013
+		
+		
+			May 12, 2009
+			U.S. Department of Health and Human Services, Public Health Service, Centers for Disease Control, National Institute for Occupational Safety and Health
+		
+	
+	Air Traffic Control Complexity Formula for Terminal and En Route Pay Setting by Facility
+	5 Appendix A of the Mediation to Finality process adopted by the Federal Aviation Administration and the National Air Traffic Controllers Association, "Air Traffic Control Complexity Formula for Terminal and En Route Pay Setting by Facility," May 12, 2009, URL: http://nwp.natca.net/Documents/Arbitration_Award.pdf [cited: 2/25/2013].
+
+
+
+	
+		FAA Aviation Forecasts Fiscal Years 1979-1990. Federal Aviation Administration, Office of Aviation Policy, 800 Independence Avenue, S.W., Washington, D.C. 20591. September 1978. 92p
+		10.1177/004728757901800127
+		cited: 2/25/2013
+		
+	
+	
+		Journal of Travel Research
+		Journal of Travel Research
+		0047-2875
+		1552-6763
+		
+			18
+			1
+			
+			
+			SAGE Publications
+		
+	
+	Federal Aviation Administration
+	6 Federal Aviation Administration, "Terminal Area Forecast Summary Fiscal Years 2011-2040," URL: http://www.faa.gov/about/office_org/headquarters_offices/apl/aviation_forecasts/taf_reports/media/TAF_summary_report_FY20 112040.pdf [cited: 2/25/2013].
+
+
+
+	
+		Runway Safety
+		
+			KimCardosi
+		
+		
+			StephanieChase
+		
+		
+			DanielleEon
+		
+		10.2514/atcq.18.3.303
+		cited: 2/25/2013
+		
+	
+	
+		Air Traffic Control Quarterly
+		Air Traffic Control Quarterly
+		1064-3818
+		2472-5757
+		
+			18
+			3
+			
+			2010
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	7 Federal Aviation Administration Air Traffic Organization, "Annual Runway Safety Report 2010," URL: http://www.faa.gov/airports/runway_safety/news/publications/media/Annual_Runway_Safety_Report_2010.pdf [cited: 2/25/2013].
+
+
+
+	
+		Distributing net-enabled federal aviation administration (FAA) weather data
+		
+			MarkSimons
+		
+		10.1109/icnsurv.2008.4559189
+		cited: 2/25/2013
+		
+	
+	
+		2008 Integrated Communications, Navigation and Surveillance Conference
+		
+			IEEE
+			
+		
+	
+	8 Federal Aviation Administration, "Low Cost Ground Surveillance (LCGS)," URL: http://www.faa.gov/about/office_org/headquarters_offices/ang/offices/ac_td/td/projects/lcgs/ [cited: 2/25/2013].
+
+
+
+	
+		Comparing European ATM master plan and the NextGen implementation plan
+		
+			DavidBatchelor
+		
+		10.1109/icnsurv.2015.7121357
+		cited: 2/25/2013
+		
+	
+	
+		2015 Integrated Communication, Navigation and Surveillance Conference (ICNS)
+		
+			IEEE
+			March 2012
+		
+	
+	NextGen Implementation Plan
+	9 Federal Aviation Administration, "NextGen Implementation Plan," March 2012, URL: http://www.faa.gov/nextgen/implementation/media/NextGen_Implementation_Plan_2012.pdf [cited: 2/25/2013].
+
+
+
+	
+		Characterization of Days Based on Analysis of National Airspace System Performance Metrics
+		
+			GanoChatterji
+		
+		
+			BassamMusaffar
+		
+		10.2514/6.2007-6449
+	
+	
+		AIAA Guidance, Navigation and Control Conference and Exhibit
+		Hilton Head, South Carolina
+		
+			American Institute of Aeronautics and Astronautics
+			Aug. 20-23, 2007
+		
+	
+	AIAA-2007-6449
+	10 Chatterji, G. B., and Musaffar, B., "Characterization of Days Based on Analysis of National Airspace System Performance Metrics," AIAA-2007-6449, Proceedings of AIAA Guidance, Navigation and Control Conference and Exhibit, Hilton Head, South Carolina, Aug. 20-23, 2007.
+
+
+
+	
+		Automating the Process of Terminal Area Node-Link Model Generation
+		
+			Hak-TaeLee
+		
+		
+			ThomasRomer
+		
+		10.2514/6.2008-7101
+	
+	
+		AIAA Modeling and Simulation Technologies Conference and Exhibit
+		Honolulu, Hawaii
+		
+			American Institute of Aeronautics and Astronautics
+			August 18-21, 2008
+			12
+		
+	
+	AIAA 2008-7101
+	Lee, H., and Romer, T. F., "Automating the Process of Terminal Area Node-Link Model Generation," AIAA 2008-7101, Proc. AIAA Modeling and Simulation Technologies Conference and Exhibit, Honolulu, Hawaii, August 18-21, 2008. 12
+
+
+
+	
+		Characterization of Tactical Departure Scheduling in the National Airspace System
+		
+			RichardCapps
+		
+		
+			ShawnEngelland
+		
+		10.2514/6.2011-6835
+	
+	
+		11th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference
+		Virginia Beach, VA
+		
+			American Institute of Aeronautics and Astronautics
+			September 20-22, 2011
+		
+	
+	AIAA 2011-6835
+	Capps A., and Engelland, S. A., "Characterization of Tactical Departure Scheduling in the National Airspace System," AIAA 2011-6835, Proc. AIAA Aviation Technology, Integration and Operations Conference (ATIO), Virginia Beach, VA, September 20-22, 2011.
+
+
+				
+			
+		
+	

file129.txt

@@ -0,0 +1,320 @@
+
+	
+	
+		
+I. Introductionhis paper is motivated by the need for improving departure scheduling advisories.A specific application is for Precision Departure Release Capability (PDRC) which has the goal of using surface trajectory based off-time predictions for Call for Release (CFR).Accurate wheels-off time estimation is important to PDRC both for 1) providing an initial estimate to the Traffic Management Advisor (TMA) for inserting a flight into a constrained overhead stream and for scheduling an inbound flight to a TMA-metered airport, and 2) determining when to release the aircraft from the gate so that its actual wheels-off time corresponds to the wheels-off time coordinated with the Air Route Traffic Control Center.Improved wheels-off time prediction would also benefit traffic flow management (TFM).TFM techniques first estimate traffic demand considering both airborne aircraft that are detected by the air traffic control radars and aircraft on the ground that are scheduled to depart within the forecasting interval, and then
+II. Dallas-Fort Worth Airport Geometry and OperationsDallas-Fort Worth International Airport (DFW) is the third busiest airport in the United States.According to the March 2011 FAA Administrator's Factbook 7 , 652,000 operations were conducted at DFW in 2010 compared to 950,000 operations at Hartsfield-Jackson Atlanta International, the busiest US airport, and 883,000 operations at Chicago O'Hare International, the second busiest US airport.DFW is physically one of the largest airports in the United States and the world with an area spanning five nautical-miles east to west and three nautical-miles north to south.The airport has seven physical runways shown in Fig. A-1 in the Appendix.These runways are operated in the south-flow and north-flow configurations.The runways in south-flow configuration are designated as, 13L, 13R, 17L, 17C, 17R, 18L and 18R.The runways in the north-flow configuration are designated as, 31L, 31R, 35L, 35C, 35R, 36L and 36R.These designations also indicate the runway heading with respect to north; they are physically painted on the two ends of each runway.To get insight into DFW operations, hourly runway configurations (arrival and departure runways) for each day in 2011 were obtained by generating the "Daily Weather by Hour Report," from the ASPM database.These data were then processed to determine runways that are used for both arrivals and departures, runways that are used only for arrivals and runways that are used solely for departures.Table 1 lists the runways and the time duration of usage during 2011.This table shows that 10 of the 14 runways (considering the same physical runway to be two different runways based on the approach direction) are used both for arrivals and departures.Of these, 18R is used most often.Summing the runway utilization in both directions of the physical runways, it is seen that 17C/35C, 17R/35L, 18L/36R, 18R/36L, 17L/35R and 13R/31L usage is within 20% of each other.Runway 13L/31R usage is 27% of the usage of the most utilized runway, 18R/36L.Data in Table 1 show that runways 17C, 17R, 18L and 18R are utilized more often compared to the other runways; thus, south-flow is the dominant flow direction in DFW operations.Further analysis of ASPM runway configuration data revealed that 76 arrivaldeparture runway configurations out of 254 theoretically possible configurations were used in 2011.Number of theoretically possible combinations is obtained using,∑ ∑ ≤ ≤ ≤ ≤         +         = 7 1 7 1 7 7 k k k k n (1)where        k 7means the number of ways in which "k" runways can be chosen out of seven runways.The two summations are for the seven south-flow and seven north-flow runways.Table 2 lists the top ten configurations along with the percentage of time they were used with respect to 8,760 hours in the year.Observe that the ten configurations listed in Table 2 were used 95% of the time.Summing the percentage use of the 76 south-flow and north-flow configurations, it is seen that the south-flow configuration is used 70% of the time and the north-flow configuration is used 30% of the time.Reference 6 also cited the same numbers.Passenger and cargo traffic in and out of the 233 gates cross 59 spots to exit and enter a network of 96 taxiways to travel to and from the seven physical runways.The taxiways intersect at 215 locations.This includes intersection   of taxiways with runways.The taxiways consist of 360 segments partitioned by the intersections.These numbers were obtained by analyzing DFW geometry data derived from SODAA.The DFW airport design is also characterized by multiple runways, high-speed exit taxiways, non-intersecting runways, three towers, and perimeter taxiways near 35L and 35C that permit aircraft to taxi to and from the gates without crossing active runways.These features reduce congestion on the airport surface; delays are mostly weather related and independent of surface operations. 6This suggests that it might be possible to estimate taxi time with reasonable accuracy for different weather conditions at DFW.To estimate wheels-off time, estimates of gate departure time and taxi time are needed.Of these two estimates, gate departure time estimate is much more difficult to compute.In the absence of airline provided gate departure time, the choices are limited to the scheduled departure time from the Official Airline Guide (OAG) and the proposed departure time included in the filed flightplan.These times provide approximate estimates of gate-out time.Reference 1 proposes the use of pre-departure event times to improve estimate of gate-out time.The central idea employed in Ref. 1 is that gate-out time becomes more certain as completion of each step taken by airline and air traffic for preparing the flight for departure are reported via the Aircraft Communication Addressing and Reporting System (ACARS).Taxi time is a function of distance between the gate and the runway, taxi-speed and congestion of the surface.Since a taxi route is specified, distance along the route can be determined.Taxi-speed differences between flights are significant; taxi-speed is a function of pilot preference, stops needed at intersections and congestion along the route to the runway threshold.To determine gate to runway distances and taxi-out times to the runways at DFW, one week-spanning 7 August 2011 through 13 August 2011 of surface traffic data were obtained by processing SMS logs.The chosen seven days had good weather, and consisted of 6,284 departures.After discarding flights with more than 60 minutes of gate departure delay and 30 minutes of gate to runway entry time, 5,822 departures were considered for further analysis.It was determined that 393 unique gate-runway combinations were used by these aircraft.Figure 1 shows the gate to runway distance distribution and Fig. 2 shows the average taxi speed distribution of 5,822 departures.Average taxi speed is obtained as the ratio of gate to runway distance to taxi-out time.Average, standard deviation and maximum gate to runway distance were found to be 1.6, 0.6 and 4.0 nautical-miles, respectively.The average, standard deviation and maximum of the average taxi-out speed were found to be 13, 3.6 and 34.5 knots, respectively.Average and standard deviation of the taxi-out time were determined to be 7.8 and 3.6 minutes, respectively.
+III. Correlation Between Metrics and Gate to Wheels-off Time and Gate Departure DelayCorrelation coefficients between gate to wheels-off time (SMS Log Data Item 4 in the Appendix) and variables selected from the two sets discussed in the SMS Log Data and ASPM Data sections in the Appendix were computed to make an assessment of their suitability as predictors of gate to wheels-off time in a neural network and a linear model framework.The procedure was repeated for correlations with respect to gate departure delay (SMS Log Data Item 15 in the Appendix) to ascertain the ability of these variables to predict gate departure delay.Prior to computing the correlation coefficients, data were conditioned as follows.SMS Log data for flights with absolute value of the gate departure delay greater than the specified threshold of 60 minutes were discarded.The data were found to contain large negative gate departure delays.Another check was performed for gate to runway taxi time with a threshold value of 30 minutes.Only flights with gate to runway entry time of less than 30 minutes were considered.Total number of samples prior to pruning is 6,284.After pruning 5,822 remain.462 samples removed represent a 7.4% loss with respect to 6,284 samples.Table 3 shows the correlation (with 100% being perfect correlation) and p-value of the selected variables with gate to wheels-off time.Here, correlation means the cross-correlation coefficient derived from the covariance matrix of the two variables being compared and the p-value is the probability of obtaining a correlation as large as the observed value by random chance, when the true correlation between the variables is zero.p-value of less than 0.05 is considered to be significant.The gate to runway distance has the highest correlation followed by average taxi-out delay and number of departures on the surface.Wind angle, visibility and temperature were found to be negatively correlated.Weak negative correlation with visibility and temperature is reasonable in that as visibility and temperature decrease, one would expect taxi-time to increase a bit.Wind angle correlation is difficult to interpret without examining wind velocity components relative to the surface trajectory.Few variables that are independent of each other (contain different type of information) with high correlation can be used to develop a linear model or a neural network model for predicting gate to wheels-off time.Such models could be adequate for CFR at DFW because in current operations, CFR is initiated after pushback from the gate.This means that the gate departure time is known.At some other airports, for example at San Jose International airport, airlines are asked to inform air traffic control (ATC) some time (for example, 15 minutes) prior to ready for departure when the flight is impacted by CFR.Gate departure time uncertainty can be expected to be small in this scenario therefore wheels-off time could be predicted with these models.For predicting gate departure time, the only available information about when an aircraft might leave the gate is the scheduled gate departure time.Improvement beyond the scheduled gate departure time is possible if metrics derived from observed or estimated airport state are found to be good predictors of gate departure delay.This is examined next.Table 4 shows the correlation of the selected metrics with gate departure delay.The table shows that time of day at the scheduled gate departure time has the highest correlation followed by average gate departure delay in the previous 15-minutes.Average taxi-in delay in previous 15-minutes is similarly correlated with gate departure delay.While unexpected, temperature was found to be mathematically correlated to gate departure delay.It turns out that temperature increase from morning to afternoon and then decrease in the evening is somewhat correlated to the schedule of the departure pushes.The degree of correlation can be expected to change with different weather conditions.Comparing Tables 3 and4, it is seen that the degree of correlation of the selected metrics with gate departure delay is much lower compared to with gate to wheels-off time.This suggests that constructing a model for reliable prediction of gate departure delay based on these metrics might be difficult.Furthermore, wheels-off time prediction at the scheduled gate departure time requires that the metrics computed based on airport state data at scheduled departure time be correlated to gate to wheels-off time at actual gate departure time.These correlations are likely to be worse compared to the correlations in Table 3 which are based on airport state data at or close to the actual gate departure time and not at an earlier time.These issues have not been examined further in this paper.The rest of the paper assumes that gate departure time is known and a prediction of gate to wheels-off time is needed for wheels-off time estimate.
+IV. Neural Network Model and ResultsA three-layer neural network with seven nodes in the input layer, 20 nodes in the hidden layer and one node in the output layer was designed to predict gate to wheels-off time.Such a neural network is shown in Fig. 3.The seven selected inputs are, 1) gate to runway distance, 2) average taxi-out delay in previous 15-minutes, 3) number of departures on surface at actual gate departure time, 4) average taxi-out delay of departures on same runway in previous 15-minutes, 5) average taxi-out delay of departures to same fix in previous 15-minutes, 6) wind angle and 7) ATC set airport arrival rate.Gate departure count in previous 15 minutes (#6 in Table 3) and number of departures to same fix in previous 15 minutes (#7 in Table 3) were not considered as inputs for the neural network because they were found to be significantly correlated to the other inputs.For example, correlation between metric #6 and #3 is 53.5% and between #7 and #5 is 43.5%.The single output is the gate to wheels-off time.Maximum absolute value of each input was determined for the entire seven day  dataset.Inputs were then normalized with these values.The gate to wheels-off time data used for training the neural network were also normalized by the maximum absolute value obtained from seven days of gate to wheels-off time data.Figure 3 shows that the inputs are multiplied by weights and summed together with a bias at each hidden layer node and input to the sigmoid function, which is real-valued and differentiable.This means that the neural network had 140 weights and 20 biases for seven inputs and 20 nodes in the hidden layer.The output of the sigmoid functions in the hidden layer are multiplied with another set of weights and summed together with a bias and input to sigmoid functions in the output layer.Since this neural network has 20 nodes in the hidden layer and one node in the output layer, there are 20 weights and one bias between the hidden and output layers.The 160 neural network weights and 21 biases were initialized with values between -1 and 1 using a uniform random number generator.These weights were then adjusted using the standard gradientbased back-propagation algorithm in the neural network training step.Four-hundred iterations resulted in reduced error between the gate to wheels-off time predicted by the neural network and the gate to wheels-off time used for training the network as shown in Fig.     runway distance correlation with gate to wheels-off time of 66.3%.Average and standard deviation of the error with respect to the gate to wheels-off time training data turned out to be 23 seconds and 2.6 minutes, respectively.The average and standard deviation of gate to wheels-off time in the six day training set is 8.9 and 3.8 minutes, respectively.A histogram of the training error is shown in Fig. 5.Reference 8 notes that while the compliance window for CFR varies by facility and that a nationwide standard does not exist, information from traffic managers and inter-facility agreements ask for flights to depart within a three-minute window, two-minutes early to one-minute late, with respect to the coordinated departure time.The reason for allowing departures to be two-minutes early compared to one-minute late is because it is easier to slow down the flight compared to accelerating it for merging into the constrained flow.Considering the neural network predicted gate to wheels-off time to be the coordinated departure time and the difference of the actual gate to wheels-off time with respect to this predicted time to be delay, 53.5% of the flights were found to be in compliance with the CFR window based on the error distribution in Fig. 5.This increased to 61.2% when the window was expanded to allow two-minute early to two-minute late departures.For the 13 August gate to wheels-off time test data, the correlation with the neural network generated gate to wheels-off time estimate is 74%.Average and standard deviation of the error with respect to the test data are 35 seconds and 2.3 minutes.These values can be compared to the average and standard deviation of gate to wheels-off time of 8.7 minutes and 3.4 minutes.Gate to wheels-off time test data are shown in Fig. 6 and the departure distribution of the flights whose gate to wheels-off time are in Fig. 6 are shown in Fig. 7 in 15-minute bins.The error distribution is given in Fig. 8. Actual gate to wheels-off times for 59.2% of the flights in the test set were found to be within the two-minute early and one-minute late CFR window with respect to the neural network predicted gate to wheels-off times.This result compares favorably with the observation in Ref. 8, which is based on onemonth of data, that 69.2% of aircraft subject to CFRs in which TMA automation was utilized were compliant with the CFR window.Compliance of the test set improved to 66.5% when the window was expanded to allow twominute early to two-minute late departures.These results confirm that the neural network performance on the test data is as good as it is on the training data.Results discussed in this section show that the selected metrics can be used as inputs to a neural network for generating gate to wheels-off time predictions for CFR after gate pushback.The performance of the neural network can be improved further by removing outliers from the training and test sets.This would require flights with unusual delay to be identified and removed from the training and test sets based on detailed analysis of surface trajectory of each flight.To compare the results obtained with the neural network with the earlier study reported in Ref. 5, a linear model was set up as follows:∑ ≤ ≤ = 7 1 k k k x c y (2)with 1x through 7x representing the seven non-normalized neural network inputs, 1 c through 7 c representing the corresponding coefficients and y representing the non-normalized gate to wheels-off time.These seven coefficients were computed using the leastsquares method with the left and the right hand sides of Eq. ( 2) derived from six days of data that had been used earlier to train the neural network.The numerical values of the coefficients are given in Table 5. Inputs derived from one day of data used for testing the neural network were then multiplied with these coefficients and summed to generate gate to wheelsoff time predictions for comparison with the actual gate to wheels-off time.Results obtained with the linear model matched the distributions shown in Fig. 5 and 8 based on the Mahalanobis distance metric, which is the ratio of the Euclidean Norm of the error (data in Figs. 5 and8) and the standard deviation of the gate to wheels-off time distributions used for training and testing.Mahalanobis distance metric values were determined to be 49.55 and 49.02 with neural network and linear model outputs with respect to training data, respectively.With respect to test data, the values were found to be 19.42 and 19.29 with neural network and linear model outputs, respectively.Actual gate to wheels-off times of 62.1% of the flights in the test set were found to be within the two-minute early and one-minute late CFR window with the linear model.CFR compliance is a bit better than 59.2% obtained with the neural network.These results do not show a benefit of using the neural network over a simple linear model for DFW traffic.It remains to be seen if the neural network would perform better on surface data from other airports.Results obtained with both the neural network and linear model validate the suitability of the chosen metrics for predicting gate to wheels-off time.For follow on work, these metrics will be computed using data from airports where SMS will not be available and used with the neural network and linear model to assess the accuracy of gate to wheels-off time predictions.
+V. ConclusionsCorrelation of airport state metrics derived from the Aviation System Performance Metrics database and Surface Management System logs with gate to wheels-off time and gate departure delay were examined to identify metrics with significant correlation as inputs for a neural network.Gate to runway distance was found to have the highest correlation of 66.3% with gate to wheels-off time.Scheduled departure time of day was found to have the highest correlation of 13.8% with gate departure delay.Given low correlation with gate departure delay, this study did not attempt to develop a model for predicting gate departure time.Instead, gate departure time was assumed to be known.The neural network was trained with six days of data to predict gate to wheels-off time.After training, the correlation with gate to wheels-off time predicted by the neural network and that used for training increased by 6% to 72.4% compared to 66.3% correlation with gate to runway distance.Average and standard deviation of the error with respect to the gate to wheels-off time training data were found to be 23 seconds and 2.6 minutes.One day of data were used for testing the neural network.A 74% correlation was found between these test set data and the neural network generated gate to wheels-off time estimate.Average and standard deviation of the error with respect to the test data were determined to be 35 seconds and 2.3 minutes.These results show that the neural network performance on the test data is comparable to its performance on the training data.Actual wheels-off times for 59% of the departures in the test set were found to be within the two-minute early to one-minute late Call for Release window with respect to the trained neural network predicted wheels-off times.This result is comparable to 69% compliance within the Call for Release window reported in an earlier study.Results based on analysis of Dallas-Fort Worth data show that it is feasible to use the selected metrics as inputs to a neural network for generating gate to wheels-off time predictions for Call for Release after gate pushback.Results obtained with a linear model, with coefficients obtained using the least-squares method, were found to be as good as those obtained with the neural network based on the Mahalanobis distance metric.While a clear benefit of using a neural network over the simple linear model was not found for Dallas-Fort Worth traffic, it remains to be seen if it would perform better on surface data from other airports.Both the approaches suggest that the selected metrics can be used for predicting gate to wheels off time.These metrics will be computed using data from airports where the Surface Management System will be unavailable and used with the neural network and the linear model to determine the accuracy with which gate to wheels-off time can be predicted.
+Parameters for Modeling Gate to Wheels-off Time and Gate Departure DelayThe variables chosen for predicting gate to wheels-off time and gate departure delay are discussed in this section.These input variables were obtained by processing seven days, 7 August 2011 through 13 August 2011, of SMS logs and ASPM data.
+SMS Log DataSurface trajectory data, consisting of a sequence of latitudes and longitudes as a function of time, of every flight departing DFW were analyzed to determine the following:1. Actual gate-out time in hours, minutes and seconds.2. Time of day in 15-minute interval.For example, 02:45 means 2 hours and 45 minutes past 00:00 local time.Actual gate-out time is within this 15-minute bin. 3. Actual wheels-off time in hours, minutes and seconds.4. Actual wheels-off time minus actual gate-out time.This is the sum of time spent in the ramp area, time taken to taxi to the runway and time spent on runway till the wheels are off the ground.5. Flight ID. 6. Departure gate ID. 7. Departure runway ID. 8. Name of the departure fix.9. Name of the destination airport.10.Actual gate to runway distance in nautical-miles based on surface trajectory.11.Actual taxi-out time of the flight in seconds.This is the gate to runway entrance time; sum of time spent in ramp area and taxi time to the runway entry.12. Aircraft type.For example, Boeing 747-400.13.Airline name.14.Scheduled gate departure time in hours, minutes and seconds.15.Gate departure delay.This is the difference between the actual gate departure time and the scheduled departure time.16.Number of departures on the surface at the actual gate departure time.These flights are out of the gate and moving towards departure runways.17.Number of arrivals on the surface at the actual gate departure time.These flights have landed and are moving towards arrival gates.18. Number of takeoffs from the same runway as this flight in the previous 15-minute interval with respect to the time of day.19.Average taxi-out delay of departures in seconds using the same runway as this flight in the previous 15minute interval with respect to the time of day.Taxi-out delay of each departure is computed as the difference of the actual taxi-out time and the unimpeded taxi-out time, where the unimpeded taxi-out time is computed as the ratio of the actual taxi-out distance to the average speed of 13 knots.20.Number of takeoffs that used the same departure fix as this flight in the previous 15-minute interval with respect to the time of day.21.Average taxi-out delay of departures in seconds using the same fix in the previous 15-minute interval with respect to the time of day.Taxi-out delay is computed in the same manner as in Item 19.22. Number of departures to the same destination airport in the previous 15-minute interval with respect to time of day.23.Average taxi-out delay of departures in seconds to the same destination airport in the previous 15-minute interval with respect to time of day.Taxi-out delay is computed in the same way as in Item 19.24.Number of departures from all gates in the previous15-minute interval with respect to the time of day.25.Number of takeoffs from all runways in the previous15-minute interval with respect to the time of day.26.Number of arrivals at all gates in the previous15-minute interval with respect to the time of day.27.Number of landings on all runways in the previous15-minute interval with respect to the time of day.28.Time of day in 15-minute interval such that the scheduled gate-out time is within this time interval.31.Average taxi-out delay of departures from the same runway as this flight in the previous 15-minute interval with respect to time of day.Taxi-out delay is computed in the same way as in Item 19.32.Average taxi-out delay of departures through the same departure fix in the previous 15-minute interval with respect to time of day.Taxi-out delay is computed in the same way as in Item 19.33.Average taxi-out delay of departures to the same destination airport in the previous 15-minute interval with respect to time of day.Taxi-out delay is computed in the same way as in Item 19.Note that the variables 29 through 33 are with reference to time of day related to the scheduled gate departure time.The other variables are with respect to the actual gate departure time.While some variables, such as 16 and 17, are with respect to the actual gate departure time, most variables are with respect to a broader interval of the time of day.Many variables, such as 1 and 3-15 are flight specific.Other variables like 16-23 and 29-33 are aggregate metrics based on other flights.Additionally, variables like 18-27 and 31-33 are based on aggregate metrics in the 15-minute time interval just prior to either the actual gate departure time or the scheduled gate departure time.It is assumed that these variables can be computed with flight plan data and Out-Off-On-In (OOOI) data provided by airlines available in the current air traffic system.Aggregate metrics 24-27 consider traffic to and from all gates and runways.Thus, they represent general state of airport operations.Time of day variables 2 and 28 are used for relating variables derived from the SMS logs and ASPM data.Variables based on ASPM data are discussed next.
+ASPM DataFederal Aviation Administration's (FAA) Aviation System Performance Metrics (ASPM) database, which is accessible on the web for authorized users, provides detailed data on flights to and from the 77 major U. S. airports (ASPM 77 airports) and flights operated by 29 major carriers (ASPM 29 carriers).Flights operated by ASPM carriers to international and domestic non-ASPM airports are also included.ASPM database also contains information on airport weather, runway configuration, and arrival and departure rates.Data in ASPM provide insight into air traffic and air carrier activity.FAA uses these data for monitoring airport efficiency, aspects of system performance, and retrospective trend analysis studies.Two different reports were extracted from the ASPM database for Dallas-Fort Worth airport operations spanning the period of 7 August 2011 through 13 August 2011.The first report, "Daily Weather by Quarter Hour Report," provided weather (visual meteorological condition or instrument meteorological condition), ceiling in feet, visibility in statute miles, temperature in degrees Fahrenheit, wind angle in degrees, wind speed in knots, arrival/departure runway configuration, airport departure rate and airport arrival rate in fifteen minute intervals as a function of local time.The second report, "Analysis By Airport By Quarter Hour Report (compared to flight plan)," includes numbers of scheduled departures/arrivals and departures/arrivals used for metric computation, percentages of on-time gate departures, airport departures and gate arrivals, average gate departure delay, average taxi-out time, average taxi-out delay, average airport departure delay, average taxi-in delay, and average gate arrival delay.Times and delays are in minutes.Numbers of scheduled arrivals and departures are based on carrier published schedules.Numbers of arrivals and departures for metric computation are based on itinerant flights to/from the ASPM 77 airports or operated by one of the ASPM 29 carriers.General aviation and military flights are excluded.Percent on-time gate departures is computed as the ratio of the number of flights that departed within 15-minutes past the flight plan gateout time to the number of departures for metric computation.Percent on-time airport departures is given as the ratio of the number of flights that departed within 15-minutes past the flight plan wheels-off time to the number of departures for metric computation.Percent on-time gate arrivals is determined as the ratio of the number of flights that arrive at the gate less than 15-minutes late compared to the flight plan gate-out time plus the scheduled block time to the total number of arrivals for metric computation.Taxi-out/taxi-in delay is the difference between taxiout/taxi-in time and unimpeded taxi-out/taxi-in time.Airport departure delay is computed as the difference between the actual wheels-off time and the sum of flight plan gate-out time and unimpeded taxi-out time.Average gate arrival delay is determined by adding minutes of gate arrival delay of one-minute or more, and dividing it by number of arrivals for metric computation.Gate arrival delay is defined as the difference between the actual gate-in time and the flight plan gate-in time.ASPM data from the two reports were processed to determine the following:1. Time of day in 15-minute intervals in hours and minutes format with respect to 00:00 local time.Figure 1 .1Figure 1.Gate to runway distance distribution.Figure 2. Average taxi speed distribution.
+Figure 2 .2Figure 1.Gate to runway distance distribution.Figure 2. Average taxi speed distribution.
+Figure 3 .3Figure 3. Neural network.
+Figure 4 .4Figure 4. Neural network convergence.
+Figure 7 .7Figure 7. Test set departure distribution.
+Figure 6 .6Figure 6.Test set data.Figure 5. Training error distribution.
+Figure 5 .5Figure 6.Test set data.Figure 5. Training error distribution.
+Figure 8 .8Figure 8. Test error distribution.
+Figure A- 1 .1Figure A-1.Dallas-Fort Worth airport layout.
+2 .2Meteorological condition-Visual Meteorological condition (VMC) or Instrument Meteorological Condition (IMC).3. Visibility in statute miles.4. Temperature in degrees Fahrenheit. 5. Wind angle in degrees.6. Wind speed in knots.7. Runway configuration indicating runways used for arrivals and runways used for departures.
+Table 1 .12011 runway usage summary.RunwaysRunwaysRunwaysUsed for bothUsageUsedUsageUsed onlyUsageArrivals and(hours)only for(hours)for(hours)DeparturesArrivalsDepartures18R6,96017L5,84435L2,55617R6,14213R5,62413L33517C6,11918L5,16736R2,56935C2,54731L2,53736L2,52635R2,31831R2,208Downloaded by NASA AMES RESEARCH CENTRE on April 17, 2013 | http://arc.aiaa.org| DOI: 10.2514/6.2012-5699
+Table 2 .22011top-ten runway configurations.ArrivalDepartureUsage (%)Flow13R,17C,17L,18R17R,18L45.4South31R,35C,35R,36L 31L,35L,36R22.7North13R,17C,17L,18R17R,18R9.7South17C,17L,18R17R,18L4.7South13R,17C,17L,18R 13L,17R,18L3.1South13R,17C,18R17R,18L2.9South35C,35R,36L31L,35L,36R2.9North31R,35C,36L31L,35L,36R1.5North13R,17C,17L17R,18R1.2South13R,17C,17L17R,18L0.8South
+Table 3 .3Gate to wheels-off correlations.#MetricCorrelation (%)p-value1 Gate to runway distance66.30.002 Average taxi-out delay in previous 15-minutes36.70.003 Number of departures on surface at actual gate departure time36.20.004 Average taxi-out delay of departures on same runway in previous 15-minutes23.50.005 Average taxi-out delay of departures to same fix in previous 15-minutes22.90.006 Gate departure count previous 15-minutes13.20.007 Number of departures to same fix in previous 15-minutes12.60.008 Wind angle-12.00.009 ATC set airport arrival rate9.70.00Number of departures on same runway in previous 15-minutes9.40.00Average taxi-in delay in previous 15-minutes8.30.00Wind speed on surface7.50.00Average taxi-out delay of departures to same destination in previous 15-minutes6.90.00ATC set airport departure rate6.00.00Number of departures to same destination in previous 15-minutes5.50.00Number of arrivals on surface at actual gate departure time5.30.00Visibility-4.40.00Average gate departure delay in previous 15-minutes3.80.00Takeoff count previous 15-minutes3.60.00Time of day at gate departure3.00.02Landing count previous 15-minutes2.50.05Temperature-2.20.09Gate arrival count previous 15-minutes2.00.13Average gate arrival delay in previous 15-minutes1.60.23Meteorological condition (VMC or IMC)1.30.33
+Table 4 .44. Note that the error is dimensionless because the training and neural network outputs are normalized.Six days, 7 August 2011 through 12 August 2011, of data were used for training the neural network and one day, 13 August 2011, of data were used for evaluating the gate to wheels-off time estimation ability of the neural network.Correlation between the gate to wheels-off time used for training and that generated by the neural network after training on the same set of neural network input data was found to be 72.4%.This is an improvement over gate to Gate departure delay correlations.#MetricCorrelation (%)p-value1Time of day at scheduled gate departure13.80.002Average gate departure delay in previous 15-minutes11.00.003Temperature10.10.004Average taxi-in delay in previous 15-minutes9.60.005Landing count previous 15-minutes8.90.006Gate arrival count previous 15-minutes7.70.007Number of arrivals on surface at scheduled gate departure time6.50.008Wind angle-5.30.009Average gate arrival delay in previous 15-minutes5.10.0010 Average taxi-out delay of departures to same destination in previous 15-minutes4.90.0011 Takeoff count previous 15-minutes4.80.0012 ATC set airport departure rate4.20.0013 Average taxi-out delay in previous 15-minutes3.50.0114 Number of departures on surface at scheduled gate departure time3.10.0315 Gate departure count previous 15-minutes3.00.0416 Wind speed on surface-1.20.4017 Visibility1.10.4318 Average taxi-out delay of departures on same runway in previous 15-minutes0.60.6619 ATC set airport arrival rate0.60.6920 Gate to runway distance0.50.7321 Average taxi-out delay of departures to same fix in previous 15-minutes-0.40.8022 Meteorological condition (VMC or IMC)-0.10.93
+Table 5 .5Linear model coefficients.Downloaded by NASA AMES RESEARCH CENTRE on April 17, 2013 | http://arc.aiaa.org| DOI: 10.2514/6.2012-5699CoefficientValue1 c241.18282 c10.00573 c10.15214 c0.10495 c0.03696 c-0.24917 c2.4563
+Time is given in the same format as Item 2 of this list.Observe that Item 2 is with reference to actual gate-out time while Item 28 is with respect to scheduled gate-out time.29.Number of departures on the surface at the scheduled gate departure time.Similar to Item 16 except at scheduled, not actual, gate departure time.30.Number of arrivals on the surface at the scheduled gate departure time.Similar to Item 17. by NASA AMES RESEARCH CENTRE on April 17, 2013 | http://arc.aiaa.org| DOI: 10.2514/6.2012-5699Downloaded
+			Downloaded by NASA AMES RESEARCH CENTRE on April 17, 2013 | http://arc.aiaa.org| DOI: 10.2514/6.2012-5699
+		
+		
+			
+			
+			
+
+				
+
+
+	
+		Improved Prediction of Gate Departure Times Using Pre-Departure Events
+		
+			LaraCook
+		
+		
+			StephenAtkins
+		
+		
+			YoonJung
+		
+		10.2514/6.2008-8919
+	
+	
+		The 26th Congress of ICAS and 8th AIAA ATIO
+		Anchorage, Alaska
+		
+			American Institute of Aeronautics and Astronautics
+			September 14-19, 2008
+		
+	
+	AIAA 2008-8919
+	Cook, L. S., Atkins, S., and Jung, Y., "Improved Prediction of Gate Departure Times Using Pre-Departure Events," AIAA 2008-8919, Proc. 26 th Congress of International Council of the Aeronautical Sciences (ICAS), Anchorage, Alaska, September 14-19, 2008.
+
+
+
+	
+		Aircraft Taxi Times at U.S. Domestic Airports
+		
+			DerekRobinson
+		
+		
+			DanielMurphy
+		
+		10.2514/6.2010-9147
+	
+	
+		10th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference
+		Fort Worth, TX
+		
+			American Institute of Aeronautics and Astronautics
+			September 13-15, 2010
+		
+	
+	Aircraft Taxi Times at U. S. Domestic Airports
+	Robinson, D. P., and Murphy, D. J., "Aircraft Taxi Times at U. S. Domestic Airports," AIAA 2010-9147, Proc. 10th AIAA Aviation Technology, Integration and Operations Conference (ATIO), Fort Worth, TX, September 13-15, 2010.
+
+
+
+	
+		An Analytical Queuing Model of Airport Departure Processes for Taxi Out Time Prediction
+		
+			IoannisSimaiakis
+		
+		
+			NikolasPyrgiotis
+		
+		10.2514/6.2010-9148
+	
+	
+		10th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference
+		Fort Worth, TX
+		
+			American Institute of Aeronautics and Astronautics
+			September 13-15, 2010
+		
+	
+	Simaiakis, I., and Pyrgiotis, N., "An Analytical Queuing Model of Airport Departure Process for Taxi Out Time Prediction," AIAA 2010-9148, Proc. 10th AIAA Aviation Technology, Integration and Operations Conference (ATIO), Fort Worth, TX, September 13-15, 2010.
+
+
+
+	
+		Quantitative Analysis of Uncertainty in Airport Surface Operations
+		
+			DavidRappaport
+		
+		
+			PeterYu
+		
+		
+			KatyGriffin
+		
+		
+			ChrisDaviau
+		
+		10.2514/6.2009-6987
+	
+	
+		9th AIAA Aviation Technology, Integration, and Operations Conference (ATIO)
+		Hilton Head, SC
+		
+			American Institute of Aeronautics and Astronautics
+			September 21-23, 2009
+		
+	
+	Rappaport, D. B., Yu, P., Griffin, K., and Daviau, C., "Quantitative Analysis of Uncertainty in Airport Surface Operations," AIAA 2009-6987, Proc. 9th AIAA Aviation Technology, Integration and Operations Conference (ATIO), Hilton Head, SC, September 21-23, 2009.
+
+
+
+	
+		Relationship Between Airport Efficiency and Surface Traffic
+		
+			MatthewKistler
+		
+		
+			GautamGupta
+		
+		10.2514/6.2009-7078
+	
+	
+		9th AIAA Aviation Technology, Integration, and Operations Conference (ATIO)
+		Hilton Head, SC
+		
+			American Institute of Aeronautics and Astronautics
+			September 21-23, 2009
+		
+	
+	Kistler, M. S., and Gupta, G., "Relationship between Airport Efficiency and Surface Traffic," AIAA 2009-7078, Proc. 9th AIAA Aviation Technology, Integration and Operations Conference (ATIO), Hilton Head, SC, September 21-23, 2009.
+
+
+
+	
+		Air Traffic Management System Development and Integration (ATMSDI) Acquisition CTO-05--Surface Management System CTOD-2--Airport Site Surveys
+		
+			AtmsdiRaytheon
+		
+		
+			Team
+		
+		NAS2-00015
+	
+	
+		Contract Number
+		Moffett Field, CA
+		
+			June 5, 2001
+			
+		
+		
+			NASA Ames Research Center
+		
+	
+	Raytheon ATMSDI Team, "Air Traffic Management System Development and Integration (ATMSDI) Acquisition CTO-05- -Surface Management System CTOD-2--Airport Site Surveys," Contract Number NAS2-00015, NASA Ames Research Center, Moffett Field, CA 94035-1000, June 5, 2001.
+
+
+
+	
+		Foreword from Acting Administrator of the Health Resources and Services Administration (HRSA) and Associate Administrator of the Federal Office of Rural Health Policy (FORHP), U.S. Department of Health and Human Services
+		
+			JimMacrae
+		
+		
+			TomMorris
+		
+		10.1353/hpu.2016.0195
+		
+	
+	
+		Journal of Health Care for the Poor and Underserved
+		Journal of Health Care for the Poor and Underserved
+		1548-6869
+		
+			27
+			4A
+			
+			March 2011
+			Project MUSE
+		
+	
+	Administrator's Fact Book. cited: 4/24/2012
+	Assistant Administrator for Financial Services, "Administrator's Fact Book," Federal Aviation Administration, U. S. Department of Transportation, March 2011, URL: http://www.faa.gov/about/office_org/headquarters_offices/aba/admin_factbook/media/201103.pdf [cited: 4/24/2012].
+
+
+
+	
+		Characterization of Tactical Departure Scheduling in the National Airspace System
+		
+			RichardCapps
+		
+		
+			ShawnEngelland
+		
+		10.2514/6.2011-6835
+	
+	
+		11th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference
+		Virginia Beach, VA
+		
+			American Institute of Aeronautics and Astronautics
+			September 20-22, 2011
+		
+	
+	AIAA 2011-6835
+	Capps, A., and Engelland, S. A., "Characterization of Tactical Departure Scheduling in the National Airspace System," AIAA 2011-6835, Proc. 11th AIAA Aviation Technology, Integration and Operations Conference (ATIO), Virginia Beach, VA, September 20-22, 2011.
+
+
+
+	
+		Observations of Departure Processes at Logan Airport to Support the Development of Departure Planning Tools
+		
+			HusniRIdris
+		
+		
+			IoannisAnagnostakis
+		
+		
+			BertrandDelcaire
+		
+		
+			RJohnHansman
+		
+		
+			John-PaulClarke
+		
+		
+			EricFeron
+		
+		
+			AmedeoROdoni
+		
+		10.2514/atcq.7.4.229
+	
+	
+		Air Traffic Control Quarterly
+		Air Traffic Control Quarterly
+		1064-3818
+		2472-5757
+		
+			7
+			4
+			
+			
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	15-minute airport departure rate set by air traffic control.
+
+
+
+	
+		Estimating One-Parameter Airport Arrival Capacity Distributions for Air Traffic Flow Management
+		
+			TashaRInniss
+		
+		
+			MichaelOBall
+		
+		10.2514/atcq.12.3.223
+	
+	
+		Air Traffic Control Quarterly
+		Air Traffic Control Quarterly
+		1064-3818
+		2472-5757
+		
+			12
+			3
+			
+			
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	15-minute airport arrival rate set by air traffic control.
+
+
+
+	
+		Supplementary file 1.
+		10.7554/elife.04535.023
+		
+			null
+			eLife Sciences Publications, Ltd
+			11
+		
+	
+	Average gate departure delay in the previous 15-minute interval with respect to the time of day
+	Average gate departure delay in the previous 15-minute interval with respect to the time of day. 11. Average taxi-out delay in the previous 15-minute interval with respect to the time of day. 12. Average taxi-in delay in the previous 15-minute interval with respect to the time of day. 13. Average gate arrival delay in the previous 15-minute interval with respect to the time of day. The first 9 items are measured and directly available at the tower. The remaining four items can be calculated based on flight plan and OOOI data.
+
+
+				
+			
+		
+	

file130.txt

@@ -0,0 +1,496 @@
+
+	
+	
+		
+I. Introductionn the current national airspace system, design of sectors have evolved over a long period of time based on incremental addition of new technologies and procedures for air traffic control.Each sector has a fixed capacity.When these capacities are exceeded by traffic demand, traffic flows are restriced to bring the demand below capacity.The concept in Ref. 1 suggests that instead of restricting traffic, which causes delays, airspace capacity can be increased by partitioning the airspace differently.Motivated by this concept, several methods for airspace partitioning that are described in Refs. 2 through 6 have been developed.These methods use some measure of controller workload to guide the design.In the future, with increased level of automation, airspace design might not be guided by controller workload considerations.Depending on how different the future design is from the current design, the controller's ability to actively separate aircraft might be limited.It might be useful to carry some of the design features of the current system into the future one, if some role for human controller is envisioned in the future air traffic control system.The motivation for computing metrics for the existing sectors is to capture some of the design features of the current sectors.Since the design of current sectors is based on the routes of flight and controller workload considerations, metrics related to controller workload can be expected to capure the design features.][9][10][11] These studies are limited to sectors in few centers.A comprehensive study of sectors in all the twenty centers is unavailable.In this paper, thirty-three traffic and geometric metrics from Refs. 7 to 11 are computed for 364 higher altitude sectors in each of the twenty centers, and in eight geographical regions.Higher altitude sectors were chosen because the benefits of airspace partitioning are expected to be realized in these sectors first.Data presented in this paper describes the design of the current sectors and will be found to be useful for comparing the designs of future airspace partitions.The paper is organized as follows.Section II describes the method for computing the metrics for each sector.The maximum number of aircraft in sectors obtained using simulated track-data are compared with those obtained using actual track-data in this section.Evaluation of sectors based on the traffic metrics is described in Section III and on geometric metrics in Section IV.Finally, the paper is summarized in Section V.
+II. Computational MethodThis section describes two methods for computing the fifteen sector traffic metrics as a function of time.The first method processes the actual track-data, and the second method processes simulated traffic data.The main benefit of using actual track-data is that they best represent traffic resulting from the demand and capacity constraints of the particular day.These data reflect flow control and separation assurance actions, and airline operational control actions such as: cancellations and creation of new flights.Along with their benefits, they suffer from some data-quality issues discussed in Ref. 12 that can lead to erroneous values of the metrics.Simulated traffic data do not suffer from many of the data quality issues, but they do not include all of the realworld effects.Trajectories of aircraft are simulated using mathematical models; they can be expected to differ from those actually flown.Simulated traffic data have to be compared against actual traffic data for validation.These limitations not withstanding, a simulation offers far greater flexibility in designing scenarios, including future traffic growth scenarios, for evaluating the desired metrics.Simulation also provides a mechanism for eliminating control actions inherent in the real system.For example, aircraft are permitted to violate separation minimums, which only happen as an operational error in the real system.Given a choice between using actual position data, "track-data," and simulated data, it is often desirable to use the actual track-data because they represent reality.Due to this reason, ten of the fifteen measures have been computed using the actual track-data derived from a recorded Airline Situation Display to Industry (ASDI) file.The remaining five metrics, number of jet aircraft, number of non-jet aircraft, conflict-count, average airspeed and variance in airspeed, were computed using position data simulated using the Airspace Concept Evaluation System (ACES) with flight-plans extracted from the same ASDI file.The reason for using simulated data is that the information needed for computing these metrics is provided in the flight plan and not in the actual track-data.
+A. Track-DataThe ASDI subsystem of the Enhanced Traffic Management System (ETMS) 13 disseminates real-time air traffic data, associated with different message types, to aviation industry.Of the different ASDI messages and supported data types described in Ref. 13, only the track/flight data-block messages, TZ messages, were used for computing the traffic metrics.TZ messages contain time-stamp, ARTCC identifier, aircraft identification (ACID), groundspeed, altitude, latitude and longitude.Over 9 million TZ messages were extracted from the recorded ASDI file, containing 48-hours of traffic data spanning the period from zero Coordinated Universal Time (UTC) on 17 March 2006 to zero UTC on 19 March 2006.Traffic data for these days were selected because of high traffic-volume, low weather impact and low delays.It should be noted that the traffic patterns vary from day to day and from season to season depending on demand, capacity, wind patterns and weather.Numerical values of the metrics in the sectors can be expected to vary based on the traffic pattern.The results presented in this paper are for a nominal day, which is defined as a high-volume, low-weather, and low-delay day.The time-sorted TZ messages were stored in data structures based on ACID.Data structures were then examined for each ACID to determine all the flights associated with that ACID.Individual flights were identified based on temporal gaps in the associated time-stamps.A temporal spacing of over thirty minutes was assumed to be due to a different flight.After identifying each flight, duplicate messages within one-minute periods were removed.This reduced the data by about 15%.Each flight's latitude, longitude and altitude derived from TZ messages were used to determine the sector the flight was in at each instant of time, where the sectors were as defined in May 2007 ETMS adaptation data.An efficient procedure for locating aircraft in a sector, described in Ref. 14, was used to identify the sectors.The latitudes and longitudes were then transformed into Cartesian coordinates with respect to a horizontal frame of reference using Oblique Stereographic Projection. 2 A first-order filter with altitude time-history input was used to generate time-history of climb/descent rate of each flight.Time, Flight ID, climb/descent rate, position and the sector corresponding to the position were input into a MySQL 15 database.This database was then queried to extract data related to flights in each sector and to sort them in time.These time-sorted data were then used to determine the numbers of aircraft in climb, in cruise, and in descent, and their sum (total number of aircraft) in the sectors.The data and the results are for 364 higher altitude sectors shown in Fig. 1.Each of these sectors has a floor of 17,100 feet altitude or higher.Fifty of the sectors have a floor between 17,100 feet and 24,000 feet; the rest have a floor at or above 24,000 feet altitude.The ceilings of these sectors are between 24,000 feet and 99,900 feet altitude.Several sectors in Salt Lake, Minneapolis, New York, DC, Boston and Los Angeles centers (see Fig. 1) were not included in the analysis because their floors are below 17,100 feet altitude.
+B. ACES SimulationACES is a comprehensive computational model of the national airspace system consisting of air traffic control and traffic flow management models of ARTCCs, terminal-radar-approach-controls (TRACON), airports and the air-traffic-control-system-command-center (ATCSCC). 16It simulates flight trajectories through the enroute-phase of flight, where enroute-phase for jet aircraft is above 10,000 feet.A queuing model simulates surface movement and flight through terminal airspace.Traffic flow management and air traffic control models in ACES use airport and sector capacity thresholds for delaying flights, while they are on the ground and during their enroutephase.Some of the ACES outputs are arrival and departure counts at airports, trafficcounts in sectors and air traffic system performance metrics including arrival, departure, enroute and total delays.Earlier validation studies in Refs.17 and 18 have shown that ACES generates realistic delays and airport operational metrics similar to those observed in the real-world.Due to these capabilities, ACES was chosen for simulating traffic.
+Simulation InputsACES simulation inputs include files containing capacity data (airport arrival and departure capacities, and sector capacities), traffic data (scheduled departure times and flight-plans), and adaptation data (sector/center geometric data).The actual airport arrival and departure rates specified at the 74 major U. S. airports on 17th and 18th March 2006 were specified as airport capacities in ACES.Sector capacity data were derived from the ETMS data tables.Sector capacity is defined as the maximum number of aircraft allowed in a sector at any one time during a fifteen-minute time interval.Capacity values, known as Monitor Alert Parameter (MAP), are used in the ETMS to trigger traffic flow management initiatives for demand reduction.Capacity thresholds are set to ensure that air traffic controllers are able to separate aircraft traversing the sector airspace.MAP values and the number of sectors with those values out of 364 higher altitude   1.Flight-plans for the simulation were derived from the same ASDI file used for computing the traffic-count metrics discussed in the previous subsection.Flight connectivity data relating the same physical aircraft to two or more flights segments were obtained from the Bureau of Transportation Statistics (BTS) for the two days.Airline flight-numbers, aircraft tail-numbers and the associated flight-plans for all flights were included in the ACES input file.Scheduled departure times derived from the BTS data were assigned as departure times.Proposed departure times from flight-plan messages in the ASDI data, or actual departure times in ASDI data minus average taxi times associated with airports of departure were assigned as departure times when scheduled times were not available in the BTS data.After assigning scheduled departure times for flights, an ACES simulation was run without airport and sector capacity constraints to compute unconstrained arrival times of flights at their destination airports.These were then set to scheduled arrival times at destination airports.A series of steps described in Ref. 19 were then taken to ensure that flight connectivity was preserved and that the arrival and departure schedules were compatible with turnaround-time requirements, time required for unloading aircraft after arrival at the gate and preparing it for departure.Final traffic schedule was generated after making the required flight schedules and tail-number changes.Sector and center geometry definitions needed for the simulation were obtained from the May 2007 ETMS adaptation data.
+Simulation OutputsACES writes out identification information (ID) and position coordinates of flights that are in each sector to an output database at the specified rate during simulation.After completion of the simulation, flight IDs associated with a sector and their position time-histories were extracted from the database for computation of the five traffic metrics, number of jet aircraft, number of non-jet aircraft, conflictcount, average airspeed and variance in airspeed.This process was repeated for every sector.
+C. Comparison of Tack-Data and ACES Simulation ResultsGiven that there are differences between the actual traffic data from the field and ACES simulated data, it is necessary that a comparison between the values of metrics obtained by processing actual track-data and ACES simulation be done for establishing the validity of the results.This was accomplished by comparing 1) the total number of aircraft in higher altitude airspace and 2) numbers of higher altitude sectors grouped according to the peak total-counts (trafficcounts), peak climb-counts, peak cruise-counts and peak descent-counts in each hour simulated by ACES with those obtained by processing track messages.Numbers of aircraft in sectors shown in Fig. 1 were retrieved from the simulation output and added together to compute the total number of aircraft in ten-second intervals.This time-history is shown with that of the actual number of flights in Fig. 2. Observe that the ACES simulation starts with all aircraft on the ground, whereas in the actual air traffic system there are always flights that are airborne.Figure 2 shows that the simulated traffic catches up with the actual traffic.The general trends of simulated and actual traffic are similar for the twenty-four hours between eight UTC on 17 March 2006 and eight UTC on 18 March 2006 (location marked 32 UTC in Fig. 2).Some of the differences between the time-histories are attributable to the issues with simulated and actual flight data discussed earlier in this section.Figure 3. Sectors with peak traffic-count in the first five levels, listed in Table 2, using actual track-data.For comparison based on peak traffic-counts, aircraft were counted in each higher altitude sector.The maximum of 60 values, one for each minute, provided the maximum number of aircraft, peak trafficcount, using actual track data for that hour.Figures 3 and4 show the number of sectors with peak traffic-count values in the first five levels listed in Table 2 during each one-hour period.The graphs in the figures viewed in conjunction with Fig. 2 show that as traffic increases, there are fewer sectors with low peak traffic-counts as would be expected.The maximum numbers of sectors for the four graphs, Level 2 through 5, in Fig. 3 were found to be 217 at 28:00 UTC, 190 at 22:00 UTC, 85 at 23:00 UTC and 12 at 18:00 UTC.In Fig. 4 these are 222 at 29:00 UTC, 176 at 24:00 UTC, 55 at 23:00 UTC and 12 at 17:00 UTC.The information contained in the individual graphs in Figs. 3 and4 can be combined and presented in terms of a vertically stacked bar chart for each one-hour period of the day.Such bar charts provide cumulative information as explained by the following example.Consider the bar charts obtained by processing the actual trackdata shown in Fig. 5.Of the eight levels listed in Table 2 and included in Fig. 5, only levels one through five are visible in the bar charts.Numbers of sectors with a peak traffic-count of four aircraft or less are shown in the bottom bar charts.Numbers of sectors with the next higher level are placed on top of this layer, and so on.Cumulative counts obtained by summing the levels below each chosen level determine the numbers of sectors with peak traffic-counts below the thresholds implicitly defined by levels in Table 2.For example, the top of Level 2 histogram at 23:00 UTC indicates that at any time of the day, at least 99 sectors have a peak traffic-count of nine aircraft or less.This is defined as the lower cumulative count.The maximum value of the top of the Level 2 histogram is 364 sectors at 10:00 UTC.This maximum value is defined as upper cumulative count.Table 3 summarizes these results obtained using actual track-data and ACES simulation.The time-histories shown in Figs. 3 and4 compare reasonably well; they show that similar peak traffic-counts are obtained in approximately the same number of higher altitude sectors using actual track-data and ACES simulated track-data.Values listed in the second and third  2, using ACES data.Figure 5. Time-history of bar charts of number of sectors grouped in eight peak traffic-count levels using track-data.Table 3. Numbers of sectors with peak traffic-counts below thresholds using ACES simulation compared to those obtained using track-data.3 also suggest that the peak traffic-count statistics computed using actual track-data and ACES are comparable.Analysis of peak climb-counts, peak cruise-counts and peak descent-counts using actual track-data and ACES simulated data also compared well.
+III. Traffic MetricsA considerable amount of research has been devoted to the synthesis of traffic dependent metrics and their application to modeling sector complexity and air traffic controller's workload. 8,9Traffic metrics that were found to be especially pertinent for modeling workload perceived by controller in Ref. 8 are listed in Table 4.These metrics were computed for today's higher altitude sectors shown in Fig. 1 and the results are presented in the following subsections.
+A. Traffic-count MetricsPeak traffic-count or the maximum number of aircraft has been found to be the most important contributor to controller workload.Research has found that controller activity and attentiveness are highly correlated to peak traffic-count. 10Maximum number of aircraft in higher altitude sectors computed using actual track-data were discussed in Section II.The time histories of the numbers of sectors with peak traffic-count levels were shown in Fig. 3 and the corresponding bar charts were shown in Fig. 5.Maximum numbers of aircraft in the climb phase (peak climb-count), in the cruise phase (peak cruise-count), and in the descent phase (peak descent-count) were computed in the same manner as the peak traffic-count in a sector.Aircraft with a climb rate of 200 feet/minute or more were considered to be climbing and those with a descent rate greater than or equal to 200 feet/minute were considered to be descending.Aircraft with climb or descent rates less than 200 feet/minute were considered to be cruising.Maximum number of jet aircraft and non-jet aircraft (turboprops and piston-props) in each hour in every sector were computed using ACES simulated track-data at ten second intervals.Since actual track-data do not contain aircraft type information, flight plans do, ACES simulated track-data that are based on flight-plans were used.Although it is possible to relate actual track-data to the flight-plans using the aircraft IDs that are common to both the message types, it does require an efficient data structure, significant memory and computation.ACES simulated data were also used for computing conflict-counts.Since jet aircraft cruise at approximately 8 miles/minute, trajectory data generated by ACES were written to the output database at tensecond intervals.Aircraft within each sector were selected from the database for checking conflicts, which means that conflicts between aircraft in neighboring sectors were not checked.This limitation arose because the database did not contain information on which sector is next to which one.Not knowing the sector neighborhood relationship meant that the only way to check for all such conflicts was to consider all the flights at each time-  3).Cumulative counts out of 364 counts listed in the last two columns are based on the time history of bar charts of numbers of sectors in peak climb-count levels like the one shown in Fig. 5. Cumulative counts were discussed earlier to explain the contents of Table 3.The value of 364 in the sixth and seventh columns indicates that at no time were there more than 14 aircraft in climb in a higher altitude sector.Tables 6 through 10 should be interpreted in the same way as Table 5. Observe that the thresholds in Tables 9 and 10 are not the ones in Table 2. Since there are far fewer non-jet aircraft and conflicts compared to the number of aircraft in a sector, the thresholds for non-jet aircraft counts and conflict-counts have a smaller range compared to those in Table 2.The time to conflict is computed as the ratio of the range to the range-rate (time derivative of range).The average time-to-go, defined as C12 in Ref. 8, considers each aircraft in the sector one at a time for determining the aircraft with which it is predicted to conflict.Next, the time-to-go is used for identifying those aircraft (conflict set) with which conflict is predicted in the near term (for example, in less than ten minutes).Minimum time-to-go is then determined for each conflict set, summed and divided by the number of sets to compute the average.Minimum values of the 1) two horizontal separation metrics, 2) two vertical separation metrics, and 3) one timeto-go metric were computed for each hour using actual track-data.Time histories and historgrams of sectors were created in the same manner as discussed for the traffic-count metrics.The trends in these data are summarized in Tables 11 through 15.The category no-conflict in Tables 11 and12 means that aircraft in the sector were outside each others vertical bounds, therefore horizontal separation was not computed for them.Horizontal separation is measured in nauticalmiles and the vertical separation is measured in feet.Time-to-go is measured in seconds.The categories no-conflict in Tables 13 and14 means that aircraft were outside each others horizontal bounds, and 'at most one aircraft' in
+C. Flow MetricsThe three flow metrics discussed in this section are, average transit-time, average airspeed and the variance in airspeed of aircraft in the sector.Sector transit-time is defined as the time taken by the aircraft to cross the sector.The difference between the entry and exit times, obtained by processing the actual track-data provided the transit-time.Average transit-time for each sector was obtained by considering the transittime of all the aircraft that went though the sector during the 24-hour period.Figure 6 shows the histogram of the average transit-time in minutes of 364 higher altitude sectors.The minimum, mean, standard-deviation, and maximum values of the distribution shown in Fig. 6 were found to be 2.8 minutes, 8 minutes, 2.8 minutes and 21 minutes.Average airspeed was computed every ten-second in each sector using the ACES simulated data.Maximum average-airspeed within each hour was computed with 360 such values.Time histories and bar charts of sectors were then created with the maximum average-airspeed values.Results are summarized in Fig. 7 and Table 16.Observe that airspeed of higher altitude traffic generally lies in the 400 knots to 500 knots range.Like the average airspeed, the variance of airspeed (knots 2 ) was computed at ten-second intervals and then the maximum value was selected for each hour for each sector.Variance of airspeed is defined in Eq. 26 in Ref. 8. Time history of the bar charts of sectors with the maximum airspeed-variance in six-levels is shown in Fig. 8. Table 17 lists the five-levels, minimum, maximum, average, and maximum number of sectors at each of the five-levels, and the cumulative lower and upper bounds obtained using the bar charts in Fig. 8.The computations done with airspeed could have also been done with groundspeed, which is available in the actual track-data, but it was not used because groundspeed depends on winds, which vary from day to day.Airspeed on the other hand is a function of aircraft performance characteristics and is independent of winds.
+IV. Geometric MetricsGeometric metrics are described in this section.Sector geometry features such as airways, navigational aids and airports define the kind and the frequency of tasks performed by the controller.Therefore, they contribute to controller workload directly and operational errors indirectly. 10,11Sector geometry metrics, including those proposed in Refs. 10 and 11, considered in this study are listed in Table 18.Data resulting from computations on higher altitude sectors are discussed in the subsections below.Geometric metrics listed in Table 18 are grouped into five categories.The first category "Geographical Location" consists of metrics one through three.The second category "Sector Dimensions" consists of metrics four through seven.The third category "Shape Attributes" consists of metrics eight through ten.The fourth category "Route Attributes" comprises of metrics 11 through 15.The last category "NeighborhoodAttributes" contains metrics 16 through 18.
+A. Geographical LocationThe number of higher altitude sectors in each center was determined from the first three characters of Sector IDs of sectors with a base of 17,100 feet and above.For example, "ZAB37" indicates that the sector belongs to the Albuquerque center (ZAB).Counting all these sectors shown in Fig. 1 with the same first three letters provided the number of sectors in each center shown in Table 19.The 20 centers listed in Table 19 are organized in eight geographical regions listed in Table 20.The number of sectors in each region is obtained by summing the values corresponding to the centers that form the region.These results are given in the fourth column of Table 20.The numbers of lowaltitude, high-altitude and super-high-altitude sectors as defined in May 2007 ETMS adaptation data were found to be five, 248 and 111, respectively.The five low-altitude sectors had a base at or above 17,100 feet altitude.
+B. Sector DimensionsTo determine the dimensions of each sector, the volume of each subsector was computed.A subsector is defined as a polygonal prism with a boundary defined by a polygon and a constant height.The complex geometric shape of a sector is achieved by placing subsectors on top, and to the side of other subsectors.The volume of a sector is the sum of the volumes of its subsectors.Figure 9 shows the histogram of the volume of sectors, in cubic-nautical-miles.Height of each sector was determined by subtracting the lower bound of the lowest subsector from the upper bound of the highest subsector.The distribution of the heights of the sectors is shown in Fig. 10.The histogram in Fig. 10 shows a bi-modal distribution with sector below 16,000 feet and above 56,000 feet.Dividing sector volumes with sector heights resulted in the reference-areas of the sectors.The distribution of the sector reference-areas is shown in Fig. 11.Reference-lengths of the sectors were obtained by taking the square-root of the reference-areas.The histogram of the reference-lengths is given in Fig. 12.Minimum, mean, maximum and standard deviation values of sector volume, height, reference-area and reference-length are summarized in Table 21.
+C. Shape AttributesAspect-ratio, the ratio of the length to width, of the sectors was determined by computing the moments of inertia of the sectors.Prior to the computation of moments of inertia, centroids of the sectors had to be computed.Since a sector is composed of subsectors, the centroid of each subsector was determined first, and then, these centroids were weighted with the volumes of the subsectors to obtain the centroid of the sector.The moment of inertia tensor of  each subsector was computed with respect to the frame of reference located at the centroids, and then the parallel axis theorem was employed to determine the moment of inertia tensor with respect to the frame of reference located at the centroid of the sector.These moment of inertia tensors were then summed up to determine the moment of inertia tensor of the sector.Mathematical expressions for computing the centroid and the moment of inertia tensor for a two-dimensional polygonal object are given in Ref. 20.These equations were extended to three-dimensional polygonal prisms for generating the results discussed here.The eigenvalues of the moment of inertia tensor are the principal moments of inertia about the principal axes, which are the eigenvectors corresponding to the eigenvalues.If 11 I , 22 I and 33 I are the principal moments of inertia, the dimensions of a rectangular prism with the same principal moments of inertia as the sector (polygonal prism) are:( )3 1 ; 12 ! ! " = i I S V l ii i (1)where, ( )33 22 11 2 1 I I I S + + =and V is the volume of the sector.The aspect-ratio is given by the ratio of dimensions in Eq. (1), j i l l / such that j i ! .Figure 13 shows the distribution of the aspect-ratio of the 364 sectors.Sectors with an aspect-ratio close to one are of square shape, while the sector with an aspect-ratio closer to seven is a highly elongated rectangle.The minimum, mean, maximum and standard deviation aspect-ratio are 1.0, 2.0, 6.6, and 0.86, respectively.Reference 7 notes that the traffic pattern is usually highly parallel and less complicated in elongated sectors.If the principal moments of inertia are distinct, the principal axes are uniquely specified.If two or all three of the principal moments are the same, there is no choice of a preferred axis.Once the three principal axes were obtained, the two principal axes with larger projections on the horizontal plane were selected.Of these, the one with the larger eigenvalue was chosen as the preferred axis of the sector. Figure 14 shows the preferred axis of the sectors.This figure shows that most of the sectors are aligned along East-West direction.Along the East and West Coasts, sectors are oriented in the North-South direction.It is interesting that the sectors are aligned along the major traffic flow directions.The number of subsectors in a sector is an indicator of the shape complexity of the sector.Table 22 lists the number of sectors containing the same number of subsectors.This table shows that most of the sectors have a single subsector.One sector has 13 subsectors.The mean and the standard deviation of the number of subsectors were found to be 1.7 and 1.4.
+D. Route AttributesThree types of navigation aids were counted to determine the number of navigation aids in sectors.These three types of radio navigation aids commonly employed by aircraft for navigation along routes are Very-High-Frequency Omnidirectional Range (VOR), VOR collocated with Distance Measuring Equipment (VOR-DME), and VOR collocated with Tactical Air Navigation System (VORTAC).The known latitude and longitude of each navigation aid were used for locating it in the sector. Figure 15 shows the distribution of sectors as a function of the number of navigation aids enclosed within their boundaries.There are a total of 1055 navigation aids.Some of these are shared by multiple sectors.Six sectors had no navigation aids and one sector had a maximum of 14 navigation aids.The mean and standard deviation of the number of navaids in a sector is 3.9 and 2.3.Next, the number of intersections in sectors was computed.Intersections are defined as locations where airways, Victor Airways and Jet Routes, intersect.Intersections are specified by latitude, longitude and altitude.This makes it possible to locate them in sectors.Figure 16 shows the histogram of sectors based on the number of intersections enclosed within the sectors.Sixty-four sectors had no intersections, 328 sectors had ten or fewer intersections, 36 sectors had more than 10 intersections, and one sector had 22 intersections.Mean and standard deviation were determined to be 4.3 and 4.1 intersections.Airways in sectors were determined using the association between the intersections and airways.Since an airway can be associated with several intersections within a sector, each associated airway is counted only once.An airway that went across the sector without passing through an intersection in the sector could not be counted.It may be possible to improve the airway count by using the association between the navaids and airways in addition to that between intersections and airways.Victor Airways were counted below 18,000 feet altitude and Jet Routes at or above 18,000 feet altitude.Both were counted in sectors whose base was below 18,000 feet and top above 18,000 feet altitude.A histogram of sectors as a function of the number of airways is given in Fig. 17.Airways were not found in sixty-four sectors because those sectors did not have any intersections.334 sectors had ten or fewer airways while 30 sectors had more than ten airways.The maximum number of airways was found to be 17 in only one sector.Mean and standard deviation values were determined to be 5 and 3.7 airways.Seventy-four major airports in the United States that are in the Federal Aviation Administration's (FAA) Aviation System Performance Metrics (ASPM) 21 were located within the horizontal confines of the higher altitude sectors and counted.Table 23 lists the numbers of sectors and the corresponding numbers of 74 ASPM airports.Eightythree sectors had one or more major U. S. airports.Only 16 sectors had two or more airports.Controllers have to ensure that aircraft do not enter a Special Use Airspace (SUA).This monitoring function adds to controller workload.Prohibited areas, military operations areas, alert areas, warning areas, and national security areas are considered to be SUA.Boundary and height information of the 1017 SUAs obtained from the FAA were used for locating them in sectors.Of the 364 sectors, only two sectors, ZHU 59 and ZDV 47 were found to completely contain one and two SUAs, respectively.The rest partially contained the SUAs.Statistics of sectors partially containing SUAs are discussed in the next subsection.
+E. Neighborhood AttributesSectors above, below and to the sides of sectors were counted to determine the number of surrounding sectors.Minimum, mean, maximum and standard deviation of the number of surrounding sectors were determined to be 4, 13.3, 30 and 3.9.Sector 22 in Atlanta Center was found to have 30 sectors surrounding it.There were 202 sectors that had 13 or fewer sectors surrounding them, while 162 sectors had more than 13 sectors surrounding them.The distribution of the sectors as a function of number of surrounding sectors is given in Fig. 18.SUA polygons were related to a data structure of grid cells using the method described in Ref. 14.The mapping of the grid cells to the sectors was then used to identify the set of sectors that the SUAs could be in.The altitude range of the sectors was compared with the altitude range of the SUAs to determine if there was any overlap.Sectors with overlap were deemed to partially or fully contain these SUAs.If a SUA was associated with a single sector, that sector was considered to completely contain that SUA.In instances where the SUA was associated with more than one sector, the SUA was considered to be partially contained in the associated sectors.With the elimination of two sectors that fully contained SUAs, analysis was done on the remaining 362 sectors.SUAs were not found in 268 sectors.Eighty-one sectors were found to have five or fewer SUAs.Only 15 sectors were found to have more than five SUAs.Sector 26 in Houston Center contained 19 SUAs.Table 24 summarizes these results.Distance with respect to the center of the sector to the closest one of the 74 ASPM airports outside the sector was determined by first establishing which of them were within the confines of the sector boundary.These airports were excluded and distances to the airports outside the sector boundary were computed.The minimum of these distances gave the distance to the closest airport outside the sector. Figure 19 gives the histogram of the sectors with respect to distances to the closest airports.The closest airport outside a sector was 20.6 nautical-miles.The average and maximum distances were 116.5 and 422.2 nautical-miles.The standard deviation was 57.8 nautical-miles.A major airport was within 200 nautical-miles from the 335 sectors (both inside and outside the sector).Only 29 sectors were farther than 200 nautical-miles from a major airport.
+V. SummaryThis paper was motivated by the problem of determining the design features of the current sectors so that future designs can be compared with the current design.Since the current design is based on tools and techniques used by controllers, a large departure from this design guided by automation needs will have implications in controllers being able to manage traffic in the new sectors.Since this study is focused on current sectors, fifteen traffic metrics related and eighteen geometric metrics related to controller workload were used to characterize the design of current sectors.Numerical values of these metrics were computed for sectors with a base of 17,100 feet and above in the current U. S. airspace.The fifteen traffic metrics were classified into three categories: seven traffic-count metrics, five separation metrics and three flow metrics.The eighteen geometric metrics were classified into five categories: three geographical location metrics, four sector dimension metrics, three shape attribute metrics, five route attribute metrics and three neighborhood attribute metrics.Ten out of the fifteen traffic metrics were computed using actual aircraft position data obtained from the field and the remaining five were computed using simulated aircraft position data.Use of simulated data provided an easy means of computing conflict-count, aircraft type and airspeed metrics.Conflictcounts are difficult to determine from actual data because controllers make sure that aircraft are separated.Only in rare occasions operational errors occur when separation minimums are violated.Aircraft type and airspeed information is available in the flight-plan therefore it is straightforward to carry this information in a simulation.It is possible to relate the flight-plan data to aircraft position update messages using aircraft identification tag that is common to both, but the processing is more involved.Maximum numbers of aircraft in sectors during the 24-hour period were computed using both actual and simulated data.Time histories were presented to show that the results obtained using simulated data compare very well with those obtained using actual data.Data corresponding to the distribution of sectors as a function of the traffic and geometric metric values were provided in tables and in bar charts.These results show that most sectors in the current airspace have fewer than 20 aircraft at any given time.Most sectors have less than five aircraft in climb phase, fifteen in cruise phase and five in descent phase.Most of the traffic at higher altitude sectors is jet traffic.It was shown that about 98% of the sectors have fewer that three pairs of aircraft in conflict in simulation.Horizontal and vertical separation metrics indicated that aircraft fly at the same altitude in most sectors.Airspeed was found to lie in a narrow range of 400 to 500 knots.Sector transit time was found to be normally distributed with a mean of eight minutes and standard deviation of three minutes.The maximum transit time was found to be 21 minutes.A wide variation was found in sector volume, area, height and length.Most sectors were found to be elongated with an aspect ratio of two and aligned with the main traffic flows.The number of subsectors, which is a measure of sector shape complexity, was found to be less than three in most sectors.Of the 364 sectors considered in this study, 328 sectors had ten or fewer airway intersections.Maximum number of airways in a sector was determined to be 17.On an average, a sector was surrounded by 13 other sectors.The maximum number of sectors surrounding a sector was found to be 30.None of the 1017 Special Use Airspaces (SUAs) were found in 268 sectors.Eighty-one sectors had five or fewer SUAs.The maximum number of SUAs in a sector was determined to be 19.A major U. S. airport was found within 200 nautical-miles from each of the 335 sectors.Only 29 sectors were farther than 200 nautical-miles from one of the 74 major U. S. airports.Figure 1 .1Figure 1.View of 364 higher altitude sectors with a floor at or above 17,100 feet altitude.
+Figure 2 .2Figure 2. Actual and ACES simulated aircraftcounts.
+Figure 4 .4Figure 4. Sectors with peak traffic-count in the first five levels, listed in Table2, using ACES data.
+Figure 9 .9Figure 9. Sector volume histogram.Figure10.Sector height histogram.
+Figure 11 .11Figure 11.Sector reference-area histogram.Figure12.Sector reference-length histogram.
+Figure 12 .12Figure 11.Sector reference-area histogram.Figure12.Sector reference-length histogram.
+Figure 13 .13Figure 13.Sector aspect-ratio histogram.Figure14.Preferred axis of the sectors.
+Figure 14 .14Figure 13.Sector aspect-ratio histogram.Figure14.Preferred axis of the sectors.
+Figure 16 .16Figure 16.Sector intersections histogram.
+Figure 17 .17Figure 17.Sector airways histogram.
+Figure 18 .18Figure 18.Surrounding sectors histogram.
+Figure 19 .19Figure 19.Distance to closest airport outside the sector histogram.
+Table 1 .1Sector capacities.MAP No. ofsectors911011151210131514231562164517281882193820222131231
+Table 1 .1The average MAP value is approximately 17 and the most frequent MAP value is 18 for data in Table
+Table 2 .2Peak traffic-count levels.Level Peak Traffic-Count Range10 -425 -9310 -14415 -19520 -24625 -29730 -348> 34
+Table 4 .4Traffic metrics.NumberMetricDataType1Maximum number of aircraftActualCount2Maximum number of aircraft in climbActualCount3Maximum number of aircraft in cruise ActualCount4Maximum number of aircraft inActualCountdescent5Maximum number of jet aircraftSimulated Count6Maximum number of non-jet aircraftSimulated Count7Peak conflict-countSimulated Count8Average horizontal separation betweenActualSeparationaircraft in sector9Minimum horizontal separationActualSeparationbetween aircraft in sector10Average vertical separation betweenActualSeparationaircraft in sector11Minimum vertical separation betweenActualSeparationaircraft in sector12Average time-to-go to conflictActualSeparation13Sector average transit timeActualFlow14Average airspeedSimulated Flow15Variance in airspeed of aircraftSimulated Flow
+Table 5 .5Numbers of sectors with peak climb-counts using actual track-data.Level Min. Mean Max. Threshold CumulativeCumulativeLowerUpper1282323363528236321408010359364301515364364
+Table 7 .7Numbers of sectors with peak descent-counts using actual track-data.Level Min. Mean Max. Threshold CumulativeCumulativeLowerUpper1323345364532336420194110363364300115364364
+Table 8 .8Numbers of sectors with peak jet-counts using ACES data.Level Min. Mean Max. Threshold CumulativeCumulativeLowerUpper171123415734122213421910144363319417515305364402252203523645031225362364600230364364
+Table 9 .9Numbers of sectors with peak non-jet-counts using ACES data.Level PeakMin. Mean Max. Threshold CumulativeCumulativeNon-LowerUpperjetCount102022783611202361213661172306364320164333493644304154358364540165362364650026364364
+Table 6 .6Numbers of sectors with peak cruise-counts using actual track-data.These distances are added and the average minimum separation is computed.The second horizontal separation metric, C9 in Ref. 8, is the minimum horizontal separation obtained within the altitude bands.It is the minimum of the horizontal distances, rather than the average, computed in the previous metric.The two vertical separation metrics are computed in a similar manner as the horizontal separation metrics.The first vertical separation metric, C8 in Ref. 8, places a horizontal neighborhood of ten nautical-miles around each aircraft in the sector.Vertical separation distance to the closest aircraft in the horizontal neighborhood is computed.The distances are then summed up to determine the average vertical separation.The second metric, C10 in Ref. 8, is the minimum of these vertical separation distances.Level Min. Mean Max. Threshold CumulativeCumulativeLowerUpper1191363405193402241402201017536430701491532436440153620355364502625361364600330363364700035363364800142364364
+Table 10 .10Numbers of sectors with peak conflict-counts using ACES data.Level PeakMin. Mean Max. Threshold CumulativeCumulativeConflict-LowerUpperCount10361423471363472117154232222436432045903314364430174143473645404125359364650146362364760027363364870018364364
+Table 1111Level Min.Min. Mean Max. Threshold CumulativeCumulativeSeparationLowerUpperRange10-47612216057616025-9156095109321439-165601411911662623644No-041202! 0364364conflictTable 12. Numbers of sectors with minimum minimum-horizontal-separation (nautical-miles) using actual track-data.Level Min.Min. Mean Max. Threshold CumulativeCumulativeSeparationLowerUpperRange10-49326834459334425-91422341011036139-165233901661623644No-041202! 0364conflict. Numbers of sectors with minimum average-minimum-horizontal-separation (nautical-miles) using actual track-data.
+Table 13 .13Numbers of sectors with minimum average-minimum-vertical-separation (feet) using actual track-data.Level Min.Min. Mean Max. Threshold CumulativeCumulativeSeparationLowerUpperRange10-999104281353100010435321000-970011224397011223643No-060242! 0364364conflict
+Table 14 .14Numbers of sectors with minimum minimum-vertical-separation (feet) using actual track-data.Level Min.Min. Mean Max. Threshold CumulativeCumulativeSeparationLowerUpperRange10-999105287359100010535921000-97005163497011223643No-060242! 0364364conflict
+Table 15 .15Numbers of sectors with minimum average-minimum-time-to-go (seconds) using actual track-data.Level Min.Min. Mean Max. Threshold CumulativeCumulativeTime-to-LowerUppergo Range10-11913126211120132112120-2391594144240343363240-35982658360423574360-47911124480483585480-5991724600593616At most399305! 0oneaircraftFigure 7. Time-history of histograms of number of sectors grouped in six peak average-airspeed levels in Table16using ACES data.Table16.Numbers of sectors with maximum average-airspeed (knots) using ACES data.Figure8.Time-history of histograms of number of sectors grouped in five peak airspeed-variance levels in Table17using ACES data.
+Table 17 .17Numbers of sectors with maximum airspeed-variance (knots 2 ) using ACES data.Level Max.Min. Mean Max. Threshold CumulativeCumulativeAirspeed-LowerUppervarianceRange10-29331373273033327230-592513318760197358360-8957413890326364490-119018371203613645120-140013141364364
+Table 18 .18Sector geometry metrics.NumberMetricCategory1Number of sectors in centersGeographical2Number of sectors in eight airGeographicaltraffic control regions3Sector-type: low, high or super-Geographicalhigh4Sector volumeDimensions5Sector heightDimensions6Sector areaDimensions7Sector lengthDimensions8Aspect ratio -length/widthShape9Principal directionShape10Number of subsectorsShape11Number of navaidsRoute12Number of intersectionsRoute13Number of airwaysRoute14Number of airportsRoute15Special Use Airspace completelyRouteinside16Number of surrounding sectorsNeighborhood17Special Use Airspace containedNeighborhoodpartially18Distance to closest major airportNeighborhood
+Table 19 .19Numbers of sectors in centers.Center ID ZAB ZAU ZBW ZDC ZDV ZFW ZHU ZIDZJX ZKC# sectors23251319281817252027Center ID ZLA ZLC ZMA ZME ZMP ZNY ZOA ZOB ZSEZTL# sectors168621181011271022
+Table 20 .20Centers in eight regions.RegionsCode Centers# sectorsCentral RegionACEZKC27Eastern RegionAEA ZNY, ZDC29Great Lakes RegionAGL ZMP, ZAU, ZID, ZOB95New England RegionANE ZBW13Northwest Mountain Region ANM ZSE, ZLC, ZDV46Southern RegionASOZME, ZTL, ZJX, ZMA69Southwest RegionASW ZAB, ZFW, ZHU58Western Pacific RegionAWP ZOA, ZLA27
+Table 21 .21Summary of sector dimensions.MetricMinimum MeanMaximum Std. Dev.Volume (cubic-1,909 77,775572, 21086,034nautical-mile)Height (feet)4,000 47,54182,85030,164Area (square-1,6789,39745,7786,369nautical-mile)Length (nautical-419321428mile)
+Table 22 .22Numbers of subsectors in sectors.# Subsectors# Sectors1224297322465563718391121131
+Table 23 .23Numbers of 74 ASPM airports in sectors.# Airports# Sectors02811672103551
+Table 24 .24Number of SUAs contained partially in sectors.# SUAs# Sectors# SUAs# Sectors0268711318222291312101481115812363191
+		
+		
+
+			
+AcknowledgementsThe authors thank Dr. Hak-Tae Lee of University of California Santa Cruz for generalizing the moment of inertia equations from two-dimensional polygons to three-dimensional polygonal prisms and providing the Java program that was used for computing the principal directions and aspect ratio of sectors.We also thank Christopher Gleim from the University of Louisville who as a part of his internship task wrote the computer program for the Special Use Airspace analysis discussed in this paper.
+			
+
+			
+
+				
+
+
+	
+		Initial Concepts for Dynamic Airspace Configuration
+		
+			ParimalKopardekar
+		
+		
+			KarlBilimoria
+		
+		
+			BanavarSridhar
+		
+		10.2514/6.2007-7763
+	
+	
+		7th AIAA ATIO Conf, 2nd CEIAT Int'l Conf on Innov and Integr in Aero Sciences,17th LTA Systems Tech Conf; followed by 2nd TEOS Forum
+		Belfast, Northern Ireland
+		
+			American Institute of Aeronautics and Astronautics
+			September 18-20, 2007
+		
+	
+	Kopardekar, P., Bilimoria, K., and Sridhar, B., "Initial Concepts for Dynamic Airspace Configuration," Proceedings of 7th AIAA Aviation Technology, Integration and Operations Conference (ATIO), Belfast, Northern Ireland, September 18-20, 2007.
+
+
+
+	
+		A Weighted-Graph Approach for Dynamic Airspace Configuration
+		
+			StephaneMartinez
+		
+		
+			GanoChatterji
+		
+		
+			DengfengSun
+		
+		
+			AlexandreBayen
+		
+		10.2514/6.2007-6448
+	
+	
+		AIAA Guidance, Navigation and Control Conference and Exhibit
+		Hilton Head, SC
+		
+			American Institute of Aeronautics and Astronautics
+			August 20-23, 2007
+		
+	
+	Martinez, S., Chatterji, G. B., Sun, D., and Bayen, A., "A Weighted-Graph Approach for Dynamic Airspace Configuration," Proceedings of AIAA Guidance, Navigation, and Control Conference, Hilton Head, SC, August 20-23, 2007.
+
+
+
+	
+		Temporal and Spatial Distribution of Airspace Complexity for Air Traffic Controller Workload-Based Sectorization
+		
+			ArashYousefi
+		
+		
+			GeorgeDonohue
+		
+		10.2514/6.2004-6455
+	
+	
+		AIAA 4th Aviation Technology, Integration and Operations (ATIO) Forum
+		Forum, Chicago, IL
+		
+			American Institute of Aeronautics and Astronautics
+			September 20-22, 2004
+		
+	
+	Yousefi, A., and Donohue, G. L., "Temporal and Spatial Distribution of Airspace Complexity for Air Traffic Controller Workload-Based Sectorization," Proceedings of AIAA 4th Aviation Technology, Integration and Operations (ATIO) Forum, Chicago, IL, September 20-22, 2004.
+
+
+
+	
+		Comparison of algorithms for the dynamic resectorization of airspace
+		
+			WilliamPawlak
+		
+		
+			VikasGoel
+		
+		
+			DavidRothenberg
+		
+		
+			ChristopherBrinton
+		
+		10.2514/6.1998-4106
+	
+	
+		Guidance, Navigation, and Control Conference and Exhibit
+		Boston, MA
+		
+			American Institute of Aeronautics and Astronautics
+			August 10-12, 1998
+		
+	
+	Pawlak, W. S., Goel, V., Rothenberg, D. B., and Brinton, C. R., "Comparison of Algorithms for the Dynamic Resectorization of Airspace," Proceedings of AIAA Guidance, Navigation, and Control Conference, Boston, MA, August 10-12, 1998.
+
+
+
+	
+		Air traffic complexity indicators & ATC sectors classification
+		
+			RChristien
+		
+		
+			ABenkouar
+		
+		
+			TChaboud
+		
+		
+			PLoubieres
+		
+		10.1109/dasc.2002.1067924
+	
+	
+		Proceedings. The 21st Digital Avionics Systems Conference
+		The 21st Digital Avionics Systems ConferenceBudapest, Hungary
+		
+			IEEE
+			June 23-27, 2003
+		
+	
+	Christien, R., and Benkouar, A., "Air Traffic Complexity Indicators and ATC Sectors Classification," Proceedings of 5 th USA/Europe Air Traffic management R&D Seminar, Budapest, Hungary, June 23-27, 2003.
+
+
+
+	
+		Geometric Algorithms for Optimal Airspace Design and Air Traffic Controller Workload Balancing
+		
+			AmitabhBasu
+		
+		
+			JosephS BMitchell
+		
+		
+			GirishkumarSabhnani
+		
+		10.1137/1.9781611972887.8
+	
+	
+		2008 Proceedings of the Tenth Workshop on Algorithm Engineering and Experiments (ALENEX)
+		Northampton, MA
+		
+			Society for Industrial and Applied Mathematics
+			November 10-11, 2006
+			
+		
+	
+	Basu A., Mitchell, J. S. B., and Sabhnani, G., "Geometric Algorithms for Optimal Design and Air Traffic Controller Workload Balancing," 16 th Fall Workshop on Computational and Combinatorial Geometry, Northampton, MA, November 10- 11, 2006.
+
+
+
+	
+		Investigation of En route Metrics for Model Validation and Airspace Design Using the Total Airport and Airspace Modeler (TAAM)
+		
+			AYousefi
+		
+		
+			GLDonohue
+		
+		
+			KMQureshi
+		
+	
+	
+		Proceedings of the 5 th Eurocontrol/FAA ATM R&D Conference
+		the 5 th Eurocontrol/FAA ATM R&D ConferenceBudapest, Hungary
+		
+			June 23-27, 2003
+		
+	
+	Yousefi, A., Donohue, G. L., and Qureshi, K. M., "Investigation of En route Metrics for Model Validation and Airspace Design Using the Total Airport and Airspace Modeler (TAAM)," Proceedings of the 5 th Eurocontrol/FAA ATM R&D Conference, Budapest, Hungary, June 23-27, 2003.
+
+
+
+	
+		Measures for air traffic controller workload prediction
+		
+			GanoChatterji
+		
+		
+			BanavarSridhar
+		
+		10.2514/6.2001-5242
+	
+	
+		1st AIAA, Aircraft, Technology Integration, and Operations Forum
+		Los Angeles, CA
+		
+			American Institute of Aeronautics and Astronautics
+			October 16-18, 2001
+		
+	
+	Chatterji, G. B., and Sridhar, B., "Measures for Air Traffic Controller Workload Prediction," Proceedings of AIAA Aircraft, Technology, Integration, and Operations Forum, Los Angeles, CA, October 16-18, 2001.
+
+
+
+	
+		Airspace Complexity Measurement: An Air Traffic Control Simulation Analysis
+		
+			PKopardekar
+		
+	
+	
+		Proceedings of 7th USA/Europe ATM R&D Seminar
+		7th USA/Europe ATM R&D SeminarBarcelona, Spain
+		
+			July 2-5, 2007
+		
+	
+	Kopardekar, P., et al, "Airspace Complexity Measurement: An Air Traffic Control Simulation Analysis," Proceedings of 7th USA/Europe ATM R&D Seminar, Barcelona, Spain, July 2-5, 2007.
+
+
+
+	
+		Using Knowledge Exploration Tools to Study Airspace Complexity in Air Traffic Control
+		
+			RichardHMogford
+		
+		
+			ElizabethDMurphy
+		
+		
+			JeremyAGuttman
+		
+		10.1207/s15327108ijap0401_2
+	
+	
+		The International Journal of Aviation Psychology
+		The International Journal of Aviation Psychology
+		1050-8414
+		1532-7108
+		
+			4
+			1
+			
+			January, 1994
+			Informa UK Limited
+		
+	
+	Mogford, R. H., Murphy, E. D., and Guttman, J. A.,"Using Knowledge Exploration Tools to Study Airspace Complexity in Air Traffic Control," International Journal of Aviation Psychology, vol. 4, No. 1, January, 1994, pp. 29-45.
+
+
+
+	
+		Static Sector Characteristics and Operational Errors
+		
+			SGoldman
+		
+		
+			CManning
+		
+		
+			EPfleiderer
+		
+		
+	
+	Report
+	Goldman, S., Manning, C., and Pfleiderer, E., "Static Sector Characteristics and Operational Errors," Report No.
+
+
+
+	
+		Federal Aviation Administration: 800 Independence Ave. S.W., Washington, DC 20591: Internet:
+		
+			Dot/Faa
+		
+		10.4135/9781483384757.n93
+		AM-06/4
+	
+	
+		Federal Regulatory Directory: The Essential Guide to the History, Organization, and Impact of U.S. Federal Regulation
+		Washington, DC
+		
+			CQ Press
+			20591. March, 2006
+			
+		
+	
+	DOT/FAA/AM-06/4, Office of Aerospace Medicine, Federal Aviation Administration, 800 Independence Ave., S. W., Washington, DC 20591, March, 2006.
+
+
+
+	
+		Analysis of ETMS Data Quality for Traffic Flow Management Decisions
+		
+			GanoChatterji
+		
+		
+			BanavarSridhar
+		
+		
+			DouglasKim
+		
+		10.2514/6.2003-5626
+		ASDI-FD-001
+	
+	
+		AIAA Guidance, Navigation, and Control Conference and Exhibit
+		Austin, TX
+		
+			American Institute of Aeronautics and Astronautics
+			August 11-14, 2003. August 4, 2000
+		
+		
+			Volpe National Transportation Systems Center
+		
+	
+	Analysis of ETMS Data Quality for Traffic Flow Management Decisions. Version 4.0, Report No. Automation Applications Division, DTS-56, 55 Broadway Street, Cambridge, MA 02142
+	Chatterji, G. B., Sridhar, S., Kim, D., "Analysis of ETMS Data Quality for Traffic Flow Management Decisions," AIAA- 2003-5626, Proceedings of AIAA Guidance, Navigation, and Control Conference, Austin, TX, August 11-14, 2003. 13 Volpe National Transportation Systems Center, "Aircraft Situation Display To Industry: Functional Description and Interface Control Document," Version 4.0, Report No. ASDI-FD-001, Volpe National Transportation Systems Center, Automation Applications Division, DTS-56, 55 Broadway Street, Cambridge, MA 02142, August 4, 2000.
+
+
+
+	
+		Description and Analysis of a High Fidelity Airspace Model for the Airspace Concept Evaluation System
+		
+			ScottSahlman
+		
+		10.2514/6.2007-6877
+		
+	
+	
+		AIAA Modeling and Simulation Technologies Conference and Exhibit
+		Hilton Head, SC
+		
+			American Institute of Aeronautics and Astronautics
+			August 20-23, 2007. 14 July 2008
+		
+	
+	AIAA-2007-6877
+	Sahlman, S., "Description and Analysis of a High Fidelity Airspace Model for the Airspace Concept Evaluation System," AIAA-2007-6877, Proceedings of AIAA Modeling and Simulation Technologies Conference and Exhibit, Hilton Head, SC, August 20-23, 2007. 15 URL: http://www.mysql.com/[cited 14 July 2008].
+
+
+
+	
+		Build 4 of the Airspace Concept Evaluation System
+		
+			LMeyn
+		
+	
+	
+		Proceedings of AIAA Modeling and Simulation Technologies Conference and Exhibit
+		AIAA Modeling and Simulation Technologies Conference and ExhibitKeystone, Colorado
+		
+			August 21-24, 2006
+		
+	
+	AIAA-2006-6110
+	16 Meyn, L., et al, "Build 4 of the Airspace Concept Evaluation System," AIAA-2006-6110, Proceedings of AIAA Modeling and Simulation Technologies Conference and Exhibit, Keystone, Colorado, August 21-24, 2006.
+
+
+
+	
+		Validating the Airspace Concept Evaluation System for Different Weather Days
+		
+			ShannonZelinski
+		
+		
+			LarryMeyn
+		
+		10.2514/6.2006-6115
+		AIAA 2006-6115
+	
+	
+		AIAA Modeling and Simulation Technologies Conference and Exhibit
+		San Francisco, CA; Keystone, CO
+		
+			American Institute of Aeronautics and Astronautics
+			August 15-18, 2005 18 Zelinski,. August 21-24, 2006
+		
+	
+	Proceedings of AIAA Modeling and Simulation Technologies Conference and Exhibit
+	Zelinski, S. J., "Validating The Airspace Concept Evaluation System Using Real World Data," AIAA 2005-6491, Proceedings of AIAA Modeling and Simulation Technologies Conference and Exhibit, San Francisco, CA, August 15-18, 2005 18 Zelinski, S. J., and Meyn, L., "Validating The Airspace Concept Evaluation System For Different Weather Days," AIAA 2006-6115, Proceedings of AIAA Modeling and Simulation Technologies Conference and Exhibit, Keystone, CO, August 21-24, 2006.
+
+
+
+	
+		Impact of Airport Capacity Constraints on National Airspace System Delays
+		
+			GanoChatterji
+		
+		
+			YunZheng
+		
+		10.2514/6.2007-7712
+	
+	
+		7th AIAA ATIO Conf, 2nd CEIAT Int'l Conf on Innov and Integr in Aero Sciences,17th LTA Systems Tech Conf; followed by 2nd TEOS Forum
+		Belfast, Northern Ireland
+		
+			American Institute of Aeronautics and Astronautics
+			September 18-20, 2007
+		
+	
+	AIAA-2007-7712
+	Chatterji, G. B., and Zheng, Y., "Impact of Airport Capacity Constraints on National Airspace System Delays," AIAA- 2007-7712, Proceedings of 7th AIAA Aviation Technology, Integration and Operations Conference (ATIO), Belfast, Northern Ireland, September 18-20, 2007.
+
+
+
+	
+		Automating the Process of Terminal Area Node-Link Model Generation
+		
+			Hak-TaeLee
+		
+		
+			ThomasRomer
+		
+		10.2514/6.2008-7101
+		
+	
+	
+		AIAA Modeling and Simulation Technologies Conference and Exhibit
+		Honolulu, Hawaii
+		
+			American Institute of Aeronautics and Astronautics
+			August 18-21, 2008. 14 July 2008
+		
+	
+	Lee, H., and Romer T. F., "Automating the Process of Terminal Area Node-Link Model Generation," Proceedings of AIAA Modeling and Simulation Technologies Conference and Exhibit, Honolulu, Hawaii, August 18-21, 2008. 21 URL: http://aspm.faa.gov/aspm/entryASPM.asp [cited 14 July 2008]
+
+
+				
+			
+		
+	

file131.txt

@@ -0,0 +1,454 @@
+
+	
+	
+		
+II. ApproachThe DWC candidates are shown in Table 1.DWC1 and DWC2 are the two primary candidates, both achieving a desirable unmitigated collision risk of 5%.Unmitigated collision risk is the likelihood that two aircraft would violate the DAA Well Clear definition if neither aircraft used a DAA system.DWC1 achieves the minimum maneuver initial range (MIR), which is the range between aircraft when the UAS must start maneuvering away in order to maintain DWC, whereas DWC2 is simple because it does not have a time component.DWC3 and DWC4 are backup candidates carried forward from previous analysis in case DWC1 and DWC2 do not perform well in certain categories of metrics.DWC3 achieves a "risky" unmitigated collision risk of 7%.It was once proposed for terminal area UAS operations.DWC4 achieves an unmitigated collision risk between 3.5-4% and was considered a "safer" candidate that achieves an unmitigated collision risk of less than 5%.In addition to the four DWC candidates selected in [5], the Phase 1 DWC definition was also evaluated for comparison.The DWC candidates are defined by thresholds of three parameters: horizontal miss distance (HMD), vertical separation threshold (h), and   .HMD*, h*, and   * are the specific values of these parameters used to define each DWC candidate.The HMD parameter is the predicted minimum horizontal distance during an encounter, assuming constant velocities and straight line flight.The h parameter is the current altitude difference between the two aircraft.For all candidates, h* is 450 ft, which is equal to the component used in the Phase 1 MOPS.The definition of   is𝜏 𝑚𝑜𝑑 = { - 𝑟 2 -𝐷 𝑚𝑜𝑑 2 𝑟𝑟̇, >   , 0,  ≤   (1) where  and ̇ are horizontal range and range rate, respectively, between the UAS and the intruder.  is the distance modification and defines the radius of a cylinder around the UAS.In this analysis,   is set equal to the horizontal miss distance threshold, HMD *, in the DWC definitions.A loss of DWC (LoDWC) occurs when all three parameters, computed from aircraft states, fall below their respective thresholds.One million uncorrelated encounters between one low-C-SWaP UAS and one non-cooperative intruder, and one million uncorrelated encounters between one Phase 1 UAS and one non-cooperative intruder were simulated.Uncorrelated encounters are situations where intervention from ATC is unlikely, such that aircraft blunder into close proximity.The UAS trajectory is sampled from NASA's Airspace Concept Evaluation System (ACES) UAS database [6], and the intruder trajectory is sampled from Lincoln Laboratory's Uncorrelated Encounter Model [7].Next, the encounters are simulated using the Detect and AvoID Alerting Logic for Unmanned Systems (DAIDALUS) as the DAA alerting and guidance algorithm [8].DAIDALUS is a reference algorithm used to validate the Phase 1 MOPS, and was used in this study as a representative DAA algorithm.DAIDALUS generates maneuver guidance by projecting candidate vertical and horizontal DAA maneuvers to determine which would result in conflicts, and which can be used to aid the pilot in resolving the situation manually.The evaluation for each DWC consisted of two scenarios: one nominal (or unmitigated) and one mitigated.DAIDALUS issues three types of alerts in increasing levels of severity: preventive, corrective, and warning.The lowest level, preventive, is primarily used to alert the pilot to not maneuver vertically when the aircraft are separated vertically by 450-700 feet.The second level, corrective, indicates that a LoDWC is predicted and an avoidance maneuver is necessary, but there is still time for coordination with ATC.The highest level, warning, indicates that a LoDWC is imminent, an immediate avoidance maneuver is needed, and coordination with ATC before maneuvering is not a requirement.Upon alerting, DAIDALUS generates corresponding preventive, corrective, and warning guidance indicating a range of conflict-free headings and altitudes for a pilot to select from in order to maintain DWC separation.In the event that a LoDWC cannot be avoided, DAIDALUS also generates regain DWC guidance, a range of heading or altitude that can be executed to increase separation at CPA and regain DWC effectively.DAIDALUS alerts are issued based on a buffered DWC volume.Specifically, the HMD* used by the DAA alerting and guidance algorithm is scaled by a factor of 1.52 to be consistent with the parameters referenced in the Phase 1 MOPS [3].This buffer is meant to guard against maneuvering intruders and surveillance uncertainties (there are none in this work).In the simulation, the ownship only maneuvers when a corrective or warning alert is received.For maneuver guidance computation, the low C-SWaP UAS turn rate was assumed to be 7 deg/sec, which is suitable for UAS speeds from 40 to 100 kts and results in approximately the same bank angle and load factor as the Phase 1 UAS turn rate, which was assumed to be 3 deg/sec.The simulations of low C-SWaP UAS were performed using perfect (truth) surveillance, i.e., with no track uncertainty or range limitations, in order to evaluate the mitigated performance of low C-SWaP without confounding factors (such as limited surveillance range or sensor noise).However, to assess the potential impacts of limited detection ranges on safety, simulations were also run with 2 NM, 3 NM, and 4 NM surveillance ranges applied; the results of these simulations are presented in Section III.The simulations of Phase 1 UAS were run with truth surveillance (i.e., no track uncertainty) constrained by a Phase 1 radar field, defined as ±8 NM range, ±15° elevation, and ±110° azimuth.The reason for using the Phase 1 radar surveillance volume is to faciliate comparison to a previous study of the Phase 1 UAS [9].Upon completion of the simulations, metrics are computed to compare the performance of the DWC definitions.
+A. MetricsThe safety and operational suitability metrics provide an indication of whether the system will be able to operate safely without interfering with the operations of other aircraft and without causing DAIDALUS to alert unnecessarily.These metrics and their formulation are shown in Table 2 and Table 3.A Near Mid-Air Collision (NMAC) occurs when the separation between two aircraft is less than 500 ft horizontally and 100 ft vertically [10].If the ratio is less than one, then the mitigated system reduces the risk of NMAC.For example, a risk ratio of 0.1 indicates a 90% reduction in risk.Small values are desirable.LoDWC Ratio (|, ℎ ) (|, ℎ ) Similar to the NMAC risk ratio, if the LoDWC risk ratio is less than one, then the mitigated system reduces the risk of LoDWC.Small values are desirable.Given the same risk ratio, systems with lower alert ratios are desirable, since fewer alerts indicate fewer unnecessary maneuvers.
+Alerting time and range relative to the LoDWC pointThis metric can help inform alerting timeline requirements, and thus, sensor range requirements.
+B. EncountersThis study assessed two sets of encounters: one set with encounters between one low C-SWaP UAS and one noncooperative intruder, and the other set with encounters between one Phase 1 UAS and one non-cooperative intruder.All encounters contains 1 Hz aircraft states, and a majority of encounters last 180 seconds.These encounters were generated by pairing one projected UAS trajectory generated by NASA's UAS mission flights, and one intruder trajectory sampled from MIT Lincoln Laboratory's Uncorrelated Encounter Model [7].NASA's UAS mission flights consist of 19 different types of missions, including aerial imaging and mapping, law enforcement, and air quality monitoring.The demand and mission profiles were generated based on subject matter experts' opinions and socioeconomic analysis [11].The trajectories cover the entire continental US.These aircraft models are defined in a way similar to those in the Eurocontrol Base of Aircraft Data (BADA) [12].In the low C-SWaP encounter set, two types of UAS were considered: the RA-7 AAI Shadow B and the MQ-19 AAI Aerosonde.The Phase 1 encounter set includes trajectories for seven types of UAS: Cessna 208 Caravan, Cessna 510 Citation Mustang, AAI Aerosonde, MQ-9A Reaper, RQ-4A Global Hawk, Shadow B, and Socata Trinidad.The Uncorrelated Encounter Model is derived from radar data of observed aircraft operations under visual flight rules (VFR) in the National Airspace System.The model is sampled to produce random aircraft trajectories that are statistically representative of non-cooperative trajectories.Each encounter is specified by the initial positions and orientations of the two aircraft in the simulation and the nominal dynamic maneuvers that may occur leading up to the time of closest approach (TCA).Filters were applied to the ownship and intruder speeds and altitudes to ensure that the dynamics of the sampled trajectories are within the bounds for low C-SWaP UAS or Phase 1 UAS, and the intruders they are expected to encounter.The intruder characteristics from the Uncorrelated Encounter Model are the same in both sets of encounters.The low C-SWaP UAS speeds are constrained to be between 40 and 100 kts, and the Phase 1 UAS speeds are constrained to be between 40 and 250 kts, 50 kts more than permitted by the Phase 1 MOPS so as to explore potential safety issues for faster UASs.The intruder speeds range from 0 to 170 kts, the 95 percentile speed for non-cooperative intruders in the Uncorrelated Encounter Model [7].Intruders with zero speed represent aircraft like helicopters that are hovering.The encounters occur at altitudes between 500 ft above ground level (AGL) and 10,999 ft mean sea level (MSL), in airspace classes E and G.Although Class E only goes up to 10,000 ft when it is adjacent to class B or class C airspace, altitudes up to 10,999 feet were included to represent a few UAS missions that are flown slightly above 10,000 ft.The resulting altitude and speed distributions are shown in Ownship Type
+C. Pilot Response ModelUpon alerting, DAIDALUS provides guidance indicating a range of conflict-free headings and altitudes.The SC-228 standard pilot model created by Lincoln Laboratory [13] is then used to select and execute an appropriate maneuver.Encounters with an alert within the first 5 seconds of the encounter are excluded from the analysis to ensure that the pilot response model is given adequate time to resolve the conflict.Only horizontal maneuvers were executed because vertical maneuvers against non-cooperative intruders are much less robust in most situations due to the uncertainties in non-cooperative sensors' vertical measurements.The pilot response model was executed in deterministic mode, meaning that the ownship always maneuvers horizontally in the direction of the minimum suggested maneuver.In the event that the minimum suggestion is inconclusive, the ownship will turn left, as a preference for left turns was observed in human-in-the-loop experiments [13].For this analysis, the pilot model chooses the minimum heading change suggested by the guidance bands.Although variability was observed in humanin-the loop experiments, the minimum heading change was used in the analysis to isolate the effect of the DAA Well Clear definition from other parameters.After the first alert is received, the pilot response model has a 5 sec initial delay representing the time it takes the pilot to perceive the alert and devise a plan.For corrective alerts, there is an 11 sec ATC coordination time representing the time it takes the pilot to communicate the intended maneuver with ATC and receive approval.The ATC coordination time is then followed by a 3 sec execution delay representing the time it takes the pilot to enter the maneuver command into the control station and transmit this command to the UAS.The ownship may perform multiple maneuvers per encounter to resolve a conflict.The time between pilot response model decisions is determined by the alert state, as shown in Table 4.For example, if the pilot model chooses a maneuver during a warning alert state, then the situation will be reevaluated after 9 seconds (the decision update period), and a different subsequent maneuver can be issued at that time, if needed.All delays and times (e.g., the 11 sec ATC coordination time, 3 sec execution delay, etc.) are the mean values of distributions observed from human-in-the loop studies used to build the pilot response model [13].
+III. ResultsThis section presents the metrics that were evaluated.Although HMD* and   * are not completely independent (because the definition of   * is dependent on HMD*), this study may be able to provide some insight into the effect of HMD* and   * on the metrics.Because DWC1 and DWC3 have the same   * but different HMD*s, HMD* likely causes any difference in metrics between DWC1 and DWC3.Likewise, because DWC1 and DWC2 have similar HMD* but different   *,   * likely causes any difference in metrics between DWC1 and DWC2.Results for low C-SWaP UAS will be presented first (Section A), and results for Phase 1 UAS will be presented in the following subsection (Section B).All results are for mitigated encounters, unless otherwise specified.
+A. Low C-SWaP UAS ResultsA.1 Safety Metrics Fig. 3 shows the NMAC risk ratios (left) and LoDWC ratios (right) for the four DWC candidates.The causes of NMACs after DAA maneuvers include the following:1. Intruder and ownship maneuvers 2. Surveillance volume limitation and sensor uncertainties (none in this simulation) 3. Guidance ineffectiveness or instability of guidance 4. Pilot response unable to keep up with the situation For any specific encounter leading to an NMAC, all causes could have contributed to it.For example, analysis of a few select encounters leading to NMACs indicates an intruder maneuver near the UAS, causing the conflict guidance bands to saturate, leaving no conflict-free heading available.In this situation, the WCR guidance comes up, is executed, but changes turn directions multiple times during the UAS's maneuver.Combined with the pilot response delay, this instability of guidance can cause a chase situation, resulting in the DAA system's failure to avoid an NMAC.The LoDWCs result from similar causes.Fig. 5 puts NMACs into two categories.Unresolved NMAC risk is comprised of encounters that lead to nominal NMACs (i.e., without a DAA system) and which still have NMACs with the DAA system.Induced NMAC is comprised of encounters that do not have nominal NMACs but develop into NMACs with the DAA maneuver in response to DAA guidance.The unresolved and induced LoDWCs are defined in a similar way.An important observation of the NMAC risk ratios shown in Fig. 3 is that they are all fairly small and there is no statistically significant difference among them, even when compared to the Phase 1 DWC, This suggests that, given sufficient surveillance volume (infinite for this simulation) and small surveillance uncertainties (none for this simulation), all candidate DWCs are likely to be acceptable in terms of their resulting DAA performance to avoid NMACs.Interestingly, the Phase 1 DWC does not perform better with its large volume.In reality, finite surveillance volume and sensor uncertainties will increase the NMAC risk ratios.In terms of LoDWC ratio, DAA is unable to avoid LoDWC in about 10% of the encounters.DWC2 has the lowest value of .09but is only marginally lower than DWC1's .10.DWC3, DWC4, and the Phase 1 DWC all have comparable values (0.12).Unresolved risk ratios comprise the majority of the LoDWC counts.Intruder and ownship maneuvers are likely to be the main cause of these unresolved LoDWCs.Adding a buffer to the heading selected by the pilot response model was tested in this study, but showed no improvement.
+Fig. 3 Safety RatiosOne of the trends observed among the encounters with NMACs was that the intruder or ownship had a nominal (scheduled) maneuver late during the encounter.Nominal maneuvers are maneuvers that are part of the original unmitigated encounter.To analyze the impact of this trend, NMAC risk ratios and LoDWC ratios were computed for the subset of encounters where neither the ownship nor intruder has a nominal maneuver within 30 seconds of nominal TCA (Fig. 4), and for the subset of encounters with late maneuvers-i.e., where either the ownship or intruder has a nominal maneuver within 30 seconds of nominal TCA (Fig. 5).Compared to the safety ratios for all encounters (Fig. 3), the safety ratios without maneuvering (Fig. 4) are much lower.Phase 1 now has the highest LoDWC ratio, whereas previously, the LoDWC ratios for all encounters were comparable among DWC3, DWC4, and Phase 1.Since the Phase 1 DWC is the largest and has the longest timeline, it is likely that maneuvers that occurred 30 seconds or more before the TCA contribute a sizable number to the LoDWC risk.
+Fig. 4 Safety Ratios without Ownship or Intruder ManeuveringIn contrast, the NMAC risk ratios and LoDWC ratios with late maneuvers (Fig. 5) are much higher compared to ratios where encounters with maneuvers are excluded.This suggests that one reason the risk ratios for all encounters are comparable is because the risk comes primarily from encounters with late maneuvers (and hence, late alerts), which cannot be mitigated by any DWC.DWC3 is the least robust to late maneuvers with the highest LoDWC ratio, the second-highest NMAC risk ratio, and the most unresolved NMACs.
+Fig. 5 Safety Ratios with Ownship or Intruder Late ManeuveringThe system operating characteristic (Fig. 6) allows simultaneous evaluation of safety and operational suitability.The alert ratio measures the alert frequency relative to the unmitigated NMAC frequency, so it is independent of the encounter definition.Ideally, low values of both metrics are preferred and therefore the closer a system is to the origin the better.HMD* appears to have the largest effect on alert ratio; DWC1 and DWC3 have the same   *, but DWC1 has a larger HMD* and alerts more frequently.DWC3 has the lowest alert ratio because it has the smallest HMD*.
+Fig. 6 System Operating Characteristic for Low C-SWaP Encounters
+A.2. Operational Suitability MetricsAlerting time and range are computed based on the first alert of any level that occurs in an encounter.Alerting time is the projected time to unmitigated LoDWC when the alert occurs.Only encounters that have an unmitigated LoDWC are included in this metric.Fig. 7 shows the cumulative distribution function for alerting time and range.The cumulative distribution function is the probability that alerting time or range will be less than or equal to the values on the x-axis.For example, the alerting range plot shows that 60% of encounters run with DWC4 alert at range of 3 NM or less, and all encounters run with DWC4 alert within 6 NM.Mitigated encounters that, with DAA maneuvers, still result in a LoDWC (dashed lines) have on average later alert times and shorter ranges than all encounters with an alert (solid lines).This suggests that many LoDWCs may be caused by late nominal (non-DAA) maneuvers.Alerting time and range are driven more by   than by HMD (as indicated by the larger difference between DWC1 and DWC2 than between DWC1 and DWC3).DWC2, which has no   , has the earliest alerting time relative to LoDWC and the smallest alerting range.This implies the surveillance range required to provide the alerting timeline for DWC2 is likely smaller than those for other DWCs.
+A.3. Effect of Surveillance Range on Safety MetricsTo assess the potential impact of limited surveillance ranges on safety, NMAC risk ratios and LoDWC ratios were compared for simulations run with a 2 NM, 3 NM, and 4 NM surveillance range limit (shown in Fig. 8).The NMAC risk ratios for DWC 1, 2, and 3 are largely insensitive to reduced surveillance ranges.On the other hand, the NMAC risk ratios for DWC4 and Phase 1 experience large increases when the surveillance range is reduced to 2 NM.For the DWC4 and Phase 1 volumes, the intruder is sometimes not observed until loss of Well Clear has already occurred (particularly during higher speed encounters), and DAIDALUS's regain DWC guidance is likely not as effective in avoiding NMACs as its maintain DWC guidance.The LoDWC risk ratios for DWC1 and DWC2 increase noticeably while the value for DWC3 stays constant at 2 NM.The LoDWC risk ratios for DWC4 and the Phase DWC increase the most because 2 NM is inside of their DWC volume for some encounters.
+B. Phase I UAS ResultsAnalysis of the Phase 1 UAS encounters was performed on a set of one million encounters between one Phase 1 UAS and one non-cooperative intruder.The same DWC volumes used to evaluate the low C-SWaP UAS encounters were used to evaluate the Phase 1 UAS encounters in order to understand the effect of high speed UAS and baseline any additional differences when comparing to Phase 1 results.For the Phase 1 UAS results, truth surveillance data were constrained by the Phase 1 radar field of view, defined as ±8 NM range, ±15° elevation, and ±110° azimuth.
+B.1. Safety MetricsFig. 9 shows the NMAC risk ratios (left) and LoDWC ratios (right).Results similar to those for low C-SWaP UAS are desirable because this would corroborate the notion that the same DWC can be applied to both low C-SWaP UAS and Phase 1 UAS.As with the low C-SWaP UAS results, the NMAC risk ratios and LoDWC ratios are comparable among the DWC candidates.However, the risk ratios are approximately five times larger than the risk ratios for low C-SWaP UAS, and the LoDWC ratios are approximately two times larger than the LoDWC ratios for low C-SWaP UAS.This difference is primarily due to the limited 110° bearing range, which results in undetected intruders and therefore unresolved NMACs and LoDWCs.When the Phase 1 risk ratios and LoDWC ratios are computed without bearing and elevation limitations (as shown in Fig. 10), the results are much closer to the ratios obtained using the low C-SWaP encounter set (Fig. 3).Safety Ratios with Full Field of View Fig. 11 compares the above safety metrics to the alert ratio, providing insight into the potential tradeoff between safety and operational suitability.Like the low C-SWaP encounter set (Fig. 6), HMD has the largest effect on alert ratio; larger volumes result in significantly higher alert ratios, while maintaining similar safety.
+B.3 Effect of Ownship Speed on SafetyThe Phase 1 UAS encounters encompass both low C-SWaP and high-performance aircraft against a VFR intruder.To assess the sensitivity of the safety metrics to ownship speed, the results were binned by maximum ownship speed shown in Table 5.The relative frequency of encounters in each of these bins is shown in Fig. 13.Frequency of Encounters per Bin Fig. 14 shows the risk ratios binned by maximum ownship speed.There are no mitigated NMACs when the ownship aircraft has a maximum speed greater than 200 knots (Speed Bin 4), but this could be caused by the few number of encounters in Speed Bin 4. Likewise, DWC3 appears to induce NMACs for Speed Bin 3, but this is not statistically significant.In general, the risk ratios in Speed Bin 1 are highest, mainly because UAS in Speed Bin 1 are more likely to have overtaking aircraft from the rear, outside the radar's field of view.UAS in speed bins 2, 3, and 4 usually fly faster than the intruder and are therefore less likely to have undetected intruders approach from the rear.
+Fig. 14 NMAC Risk Ratios Binned by SpeedFig. 15 shows the LoDWC ratios binned by maximum ownship speed.For slower aircraft (Speed Bin 1), LoDWC ratios are all comparable.For faster aircraft (Speed Bins 2, 3, 4),   * seems to have a larger effect (low   * leads to lower LoDWC ratios).DWC2, the only DWC with a zero   *, consistently leads to the lowest LoDWC risk ratio.Although not modeled in this work, sensor uncertainties are likely to increase the NMAC and LoDWC risk ratio.For example, a MITRE study [9] yielded a NMAC risk ratio of 0.22 for a class of UAS similar to those in the Speed Bin 1 when taking into account sensor uncertainties (compared to 0.15 without uncertainty in this simulation), and a LoDWC risk ratio of 0.42 for a class of UAS similar to those in the Speed Bin 1 when taking into account sensor uncertainties (compared to 0.28 without uncertainty in this simulation).
+Fig. 15LoDWC Ratios Binned by Speed
+IV. Conclusion and Future WorkThis analysis evaluated four potential Detect-and-Avoid (DAA) Well Clear (DWC) definitions for UAS encountering non-cooperative aircraft using safety and operational suitability metrics.Two sets of encounters were used to evaluate the metrics in the two operational contexts of interest: one with encounters between low C-SWaP UAS and non-cooperative intruders, and another with encounters between Phase 1 UAS and non-cooperative intruders.The low C-SWaP UAS analysis shows that NMAC risk and LoDWC ratios are not sensitive to DWC parameters under perfect surveillance.Furthermore, safety and operational suitability are not dependent on   *; this indicates that   * may not be necessary in a DWC definition.  * affects mainly alerting performance (timing and range), whereas safety and operational suitability are of primary concern for a DAA Well Clear definition for low C-SWaP UAS.Because   requires a larger tracking range without providing any additional safety or operational suitability benefit, the results give preference to DWC2, which has no temporal parameter.The Phase 1 UAS analysis results generally follow the same trends as the low C-SWaP UAS analysis results.In terms of safety, with no radar field of view applied, the risk ratio and LoDWC ratios for Phase 1 UAS are comparable to those for low C-SWaP UAS.This seems to support the hypothesis that the final non-cooperative DWC definition for low C-SWaP UAS is also applicable to Phase 1 UAS.With the Phase 1 radar field of view applied, NMAC and LoDWC risk ratios increase noticeably due to a large number of undetected intruders approaching from the rear of the UAS.The Phase 1 UAS analysis results were also binned by maximum ownship speed.As ownship speed increases, the LoDWC risk ratios seem to be driven more by   * than by HMD*.A smaller   * seems to have little effect on safety at low speeds and also reduces the safety ratios at high speeds.This again corroborates the result that   is not needed in a definition for low C-SWaP UAS and Phase 1 UAS against non-cooperative intruders.Based on the findings presented in this paper as well as a companion paper [15], SC-228 has selected DWC2 (2200 ft, 450 ft, 0   *) for low C-SWaP UAS and Phase 1 UAS encountering non-cooperative aircraft.The following tasks can be considered as follow-on efforts to this analysis:•Fig
+Fig. 77Fig. 7 Alerting Time and Range.Solid lines are all encounters.Dashed lines are encounters with LoDWC.
+Fig. 88Fig. 8 Safety Ratios for Limited Surveillance Ranges
+Fig. 99Fig. 9 Safety Ratios
+Fig. 11 System11Fig. 11 System Operating Characteristic for Phase 1 Encounters B.2 Operational Suitability Metrics Fig. 12 illustrates the time of alert, prior to unmitigated LoDWC (left) and range at time of first alert (right).The time of alerts has a noticeable negative portion (the non-zero cumulative frequency at 0 alert time) because many intruders enter the DWC volume undetected by the limited surveillance volume.As expected, the larger Phase 1 volume alerts sooner and at larger separations.Similar trends are seen compared to the low C-SWaP UAS analysis (Fig. 7); again, alerting time and range are driven more by   than HMD.DWC2, which has no   , has the earliest alerting time relative to LoDWC and the smallest alerting range.
+Fig. 12 Alerting12Fig. 12 Alerting Time and Range.Solid lines are all encounters.Dashed lines are encounters with LoDWC.
+Fig. 1313Fig. 13Frequency of Encounters per Bin
+Additional safety analyses using DWC2, and potentially test lower HMD values with 0   to see if the DAA Well Clear definition can be further reduced • Access the effects of intruder speed or relative speed on safety performance • Development and validation of low C-SWaP sensor requirements, and associated guidance and alerting • Human factors evaluation of low C-SWaP DWC to validate the alerting timeline, and understand ATC response to and pilot acceptability of the low C-SWaP DWC.
+
+
+
+Table 11DWC CandidatesDWC1DWC2DWC3DWC4Phase 1HMD*2000 ft2200 ft1500 ft2500 ft4000 fth*450 ft450 ft450 ft450 ft450 ft𝜏 𝑚𝑜𝑑 *15 s0 s15 s25 s35 s
+Table 22Safety Metrics
+Table 3 Operational Suitability Metrics3MetricNotesAlert Ratio𝑃(𝐴𝑙𝑒𝑟𝑡|𝑒𝑛𝑐𝑜𝑢𝑛𝑡𝑒𝑟, 𝑤𝑖𝑡ℎ 𝑚𝑖𝑡𝑖𝑔𝑎𝑡𝑖𝑜𝑛)𝑃(𝑁𝑀𝐴𝐶|𝑒𝑛𝑐𝑜𝑢𝑛𝑡𝑒𝑟, 𝑤𝑖𝑡ℎ𝑜𝑢𝑡 𝑚𝑖𝑡𝑖𝑔𝑎𝑡𝑖𝑜𝑛)
+Table 4 Pilot Response Model Decision Update Times4Alert ConditionDecision Update Period (s)No Alert24Preventive Alert15Corrective Alert9Warning Alert9Regain DAA Well Clear Guidance3
+Table 5 -Ownship Speed Bins5BinMaximum Ownship Speed Range140 -100 knots2100 -150 knots3150 -200 knots4200+ knots
+		
+		
+
+			
+This material is based upon work supported by the National Aeronautics and Space Administration under Air Force Contract No. FA8702-15-D-0001.Any opinions, findings, conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Aeronautics and Space Administration.areas.These standards are documented in the Phase 1 Minimum Operational Performance Standards (MOPS) for DAA systems [3] and air-to-air radar [4] published by the RTCA Special Committee 228 (SC-228) in 2017, as well as the corresponding Technical Standard Orders (TSO), TSO-C211 and TSO-C212 published by the Federal Aviation Administration (FAA) in October 2017.The UAS in the Phase 1 MOPS are assumed to be equipped with Automatic Dependent Surveillance-Broadcast (ADS-B)
+			
+
+			
+
+				
+
+
+	
+		Defining Well Clear for Unmanned Aircraft Systems
+		
+			StephenPCook
+		
+		
+			DallasBrooks
+		
+		
+			RodneyCole
+		
+		
+			DavisHackenberg
+		
+		
+			VincentRaska
+		
+		10.2514/6.2015-0481
+	
+	
+		AIAA Infotech @ Aerospace
+		Kissimmee, FL
+		
+			American Institute of Aeronautics and Astronautics
+			2015
+		
+	
+	S. P. Cook, D. Brooks, R. Cole, D. Hackenberg and V. Raska, "Defining Well Clear for Unmanned Aircraft Systems," AIAA Infotech@Aerospace, Kissimmee, FL, 2015.
+
+
+
+	
+		FAA Position on Building Consensus Around the SARP Well-Clear Definition
+		
+			DWalker
+		
+	
+	
+		RTCA Special Committee
+		
+			2014
+			228
+		
+	
+	D. Walker, "FAA Position on Building Consensus Around the SARP Well-Clear Definition," in RTCA Special Committee 228, 2014.
+
+
+
+	
+		Minimum Operational Performance Standard (MOPS) for Helicopter Hoist Systems
+		10.4271/as6342
+		
+			2017
+			SAE International
+		
+	
+	Minimum Operational Performance Standards (MOPS) for Detect and Avoid (DAA) Systems, DO-365, RTCA. Inc., 2017.
+
+
+
+	
+		AeroMACS minimum operational performance standards (MOPS) compliance field trials for Hitachi prototype
+		
+			RafaelApaza
+		
+		10.1109/icnsurv.2015.7121331
+	
+	
+		2015 Integrated Communication, Navigation and Surveillance Conference (ICNS)
+		
+			IEEE
+			2017
+		
+	
+	Minimum Operational Performance Standards (MOPS) for Air-to-Air Radar for Traffic Surveillance, DO-366, RTCA. Inc., 2017.
+
+
+
+	
+		Well Clear Trade Study for Unmanned Aircraft System Detect And Avoid with Non-Cooperative Aircraft
+		
+			MinghongGWu
+		
+		
+			AndrewCCone
+		
+		
+			SeungmanLee
+		
+		
+			ChristineChen
+		
+		
+			MatthewWEdwards
+		
+		
+			DevinPJack
+		
+		10.2514/6.2018-2876
+	
+	
+		2018 Aviation Technology, Integration, and Operations Conference
+		Atlanta, GA
+		
+			American Institute of Aeronautics and Astronautics
+			2018
+		
+	
+	M. G. Wu, A. C. Cone, S. Lee, C. Chen, M. W. M. Edwards, Jack and D. P., "Well Clear Trade Study for Unmanned Aircraft System Detect and Avoid with Non-Cooperative Aircraft," AIAA Aviation Conference, Atlanta, GA, 2018.
+
+
+
+	
+		Build 8 of the Airspace Concept Evaluation System
+		
+			SapaGeorge
+		
+		
+			GoutamSatapathy
+		
+		
+			VikramManikonda
+		
+		
+			KeePalopo
+		
+		
+			LarryMeyn
+		
+		
+			ToddLauderdale
+		
+		
+			MichaelDowns
+		
+		
+			MohamadRefai
+		
+		
+			RichardDupee
+		
+		10.2514/6.2011-6373
+	
+	
+		AIAA Modeling and Simulation Technologies Conference
+		
+			American Institute of Aeronautics and Astronautics
+			2011
+		
+	
+	S. George, G. Satapathy, G. Manikonda, M. Refai and R. Dupee, "Build 8 of the Airspace Concept Evaluation System," in AIAA Modeling and Simulation Technologies Conference, 2011.
+
+
+
+	
+		Extended Airspace Encounter Models for Unmanned Aircraft Sense and Avoid Safety Evaluation
+		
+			AndrewJWeinert
+		
+		
+			EricHarkleroad
+		
+		
+			JohnGriffith
+		
+		
+			MatthewWEdwards
+		
+		10.2514/6.2013-5049
+	
+	
+		AIAA Infotech@Aerospace (I@A) Conference
+		Lexington, MA
+		
+			American Institute of Aeronautics and Astronautics
+			2013
+		
+	
+	A. Weinert, E. Harkleroad, J. Griffith, M. Edwards and M. Kochenderfer, "Uncorrelated Encounter Model of the National Airspace System Version 2.0 (ATC-404)," Lincoln Laboratory, Lexington, MA, 2013.
+
+
+
+	
+		DAIDALUS: Detect and Avoid Alerting Logic for Unmanned Systems
+		
+			CesarMunoz
+		
+		
+			AnthonyNarkawicz
+		
+		
+			GeorgeHagen
+		
+		
+			JasonUpchurch
+		
+		
+			AaronDutle
+		
+		
+			MariaConsiglio
+		
+		
+			JamesChamberlain
+		
+		10.1109/dasc.2015.7311421
+	
+	
+		2015 IEEE/AIAA 34th Digital Avionics Systems Conference (DASC)
+		Prague, Czech Republic
+		
+			IEEE
+			2015
+		
+	
+	C. Muñoz, A. Narkawicz, G. Hagen, J. Upchurch, A. Dutle, M. Consiglio and J. Chamberlain, "DAIDALUS: Detect and Avoid Alerting Logic for Unmanned Systems," in 34th IEEE/AIAA Digital Avionics Systems Conference, Prague, Czech Republic, 2015.
+
+
+
+	
+		SC-228 DAA Phase 1 MOPS Verification & Validation Simulation
+		
+			MHall
+		
+		
+			JKaznocha
+		
+		
+			TLester
+		
+		
+			SSherman
+		
+		
+			NReep
+		
+		
+			2017
+			Bedford, MA
+		
+		
+			MITRE Corporation
+		
+	
+	M. Hall, J. Kaznocha, T. Lester, S. Sherman and N. Reep, "SC-228 DAA Phase 1 MOPS Verification & Validation Simulation," MITRE Corporation, Bedford, MA, 2017.
+
+
+
+	
+		Traffic Alert and Collision Avoidance System (TCAS II) Transition Program
+		
+			GaryPGambarani
+		
+		10.4271/901970
+	
+	
+		SAE Technical Paper Series
+		1, Washington, DC; B, RTCA
+		
+			SAE International
+			2008
+		
+	
+	Minimum Operational Performance Standards for Traffic Alert and Collision Avoidance System II (TCAS II) Version 7.1, Washington, DC: DO-185B, RTCA, 2008.
+
+
+
+	
+		UAS Demand Generation Using Subject Matter Expert Interviews and Socio-economic Analysis
+		
+			SricharanKAyyalasomayajula
+		
+		
+			RohitSharma
+		
+		
+			FrederickWieland
+		
+		
+			AntonioTrani
+		
+		
+			NicolasHinze
+		
+		
+			ThomasSpencer
+		
+		10.2514/6.2015-3405
+	
+	
+		15th AIAA Aviation Technology, Integration, and Operations Conference
+		
+			American Institute of Aeronautics and Astronautics
+			2015
+		
+	
+	S. Ayyalasomayajula, R. Sharma, W. F. A. Trani, N. Hinze and S. Spencer, "UAS Demand Generation Using Subject Matter Expert Interviews and Socio-Economic Analysis," in Proceedings of the AIAA Aviation Conference, 2015.
+
+
+
+	
+		BADA: An advanced aircraft performance model for present and future ATM systems
+		
+			AngelaNuic
+		
+		
+			DamirPoles
+		
+		
+			VincentMouillet
+		
+		10.1002/acs.1176
+	
+	
+		International Journal of Adaptive Control and Signal Processing
+		Int. J. Adapt. Control Signal Process.
+		0890-6327
+		
+			24
+			10
+			
+			2010
+			Wiley
+		
+	
+	A. Nuic, D. Poles and V. Mouillet, "An Advanced Aircraft Performance Model for Present and Future ATM Systems," International Journal of Adaptive Control and Signal Processing, Vol. 24, No. 10, pp. 850-866, 2010.
+
+
+
+	
+		A Model of Unmanned Aircraft Pilot Detect and Avoid Maneuver Decisions
+		
+			RGuendel
+		
+		
+			MKuffner
+		
+		
+			DMaki
+		
+		ATC-434
+		
+			2017
+		
+		
+			Massachusetts Institute of Technology Lincoln Laboratory Project
+		
+	
+	Report
+	R. Guendel, M. Kuffner and D. Maki, "A Model of Unmanned Aircraft Pilot Detect and Avoid Maneuver Decisions," Massachusetts Institute of Technology Lincoln Laboratory Project Report ATC-434, 2017.
+
+
+
+	
+		Better Bootstrap Confidence Intervals
+		
+			BradleyEfron
+		
+		10.1080/01621459.1987.10478410
+	
+	
+		Journal of the American Statistical Association
+		Journal of the American Statistical Association
+		0162-1459
+		1537-274X
+		
+			82
+			397
+			
+			1987
+			Informa UK Limited
+		
+	
+	B. Efron, "Better Bootstrap Confidence Intervals," Journal of the Americal Statistical Association, vol. 82, no. 397, pp. 171-185, 1987.
+
+
+
+	
+		Detect-and-Avoid Alerting Performance for High-Speed UAS and Non-Cooperative Aircraft
+		
+			AndrewCCone
+		
+		
+			MinghongGWu
+		
+		
+			SeungManLee
+		
+		10.2514/6.2019-3313
+	
+	
+		AIAA Aviation 2019 Forum
+		
+			American Institute of Aeronautics and Astronautics
+			2019
+		
+	
+	A. C. Cone, M. G. Wu and S. Lee, "Detect-and-Avoid Alerting Performance for High-Speed UAS and Non- Cooperative Aircraft," in AIAA Aviation Conference, 2019.
+
+
+				
+			
+		
+	

file132.txt

@@ -0,0 +1,875 @@
+
+	
+	
+		
+I. Introduction3][4][5][6] The largest environmental impacts for enroute air traffic comes from emissions of carbon-dioxide and nitrogen-oxides, and persistent contrail formations.It has been shown that commercial aircraft can reduce climate impact due to these factors by modifying their trajectories, although this often comes at the cost of increased fuel consumption. 7Such an increase in fuel consumption represents an increase in the operational cost incurred by an airline.The three largest environmental impacts for enroute air traffic include direct emissions of greenhouse gases such as carbon dioxide (CO 2 ), emissions of nitrogen oxides (NO X ), and persistent contrails.CO 2 and NO X emissions are a function of fuel burn therefore reducing fuel consumption results in emissions reductions.Various procedures have been proposed in the past to reduce the persistent contrail formation, including promising approaches based on changing aircraft flight altitudes.Mannstein 8 proposed a strategy to reduce the climate impact of contrails significantly by only small changes in individual flight altitude.Williams 9, 10 proposed strategies for contrail reduction by identifying fixed and varying maximum altitude restriction policies.However, these restrictions generally imply more fuel burn, thus more emissions.Sridhar, 11 Chen, 12 and Wei 13 proposed contrail reduction strategies by altering an aircraft's cruising altitude in a fuel-efficient way, but these strategies did not address the environmental impact from aircraft emissions.Recently, the Absolute Global Temperature Potential (AGTP), a climate assessment metric that adapts a linear system for modeling the global temperature response to aviation emissions and contrails, was introduced in Ref. 14  and 15 to study the combined effect of CO 2 emissions and contrail formation on the reduction strategies.Chen et al. 7 evaluate both the reduction in environmental cost and the increase in operational costs for the climate reduction strategy by applying the same flight altitude change for all aircraft in each of the twenty U.S. Air Traffic Control Centers.A detailed climate reduction method for individual aircraft was not addressed in that study.This paper follows the research reported in Ref. 7 and develops an climate impact reduction algorithm for individual aircraft.The goal of this work is to identify flights for which the environmental cost of climate impact reduction outweighs the increase in operational cost on an individual aircraft basis.To determine this, the changes in cost imposed upon both the airlines and society are considered using the cost of fuel and the social cost of carbon [16][17][18] by developing a trajectory modification algorithm that modifies individual aircraft trajectories.A trajectory modification algorithm has been developed to minimize the aircraft environmental cost by reducing AGTP 14,15,19 due to both contrails and CO 2 emissions.The increase in fuel consumption that leads to higher fuel costs imposed on airlines is also computed by the algorithm.This research aims to identify flights that yield the most environmental benefit for the least operational cost from climate impact reduction strategies.The remainder of the paper is organized as follows.Section II provides a description of the climate impact model, the cost model, and the climate impact reduction method.Next, Section III shows the results and analyses of climate impact reductions for flights from eight different airlines.Finally, Section IV presents a summary and conclusions.
+II. Models and Methods
+II.A. Climate Impact ModelThe climate response to aviation emission and contrails can be modeled as outputs from a series of linear dynamic systems.The carbon cycle models describe the changes to the CO 2 concentration due to the transport and absorption of CO 2 by the land mass and various ocean layers.The Radiative Forcing (RF) for CO 2 emissions is made of a steady-state component and three exponentially decaying components. 20ontrails occur at different regions of the earth and add non-uniform sources of energy to the atmosphere.The latest estimates indicate that contrails caused by aircraft may be causing more climate warming today than all the residual CO 2 emitted by aircraft. 21The net RF for contrails includes the effect of trapping outgoing longwave radiation from the Earth and that of reflecting incoming shortwave radiation from the sun.Energy Forcing (EF) is the net energy flux induced to the atmosphere by a unit length of contrail over its lifetime.Estimates of EF given the RF forcing due to contrails are described in Ref. 22.The EF is expressed as joules/km of contrails.NO X increases the amount of ozone in the atmosphere while decreasing the amount of methane in the atmosphere.The amount of ozone produced depends on the lifetime of NO X that varies from days to weeks in the upper troposphere.The RF associated with NO X is made up of short-lived positive RF due to ozone and a negative RF due to methane and methane-induced ozone and the combined effect results in a net RF due to NO X . 23Research in Ref. 6 shows NO X has relatively small effect for the climate reduction strategies compared to CO 2 and contrails, therefore its effect is ignored in this paper.The lifetime associated with different emissions and contrails varies from a few hours to several hundred years.The impact of certain gases depends on the amount and location of the emission, and the decisionmaking horizon, H in years, when the impact is estimated.These variations make it necessary to develop a common yardstick to measure the impact of various gases.Several climate metrics have been developed to assess the impact of the aviation emissions. 24Using linear climate response models, the Absolute Global Temperature Potential (AGTP) measures the mean surface temperature change because of different aircraft emissions and persistent contrail formations. 19AGTP provides a way to express the combined environmental cost of CO 2 and NO X emissions, and contrails as a function of the fuel cost.For simplicity, the RF due to contrails is assumed to be independent of the location of the contrails.The near surface temperature change ∆T for each flight can be approximated as∆T = ∆T CO2 + ∆T Con ,(1)where ∆T CO2 is the contribution to AGTP from CO 2 emissions in Kelvin (K) and ∆T Con is the contribution to AGTP from contrails in K. ∆T CO2 is a linear function of the additional CO 2 emissions and ∆T Con is a linear function of the contrail formation time.The coefficients of the linear functions, also known as pulse AGTP, depend on the linear models for RF, the specific forcing because of CO 2 , energy forcing because of contrails, energy balance model and the duration of the climate effect horizon. 14Using the coefficients described in Ref. 6, at the time horizon of H, Eq.( 1) can be rewritten as∆T H = AGTP H CO2 E CO2 + AGTP H Con L Con ,(2)where ∆T H is the temperature changes due to both CO 2 and contrails for the time horizon of H in K, AGTP H CO2 is the coefficient of AGTP due to CO 2 for the time horizon of H in K/kg, AGTP H Con is the coefficient of AGTP due to contrails for the time horizon of H in K/km, E CO2 is the amount of CO 2 emissions in kg, and L Con is the contrail length in km.A list of pulse AGTP coefficients used in this paper is shown in Table 1.The details of the fuel burn, emissions, and contrail models are described in Ref. 12.The details of the climate model can be found in Ref. 6.
+II.B. Cost ModelThe total social cost of fuel consumption is comprised of the private cost of paying for fuel, borne by airlines and in turn their passengers, and the external cost of environmental damage, borne by societies, present and future.The social cost of carbon (SCC) is the cost, in monetary terms, to society of emitting an additional metric ton of carbon dioxide.It is often used to determine how much investment should be undertaken in order to mitigate the effects of carbon dioxide emissions.It also represents the theoretical value of a carbon tax for a perfect market.This is particularly suitable because asking or requiring airlines to increase fuel costs to reduce contrail formation would be a form of tax on contrail-induced environmental damage.The United States Government combines results from the three most prominent climate models to determine a suitable measure for the social cost of carbon and recently adopted a value of $36 United States Dollars (USD) in 2007 dollars, which is equivalent to $41 USD in 2013 dollars. 25This is the value used for the purpose of this research.Fuel costs historically represent as much as 33% of aircraft operating costs with an increasing trend.The fuel cost for individual flights are likely to increase if otherwise-quasi-optimal trajectories are modified in a way that is detrimental to fuel efficiency so as to avoid contrail favorable regions.For the purpose of this work, the price of jet fuel of $4 USD per US gallon was used in this paper.The social cost of carbon can be used to quantify the environmental cost of CO 2 emission.Using the social cost of carbon dioxide as an estimate of environmental cost of CO 2 , the additional contribution to environmental cost from CO 2 emissions, ∆Cost CO2 , can be formulated as∆Cost CO2 = SCC • ∆E CO2 1000 ,(3)where SCC is the social cost of carbon in dollar per metric ton, and ∆E CO2 is the change in CO 2 emissions in kg.In order to quantify the environmental cost of contrails, the environmental cost of temperature changes, specifically one Kelvin of AGTP, was defined using the SCC and the AGTP coefficient of CO 2 for time horizon H years,ECK H = SCC 1000 • AGTP H CO2 ,(4)where ECK H is the equivalent environmental cost of temperature change in dollars per Kelvin for the time horizon of H years. Assume that the surface temperature is reduced after the climate impact reduction (∆T H < 0), the total environmental cost reduction ∆Cost H Env can be formulated as∆Cost H Env = ECK • (-∆T H ). (5)Note that ∆Cost H Env is postive after the climate impact reduction.The environmental net benefit, N B H Env , is defined asN B H Env = ∆Cost H Env -∆Cost Opr ,(6)where ∆Cost Opr is the additional operational cost of applying the climate impact reduction.Only the cost of additional fuel burn is considered as additional operational cost in this paper.If the environmental cost reduction ∆Cost H Env is greater than the additional operational cost ∆Cost Opr , the environmental net benefit N B H Env is positive.
+II.C. Climate Impact ReductionA preliminary trajectory modification algorithm has been developed.The goal of the algorithm is to reduce the total AGTP effect of a flight by modifying its trajectory.Previous study in Ref. 7 shows that the climate effect can be reduced efficiently by applying the same flight altitude change for all aircraft in each of the twenty U.S. Air Traffic Control Centers.This algorithm follows that concept and focuses on modifying the flight profile for individual aircraft; it allows aircraft to deviate no more than one flight level (2,000 feet) above or below the original flight path.It is assumed airlines choose to fly at, or close to, each aircraft optimal operating conditions and, at least approximately, along the most fuel-efficient trajectory, given the weather and traffic conditions at the time of flight.Aircraft can potentially reduce climate impact by avoiding contrail favorable regions either by climbing to a higher cruise altitude of descending to a lower cruise altitude.For most typical commercial aviation cruise altitudes, flying higher will generally yield higher fuel efficiency.However, flight ceiling and mechanical safety constraints often limit aircraft maximum cruise altitudes.The algorithm evaluates the environmental cost for the period of a flight cruise segment.The total environmental cost is calculated as the combined AGTP effect of CO 2 emissions due to fuel consumption and persistent contrail production caused by flying through contrail regions.The algorithm allows the flight to make one altitude change, meaning climbing 2,000 feet or descending 2,000 feet then returning to the original cruise altitude.The algorithm computes the combined environmental cost and operational cost of all flight segments at the cruise altitude and 2,000 feet above and below it and finds the path that will maximize the environmental net benefit.If this alternative results an environmental net benefit, then the flight path is altered to incorporate this change.Figure 1 shows an example flight modification for one of the flights tested.The grey blocks represent the contrail regions.The contrail regions were computed based on the weather data at the aircraft's take-off time and are assumed to be static during the flight.The blue line is the original flight path and the green line is the new path after modification.As indicated in the figure, the new path tried to avoid the contrail regions by flying 2,000 feet lower than the original flight path.The new path will result in reduction in ∆T Con by avoiding the contrail regions but increase ∆T CO2 due to additional fuel burn at a given time horizon H.The net changes in ∆T H is negative, meaning the net climate impact is reduced after the flight path modification.The environmental cost reduction ∆Cost H Env is increased because of the reduction in ∆T H , and the operational cost ∆Cost Opr is also increased because of the additional fuel burn.The net environmental benefit is the difference of the two costs.If the environmental cost saving is greater than the additional operational cost, it will result in environmental net benefit In reality, it is not possible to know the exact contrail regions to avoid before flying.The forecast data is required to predict the contrail regions so that the algorithm can determine the path to reduce the climate impact.Using actual weather data in the algorithm is like having perfect forecast data, which is not realistic.Figure 2 shows the same example of flight modification with the predicted and actual contrails regions.The grey blocks represent the contrail regions and the black grid blocks represent the predicted contrail regions.The predicted contrail regions were computed based on the one-hour weather forecast data at the aircraft's take-off time for the entire flight.The algorithm modified the flight trajectory based on predicted contrail regions, and use the actual contrail regions to determine the actual environmental cost.The blue line is the original flight path and the green line is the new path after modification.As indicated by the green line, because of the inaccuracy in the forecast data, the flight would fly through some contrail regions then lower the altitude before it reaches the black grid blocks.The flight would also fly back to the original cruise altitude after the predicted contrails regions is clear of contrails but the actual contrails still exist.It would still result in reduction in environmental cost but the benefit would be reduced because of the inaccuracy in the forecast data.
+III. Results
+III.A. Using actual weather dataThe trajectory modification algorithm analyzed 12,787 flights using actual flight track data from the Enhanced Traffic Management System of April 23, 2010.These are all flights carried by one of eight major US airlines that operated the most flights on the day: American Airlines (AAL), America West Airlines (AWE), ExpressJet Airlines (BTA), Delta Airlines (DAL), American Eagle Airlines (EGF), SkyWest Airlines (SKW), Southwest Airlines (SWA), and United Airlines (UAL).The contrail model uses atmospheric temperature and humidity data retrieved from the Rapid Updated Cycle (RUC) data, provided by the National Oceanic and Atmospheric Administration (NOAA).The actual and one-hour forecast data based on the take-off time were used to find the contrail regions along each flight path.The day was selected because there were large portions of US airspace covered by the contrail regions.A time horizon of 25 years was used in the climate model; the study in Ref. 7 shows that the environmental benefit after applying the climate impact reduction strategy for time horizon of 25 years is more significant than for time horizon of 50 and 100 years.Figure 3 shows the additional operational cost (the fuel cost increase in this paper), against the environmental cost reduction using actual weather data for the flights with positive net benefit for each airline, in descending order of total net benefit, after the reduction strategy.Each blue dot represents a flight with net benefit (when the environmental cost reduction is greater than operational cost) greater than zero.From a policy perspective, the most desirable flight modifications reduce the net environmental cost by the most while increasing the fuel cost by the least, as these will result in the greatest net benefit.Graphically, these points can be found in the bottom-right corner of the figure.In the figure, Airlines #1, #2, and #3 have a similar pattern.They show more blue dots at the bottom-right corner, while Airlines #4, #6, #7, and #8 show a similar pattern with less blue dots than the others and most of the dots are on the lower of the left-half side.This is mainly because Airlines #1, #2, and #3 have more long-haul flights that will benefit more from climate impact reduction by avoiding long contrails.Airlines #4, #6, #7, and #8 have more short-haul flights therefore the environmental cost reduction of each flight is smaller.Airlines #4, #6 have more blue dots than Airlines #7, and #8 simply because they have more flights during the day.Airline #5 has many short-haul flights and also some long-haul flights therefore the plot is a mix of the two patterns.These observations are consistent with the findings in Ref. 26.The climate impact reduction algorithm was able to achieve a net benefit for 3,067 of the 12,787 flights (24%).The total net benefit is $843,416, or equivalent to a reduction of around 20,000 tons of carbon emissions.The net benefit per flight is $275.Among the 3,067 flights, there are 77 flights resulting in net benefit greater than $1,000.The total net benefit among the 77 flights is $95,482, or $1240 per flight.These flights could be the most cost-efficient candidates for applying the climate reduction maneuver.The results for each of the eight airlines are summarized in Table 2.In the table, it shows Airline #3 has the highest percentage of flights resulting in net benefit, at 43.1% even though the total net benefit is not the highest.This is because Airline #3 has more long-haul and less short-haul flights than the others.Airlines #1 and #2 are next at 29.0%, then Airline #4 at 24.3%.The other four airlines, which have mostly short-haul flights, have percentages less than 20%.This suggests long-haul flights would be better candidates for climate impact reduction than the short haul flights.In reality, it is not possible to know the exact contrail regions to avoid before flying.In this subsection, the one-hour forecast data based on the flight take-off time were used to predict the contrail regions.The climate reduction algorithm used the predicted contrail regions to modify the flight trajectories and used the actual weather data to compute the environmental cost reductions.Because of the inaccuracy in the forecast data, the performance of the climate reduction algorithm was reduced.Figure 4 shows the same example in Fig. 3 using forecast data for the algorithm.Because of the inaccuracy in the forecast data, it is possible that the flights would fly through some contrail regions or would climb or descend without contrail regions present, therefore resulting in lower environmental net benefit or even negative net benefit.It can be seen in the figure that the blue dots were shifted toward the left side compared to the blue dots in Fig. 3.The red dots represent the flights with negative net benefit, where the increases in the operational costs are larger than the environmental cost reductions.The flights with negative net benefit are the group of flights with small environmental cost reduction using actual data (bottom-left corner in Fig. 3).Only applying the algorithm to flights with large net benefit (blue dots on the right side) would avoid negative net benefits.Using the one-hour forecast data, the climate reduction algorithm was still able to reduce the net benefit for 2,043 of the 12,787 flights (16%) on the selected day; the algorithm identified 2,959 flights for climate reduction and 916 of them ended up with negative net benefit because of the inaccuracy of the forecast data.Also, among the 3,067 flights that could have received net benefit if actual weather data were used, 515 of them were not identified for a maneuver using forecast data.The total environmental net benefit was reduced from $843,416 to $499,256 when using forecast data compared to knowing the actual weather condition, which is a 41% reduction.The net benefit per flight for this one day was $169.The results for each airline are summarized in Table 2. Using weather forecast data, Airlines #1, #2, and #3 still result in the most total net benefits among the eight airlines, mainly because the three airlines have mostly long-haul flights.Airline #3 remains having the most net benefit per flight.As indicated in the table, inaccurate forecast data have significant impact on the performance of the climate reduction algorithm for all airlines.
+IV. ConclusionsA algorithm has been developed that modifies the trajectories of individual flights to evaluate the effect of environmental cost and operational cost of flights in the United States National Airspace System.The algorithm identifies flights of which the environmental cost of climate impact reduction outweighs the increase in operational cost on an individual aircraft basis and modifies their trajectories to achieve the maximum environmental net benefit, which is the difference between the reduction in environmental cost and the additional operational cost.The result shows on a selected day, 24% of the flights can achieve environmental net benefit using actual weather data and 16% of the flights can achieve environmental net benefit using weather forecast data, resulting in net benefit of around $840,000 and $500,000, respectively.It also suggests that the long-haul flights would be better candidates in cost-efficient climate impact reduction than the short haul flights.Future work of this study includes using a more detail contrail model, 27 designing more operational viable routing, and update the actual and forecast weather data along the flights.Figure 1 .1Figure 1.Flight profile (blue line: baseline, green line: after reduction) and contrail regions (grey areas) on April 23, 2010.
+Figure 2 .2Figure 2. Flight profile (blue line: baseline, green line: after reduction), actual contrail regions (grey areas), and predicted contrail regions (black grids) on April 23, 2010.
+Table 1 .1Pulse AGTP values for CO2 and contrails for three different time horizonsTime HorizonH = 10 years H = 25 years H = 100 yearsAGTP H CO2 , K/kg6.0×10 -166.7×10 -165.1×10 -16AGTP H Con , K/km1.5×10 -133.0×10 -145.1×10 -15
+Table 2 .2Number of flights and net benefit (NB) before and after climate reduction algorithm using actual weather dataAirline total flights with NB%total NB NB per flight NB > $1000 total NB#1229066529.0% $202,901$30529$34234#2180152229.0% $147,772$2837$8101#3103544643.1% $141,684$31822$30236#4170641524.3% $105,976$2558$9362#5121223719.6% $69,685$2945$6042#6215934015.7% $63,912$1883$3778#7114122019.3% $58,305$2652$2318#8144322215.4% $53,181$2401$1401Total12787306724.0% $843,416$27577$95482
+Table 3 .3Number of flights and net benefit (NB) before and after climate reduction algorithm using forecast weather data
+			of 11 American Institute of Aeronautics and Astronautics Downloaded by NASA AMES RESEARCH CENTER on June 20, 2014 | http://arc.aiaa.org| DOI: 10.2514/6.2014-3016
+			of 11 American Institute of Aeronautics and Astronautics Downloaded by NASA AMES RESEARCH CENTER on June 20, 2014 | http://arc.aiaa.org| DOI: 10.2514/6.2014-3016
+			Downloaded by NASA AMES RESEARCH CENTER on June 20, 2014 | http://arc.aiaa.org| DOI: 10.2514/6.2014-3016
+			of 11 American Institute of Aeronautics and Astronautics Downloaded by NASA AMES RESEARCH CENTER on June 20, 2014 | http://arc.aiaa.org| DOI: 10.2514/6.2014-3016
+			of 11 American Institute of Aeronautics and Astronautics Downloaded by NASA AMES RESEARCH CENTER on June 20, 2014 | http://arc.aiaa.org| DOI: 10.2514/6.2014-3016
+			American Institute of Aeronautics and Astronautics Downloaded by NASA AMES RESEARCH CENTER on June 20, 2014 | http://arc.aiaa.org| DOI: 10.2514/6.2014-3016
+		
+		
+			
+			
+			
+
+				
+
+
+	
+		ex-nomination 24, 27, 31–2; in humanism: Catholic 69; technological French translation 6, 33 70; traditional 73 hyperconformity 2–3 design 81–3, 118–19; Bauhaus 71 dropping out 101–3 Internet 99; cybercops 101; cyberculture and business 9 effraction 90; break and entry 86; see implosion 4, 50, 94–8, 111, 122; and also symbolic exchange consciousness 83; and nationalism Einsteinism 18, 23 103 electricity: light 48–9; and language 49; and implosion 96–7 Japan 1 Eskimos 107–8, 110–11, 116 Jesus 104, 116 Expo ’67 5, 59, 92, 100; Christian j’explique rien 5 Pavilion 104; Québec Pavilion 5, 92 Expo ’92 4 Latin character 44; Gallic 7, 56, 57, 58; extensions of man 68, 85, 90; mediatic Gallicized name 53; opposed to 58 53; outering 12 liberalism 46, 103–4; cool media 105 families 101; human 102; mafia 101; M et M 58 McLuhan’s 56; commune-ist 116 Ma – Ma – Ma – Ma 58–9 figure and ground 21, 26, 35 Mac 53, 54, 58; Macbeth 54; MacBett French McLuhan 1, 2, 20, 76–8, 98; 57; Macheath 54; Big Mac 58 new 77 Le mac 62 Mack 55 galaxies 39, 41–2, 44, 99, 109, 116; McLuhan: Counterblast 118; Du and detribalization 107; Gutenberg cliché à l’archétype 119–20; 4, 14, 18, 26, 42–3, 47, 51, 85, Explorations in Communication 121; galactic shifts 38; galaxie 16; From Cliché to Archetype 119; MacLuhan 56; and tribalism 106 La galaxie Gutenberg 4, 44; The gap in historical experience 8, 91–2, Gutenberg Galaxy 4, 8, 18, 26, 49– 99, 106 50, 99, 107, 109; The Mechanical Gen-X 43, 105 Bride 18, 24–5, 27–9, 31–2, 34, 107; Global Village 4, 94, 100, 107, 111, Letters 15, 21, 55; The Medium is 121; global consciousness 102–3; the Massage 9, 26, 68; Message et and idiocy 12; and nomadology massage 44; Mutations 1990 44; 110–11; and teamness 9 Pour comprendre les médias 44, 87; grammatology 7, 39–41; écriture 37, 39, Through the Vanishing Point 120; 41; and logocentrism 40 Understanding Media 8, 13, 18–19, 23–4, 29, 68, 78, 85, 95; War and happenings 83, 119–20 Peace in the Global Village 16, 26 hemispheres 25 McLuhanacy 3, 84; McLuhanatic 108 McLuhan renaissance 1, 10, 12, 99
+		10.4324/9780203005217-18
+		#3 1035 435 128 65 $90,672 $208 #4 1706 388 87 48 $68,721 $177 #5 1212 215 60 29 $40,522 $188 #6 2159 306 125 69 $27,576 $90 #7 1141 231 67 53 $35,155 $152 #8 1443 213 69 31 $29,358 $138 Total 12787 2959 916 515 $499
+	
+	
+		McLuhan and Baudrillard
+		
+			Routledge
+			
+			901
+			
+		
+	
+	total flights identified flights with neg
+	total flights identified flights with neg. NB missed flights total NB NB per flight #1 2290 674 235 139 $121,351 $180 #2 1801 497 145 81 $85,901 $173 #3 1035 435 128 65 $90,672 $208 #4 1706 388 87 48 $68,721 $177 #5 1212 215 60 29 $40,522 $188 #6 2159 306 125 69 $27,576 $90 #7 1141 231 67 53 $35,155 $152 #8 1443 213 69 31 $29,358 $138 Total 12787 2959 916 515 $499,256 $169
+
+
+
+	
+		
+			IWaitz
+		
+		
+			JTownsend
+		
+		
+			JCutcher-Gershenfeld
+		
+		
+			EGreitzer
+		
+		
+			JKerrebrock
+		
+		Report to the United States Congress: Aviation and the Environment, A National Vision, Framework for Goals and Recommended Actions
+		London, UK
+		
+			December 2004
+		
+	
+	Tech. rep
+	Partnership for AiR Transportation Noise and Emissions Reduction
+	Waitz, I., Townsend, J., Cutcher-Gershenfeld, J., Greitzer, E., and Kerrebrock, J., "Report to the United States Congress: Aviation and the Environment, A National Vision, Framework for Goals and Recommended Actions," Tech. rep., Partnership for AiR Transportation Noise and Emissions Reduction, London, UK, December 2004.
+
+
+
+	
+		Potential to reduce the climate impact of aviation by flight level changes
+		
+			UlrichSchumann
+		
+		
+			KasparGraf
+		
+		
+			HermannMannstein
+		
+		10.2514/6.2011-3376
+	
+	
+		3rd AIAA Atmospheric Space Environments Conference
+		Honolulu, HI
+		
+			American Institute of Aeronautics and Astronautics
+			June 2011
+		
+	
+	Schumann, U., Graf, K., and Mannstein, H., "Potential to reduce the climate impact of aviation by flight level changes," 3rd AIAA Atmospheric Space Environments Conference, AIAA, Honolulu, HI, June 2011.
+
+
+
+	
+		Dynamics of Implementation of Mitigating Measures to Reduce Commercial Aviation's Environmental Impacts
+		
+			RahulKar
+		
+		
+			PhilippeBonnefoy
+		
+		
+			RJohnHansman
+		
+		
+			SgourisSgouridis
+		
+		10.2514/6.2009-6935
+	
+	
+		9th AIAA Aviation Technology, Integration, and Operations Conference (ATIO)
+		Hilton Head, SC
+		
+			American Institute of Aeronautics and Astronautics
+			September 2009
+		
+	
+	Kar, R., Bonnefoy, P., Hansman, R. J., and Sgouridis, S., "Dynamics of Implementation of Mitigating Measures to Re- duce Commercial Aviation's Environmental Impacts," 9th AIAA Aviation Technology, Integration, and Operations Conference (ATIO), AIAA, Hilton Head, SC, September 2009.
+
+
+
+	
+		Use of Hyperspace Trade Analyses to Evaluate Environmental and Performance Tradeoffs for Cruise and Approach Operations
+		
+			MGregoryO'neill
+		
+		
+			Jean-MarieDumont
+		
+		
+			RJohnHansman
+		
+		10.2514/6.2012-5507
+	
+	
+		12th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference and 14th AIAA/ISSMO Multidisciplinary Analysis and Optimization Conference
+		Indianapolis, IN
+		
+			American Institute of Aeronautics and Astronautics
+			September 2012
+		
+	
+	O'Neill, M. G., Dumont, J.-M., and Hansman, R. J., "Use of Hyperspace Trade Analyses to Evaluate Environmental and Performance Tradeoffs for Cruise and Approach Operations," 12th AIAA Aviation Technology, Integration, and Operations Conference (ATIO), AIAA, Indianapolis, IN, September 2012.
+
+
+
+	
+		Call for Papers
+		
+			BSridhar
+		
+		
+			NYChen
+		
+		
+			HKNg
+		
+		10.1027/2192-0923/a000067
+	
+	
+		Aviation Psychology and Applied Human Factors
+		Aviation Psychology and Applied Human Factors
+		2192-0923
+		2192-0931
+		
+			4
+			2
+			
+			June 2013
+			Hogrefe Publishing Group
+			Chicago, IL
+		
+	
+	Tenth USA
+	Sridhar, B., Chen, N. Y., and Ng, H. K., "Energy Efficient Contrail Mitigation Strategies for Reducing the Environmental Impact of Aviation," Tenth USA/Europe Air Traffic Management Research and Development Seminar , Chicago, IL, June 2013.
+
+
+
+	
+		Aircraft Trajectory Design Based on Reducing the Combined Effects of Carbon-Di-Oxide, Oxides of Nitrogen and Contrails
+		
+			BanavarSridhar
+		
+		
+			NeilYChen
+		
+		
+			HokNg
+		
+		10.2514/6.2014-0807
+	
+	
+		AIAA Modeling and Simulation Technologies Conference
+		National Harbor, MD
+		
+			American Institute of Aeronautics and Astronautics
+			Jan 2014
+		
+	
+	Sridhar, B., Chen, N. Y., and Ng, H. K., "Aircraft Trajectory Design Based on Reducing the Combined Effects of Carbon- Dioxide, Oxides of Nitrogen and Contrails," AIAA Modeling and Simulation Technologies (MST) Conference, National Harbor, MD, Jan 2014.
+
+
+
+	
+		Evaluating Tradeoff between Environmental Impact and Operational Costs for Enroute Air Traffic
+		
+			NeilYChen
+		
+		
+			BanavarSridhar
+		
+		
+			HokNg
+		
+		
+			JinhuaLi
+		
+		10.2514/6.2014-1464
+	
+	
+		AIAA Guidance, Navigation, and Control Conference
+		National Harbor, MD
+		
+			American Institute of Aeronautics and Astronautics
+			January 2014
+		
+	
+	Chen, N., Sridhar, B., Ng, H., and Li, J., "Evaluating Tradeoff between Environmental Impact and Operational Costs for Enroute Air Traffic," AIAA Guidance, Navigation and Control Conference, AIAA, National Harbor, MD, January 2014.
+
+
+
+	
+		A note on how to avoid contrail cirrus
+		
+			HermannMannstein
+		
+		
+			PeterSpichtinger
+		
+		
+			KlausGierens
+		
+		10.1016/j.trd.2005.04.012
+	
+	
+		Transportation Research Part D: Transport and Environment
+		Transportation Research Part D: Transport and Environment
+		1361-9209
+		
+			10
+			5
+			
+			September 2005
+			Elsevier BV
+		
+	
+	Mannstein, H., Spichtinger, P., and Gierens, K., "A note on how to avoid contrail cirrus," Transportation Research. Part D, Transport and environment, Vol. 10, No. 5, September 2005, pp. 421-426.
+
+
+
+	
+		Reducing the climate change impacts of aviation by restricting cruise altitudes
+		
+			VictoriaWilliams
+		
+		
+			RobertBNoland
+		
+		
+			RalfToumi
+		
+		10.1016/s1361-9209(02)00013-5
+	
+	
+		Transportation Research Part D: Transport and Environment
+		Transportation Research Part D: Transport and Environment
+		1361-9209
+		
+			7
+			6
+			
+			November 2002
+			Elsevier BV
+		
+	
+	9 Williams, V., Noland, R. B., and Toumi, R., "Reducing the climate change impacts of aviation by restricting cruise altitudes," Transportation Research. Part D, Transport and environment, Vol. 7, No. 6, November 2002, pp. 451-464.
+
+
+
+	
+		Variability of contrail formation conditions and the implications for policies to reduce the climate impacts of aviation
+		
+			VictoriaWilliams
+		
+		
+			RobertBNoland
+		
+		10.1016/j.trd.2005.04.003
+	
+	
+		Transportation Research Part D: Transport and Environment
+		Transportation Research Part D: Transport and Environment
+		1361-9209
+		
+			10
+			4
+			
+			July 2005
+			Elsevier BV
+		
+	
+	Williams, V. and Noland, R. B., "Variability of contrail formation conditions and the implications for policies to reduce the climate impacts of aviation," Transportation Research. Part D, Transport and environment, Vol. 10, No. 4, July 2005, pp. 269-280.
+
+
+
+	
+		Fuel efficient strategies for reducing contrail formations in United States airspace
+		
+			BanavarSridhar
+		
+		
+			NeilYChen
+		
+		10.1109/dasc.2010.5655533
+	
+	
+		29th Digital Avionics Systems Conference
+		Salt Lake City, UT
+		
+			IEEE
+			October 2010
+		
+	
+	Sridhar, B., Chen, N. Y., and Ng, H. K., "Fuel Efficient Strategies for Reducing Contrail Formations in United State National Air Space," 29th Digital Avionics Systems Conference, Salt Lake City, UT, October 2010.
+
+
+
+	
+		Tradeoff Between Contrail Reduction and Emissions in United States National Airspace
+		
+			NeilYChen
+		
+		
+			BanavarSridhar
+		
+		
+			HokKNg
+		
+		10.2514/1.c031680
+	
+	
+		Journal of Aircraft
+		Journal of Aircraft
+		0021-8669
+		1533-3868
+		
+			49
+			5
+			
+			2012
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	Chen, N. Y., Sridhar, B., and Ng, H. K., "Tradeoff between Contrail Reduction and Emissions in United States National Airspace," Journal of Aircraft, Vol. 49, No. 5, 2012, pp. 1367-1375.
+
+
+
+	
+		A Linear Programming Approach to the Development of Contrail Reduction Strategies Satisfying Operationally Feasible Constraints
+		
+			PengWei
+		
+		
+			BanavarSridhar
+		
+		
+			NeilChen
+		
+		
+			DengfengSun
+		
+		10.2514/6.2012-4754
+	
+	
+		AIAA Guidance, Navigation, and Control Conference
+		Minneapolis, MN
+		
+			American Institute of Aeronautics and Astronautics
+			August 2012
+		
+	
+	Wei, P., Sridhar, B., Chen, N., and Sun, D., "A Linear Programming Approach to the Development of Contrail Reduc- tion Strategies Satisfying Operationally Feasible Constraints," AIAA Guidance, Navigation and Control Conference, AIAA, Minneapolis, MN, August 2012.
+
+
+
+	
+		Integration of Linear Dynamic Emission and Climate Models with Air Traffic Simulations
+		
+			BanavarSridhar
+		
+		
+			HokNg
+		
+		
+			NeilChen
+		
+		10.2514/6.2012-4756
+	
+	
+		AIAA Guidance, Navigation, and Control Conference
+		Minneapolis, MN
+		
+			American Institute of Aeronautics and Astronautics
+			August 2012
+		
+	
+	Sridhar, B., Ng, H., and Chen, N., "Integration of Linear Dynamic Emission and Climate Models with Air Traffic Simulations," AIAA Guidance, Navigation and Control Conference, AIAA, Minneapolis, MN, August 2012.
+
+
+
+	
+		Aircraft Trajectory Optimization and Contrails Avoidance in the Presence of Winds
+		
+			BanavarSridhar
+		
+		
+			HokNg
+		
+		
+			NeilChen
+		
+		10.2514/6.2010-9139
+	
+	
+		10th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference
+		Brisbane, Australia
+		
+			American Institute of Aeronautics and Astronautics
+			September 2012
+		
+	
+	Sridhar, B., Ng, H., and Chen, N., "Uncertainty Quantification in the Development of Aviation Operations to Reduce Aviation Emissions and Contrails," 28th International Congress of the Aeronautical Sciences, AIAA, Brisbane, Australia, September 2012.
+
+
+
+	
+		Estimating the Social Cost of Carbon for Use in U.S. Federal Rulemakings: A Summary and Interpretation
+		
+			MichaelGreenstone
+			
+				California Environmental Protection Agency
+			
+		
+		
+			ElizabethKopits
+			
+				California Environmental Protection Agency
+			
+		
+		
+			AnnWolverton
+			
+				California Environmental Protection Agency
+			
+		
+		10.3386/w16913
+	
+	
+		California Air Resources Board Quarterly Auction
+		
+			2
+			2011. 2013
+			National Bureau of Economic Research
+		
+	
+	A. R. B.
+	Greenstone, M., Kopits, E., Wolverton, A., Greenstone, M., Kopits, E., and Wolverton, A., "Estimating the Social Cost of Carbon for Use in U.S. Federal Rulemakings: A Summary and Interpretation," 2011. 17 California Environmental Protection Agency, A. R. B., "California Air Resources Board Quarterly Auction 2," 2013.
+
+
+
+	
+		Consideration of costs and damages avoided and/or benefits gained
+		
+			RJKlein
+		
+		
+			SHuq
+		
+		
+			FDenton
+		
+		
+			TEDowning
+		
+		
+			RGRichels
+		
+		
+			JBRobinson
+		
+		
+			FLToth
+		
+		
+			OCanziani
+		
+		
+			JPalutikof
+		
+		
+			PVan Der Linden
+		
+		
+			CHanson
+		
+	
+	
+		Climate Change 2007: Impacts, Adaptation and Vulnerability Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change
+		
+			MParry
+		
+		Cambridge, United Kingdom and New York, NY
+		
+			Cambridge University Press
+			2007
+		
+	
+	Klein, R. J., Huq, S., Denton, F., Downing, T. E., Richels, R. G., Robinson, J. B., and Toth, F. L., "Consideration of costs and damages avoided and/or benefits gained," In: Climate Change 2007: Impacts, Adaptation and Vulnerability Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, edited by M. Parry, O. Canziani, J. Palutikof, P. van der Linden, and C. Hanson, Cambridge University Press, Cambridge, United Kingdom and New York, NY, 2007.
+
+
+
+	
+		Transport impacts on atmosphere and climate: Shipping
+		
+			VeronikaEyring
+		
+		
+			IvarS AIsaksen
+		
+		
+			TerjeBerntsen
+		
+		
+			WilliamJCollins
+		
+		
+			JamesJCorbett
+		
+		
+			OyvindEndresen
+		
+		
+			RoyGGrainger
+		
+		
+			JanaMoldanova
+		
+		
+			HansSchlager
+		
+		
+			DavidSStevenson
+		
+		10.1016/j.atmosenv.2009.04.059
+	
+	
+		Atmospheric Environment
+		Atmospheric Environment
+		1352-2310
+		
+			44
+			37
+			
+			2010
+			Elsevier BV
+		
+	
+	Eyring, V., Isaksen, I. S., Berntsen, T., Collins, W. J., Corbett, J. J., Endresen, O., Grainger, R. G., Moldanova, J., Schlager, H., and Stevenson, D. S., "Transport impacts on Atmosphere and Climate: Metrics," Atmosphere Environment, Vol. 44, No. 37, 2010, pp. 4648-4677.
+
+
+
+	
+		Anthropogenic and Natural Radiative Forcing pages 705 to 740
+		
+			PForster
+		
+		
+			VRamaswamy
+		
+		
+			PArtaxo
+		
+		
+			TBerntsen
+		
+		
+			RBetts
+		
+		
+			DFahey
+		
+		
+			JHaywood
+		
+		
+			JLean
+		
+		
+			DLowe
+		
+		
+			GMyhre
+		
+		
+			JNganga
+		
+		
+			RPrinn
+		
+		
+			GRaga
+		
+		
+			MSchulz
+		
+		
+			RVDorland
+		
+		
+			MManning
+		
+		
+			ZChen
+		
+		
+			MMarquis
+		
+		
+			KAveryt
+		
+		
+			MTignor
+		
+		
+			HMiller
+		
+		10.1017/cbo9781107415324.019
+	
+	
+		Climate Change 2013 - The Physical Science Basis
+		
+			SSolomon
+		
+		
+			DQin
+		
+		Cambridge, United Kingdom and New York, NY
+		
+			Cambridge University Press
+			2007. 2007
+			
+		
+	
+	Forster, P., Ramaswamy, V., Artaxo, P., Berntsen, T., Betts, R., Fahey, D., Haywood, J., Lean, J., Lowe, D., Myhre, G., Nganga, J., Prinn, R., Raga, G., Schulz, M., and Dorland, R. V., "Changes in Atmospheric Constituents and in Radiative Forcing," In: Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, edited by S. Solomon, D. Qin, M. Manning, Z. Chen, M. Marquis, K. Averyt, M.Tignor, and H. Miller, Cambridge University Press, Cambridge, United Kingdom and New York, NY, 2007.
+
+
+
+	
+		Seeing through contrails
+		
+			OlivierBoucher
+		
+		10.1038/nclimate1078
+	
+	
+		Nature Climate Change
+		Nature Clim Change
+		1758-678X
+		1758-6798
+		
+			1
+			1
+			
+			2011
+			Springer Science and Business Media LLC
+		
+	
+	Boucher, O., "Atmospheric science: Seeing through contrails," Nature Climate Change, Vol. 1, 2011, pp. 24-25.
+
+
+
+	
+		Potential to reduce the climate impact of aviation by flight level changes
+		
+			UlrichSchumann
+		
+		
+			KasparGraf
+		
+		
+			HermannMannstein
+		
+		10.2514/6.2011-3376
+		AIAA-2011-3376
+	
+	
+		3rd AIAA Atmospheric Space Environments Conference
+		Honolulu, HI
+		
+			American Institute of Aeronautics and Astronautics
+			June 2011
+		
+	
+	Schumann, U., Graf, K., and Mannstein, H., "Potential to Reduce the Climate Impact of Aviation by Flight Level Changes," AIAA Modeling and Simulation Technologies Conference, AIAA-2011-3376, AIAA, Honolulu, HI, June 2011.
+
+
+
+	
+		Latitudinal variation of the effect of aviation NOx emissions on atmospheric ozone and methane and related climate metrics
+		
+			MOKöhler
+		
+		
+			GRädel
+		
+		
+			KPShine
+		
+		
+			HLRogers
+		
+		
+			JAPyle
+		
+		10.1016/j.atmosenv.2012.09.013
+	
+	
+		Atmospheric Environment
+		Atmospheric Environment
+		1352-2310
+		
+			64
+			
+			2013
+			Elsevier BV
+		
+	
+	Kohler, M., Radel, G., Shine, K., H.L.Rogers, and Pyle, J., "Latitudinal variation of the effect of aviation NO emissions on atmospheric ozone and methane and related climate metrics," Atmosphere Environment, Vol. 64, 2013, pp. 1-9.
+
+
+
+	
+		Transport impacts on atmosphere and climate: Metrics
+		
+			JSFuglestvedt
+		
+		
+			KPShine
+		
+		
+			TBerntsen
+		
+		
+			JCook
+		
+		
+			DSLee
+		
+		
+			AStenke
+		
+		
+			RBSkeie
+		
+		
+			GJ MVelders
+		
+		
+			IAWaitz
+		
+		10.1016/j.atmosenv.2009.04.044
+	
+	
+		Atmospheric Environment
+		Atmospheric Environment
+		1352-2310
+		
+			44
+			37
+			
+			2010. May 2013
+			Elsevier BV
+		
+	
+	Tech. rep.
+	Atmospheric Environment
+	Fuglestvedt, J., Shine, K., Berntsen, T., Cook, J., Lee, D., Stenke, A., Skeie, R., Velders, G., and Waitz, I., "Transport impacts on atmosphere and climate: Metrics," Atmospheric Environment, Vol. 44, No. 37, 2010, pp. 4648 -4677. 25 United States Government, "Technical Support Document: -Technical Update of the Social Cost of Carbon for Regulatory Impact Analysis," Tech. rep., Under Executive Order 12866, Interagency Working Group on Social Cost of Carbon, May 2013.
+
+
+
+	
+		Evaluating Tradeoff between Environmental Impact and Operational Costs for Enroute Air Traffic
+		
+			NeilYChen
+		
+		
+			BanavarSridhar
+		
+		
+			HokNg
+		
+		
+			JinhuaLi
+		
+		10.2514/6.2014-1464
+	
+	
+		AIAA Guidance, Navigation, and Control Conference
+		Minneapolis, MN
+		
+			American Institute of Aeronautics and Astronautics
+			August 2012
+		
+	
+	Chen, N., Sridhar, B., Li, J., and Ng, H., "Evaluating Tradeoff between Environmental Impact and Operational Costs for Enroute Air Traffic," AIAA Guidance, Navigation and Control Conference, AIAA, Minneapolis, MN, August 2012.
+
+
+
+	
+		Evaluation of Aircraft Contrails using Dynamic Dispersion Model
+		
+			JinhuaLi
+		
+		
+			FabioCaiazzo
+		
+		
+			NeilYChen
+		
+		
+			BanavarSridhar
+		
+		
+			HokNg
+		
+		
+			StevenBarrett
+		
+		10.2514/6.2013-5178
+	
+	
+		AIAA Guidance, Navigation, and Control (GNC) Conference
+		Boston, MA
+		
+			American Institute of Aeronautics and Astronautics
+			August 2013
+		
+	
+	Li, J., Caiazzo, F., Chen, N. Y., Sridhar, B., Ng, H., and Barrett, S., "Evaluation of Aircraft Contrails using Dynamic Dispersion Model," AIAA Guidance, Navigation and Control Conference, AIAA, Boston, MA, August 2013.
+
+
+				
+			
+		
+	

file133.txt

@@ -0,0 +1,432 @@
+
+	
+	
+		
+I. IntroductionT O ENSURE smooth air traffic flow and safety in the presence of disruptions caused by uncertainties, innovative modeling and design methods are needed in traffic flow management.One of the main functions of traffic flow management is to predict and resolve demand-capacity imbalances at the sector level.Thus, an accurate sector prediction model that can account for traffic flow uncertainty and weather impact is an essential component of traffic flow management.Efforts have been made in the past to perform sector-demand predictions.Traditionally, models used in air traffic control and flow management are based on simulating the trajectories of individual aircraft.Deterministic forecasting of sector demand is routinely done within the enhanced traffic management system (ETMS), which relies on the computation of each aircraft's entry and exit times at each sector along the path of flight.Gilbo and Smith [1] proposed, acknowledging the uncertainty in the predictions, a regression model for improving aggregate traffic demand prediction in ETMS.A more recent traffic flow management simulation tool, the Future Automation Concepts Evaluation Tool (FACET) [2], was used to propagate the trajectories of the proposed flights forward in time and use them to count the number of aircraft in each sector for demand forecasting and establish confidence bounds on the forecasts [3].These trajectory-based models predict the behavior of the National Airspace System adequately for short durations of up to 20 min and their accuracy is impacted by weather and trajectory prediction uncertainties [4][5][6].In addition, these prediction models are openloop, which means the traffic flow management (TFM) actions are not accounted in the models; therefore, the prediction does not reflect the actual sector demand after the TFM management actions.The objective of this paper is to develop an empirical sectordemand prediction model that accounts for TFM actions, including air traffic control and airline actions, and that accounts for both shortterm (less than 30 min) and midterm (30 min to 2 h) predictions.The model consists of two parts: the open-loop prediction and the TFM action model.The open-loop predictions, similar to the traditional methods, are used to determine the possibility of demand-capacity imbalances at a future time, and help decide whether to activate the TFM action.The TFM action model simulates the demand reduction caused by the TFM actions.The closed-loop prediction represents the net result of the open-loop prediction and the TFM actions.The periodic autoregressive model and its variants [7,8] were used to build the model.The model considers both historical traffic flows to capture the midterm trend and flows in the near past to capture the transient response.In addition, for severe weather cases, the weatherimpacted TFM action was modeled using weather forecast information.The proposed model provides both open-and closed-loop sector-demand predictions.Open-loop prediction is adequate for short durations.When looking at predictions for long durations, open-loop models produce large errors due to their inability to capture traffic initiatives and airline actions during the planning period.A combination of closed-loop and open-loop models provide decision-makers the full range of traffic behavior.The remainder of the paper is organized as follows.Section II provides the sector-demand data and a description of the open-and closed-loop sector-demand prediction models.Next, in Sec.III, a weather factor is introduced and the TFM action model that considers weather is described.The results and performance of the models are demonstrated in Sec.IV.Finally, a summary and conclusions are presented in Sec.V.
+II. Data and Model
+A. Sector-Demand DataThe air traffic demand data were recorded from the Aircraft Situation Display to Industry (ASDI) data generated by the Federal Aviation Administration's ETMS.The ASDI data provide the locations of all aircraft at 1 min intervals.The sector demand, defined as the number of aircraft in each sector at a given time, can be computed using the ASDI data.Since traffic flow management decisions are made by comparing the peak number of aircraft in a sector during a 15 min interval with the sector's monitor alert parameter (MAP) value, the 15 min peak sector demand was used to build the models.A day is defined as a 24 h interval starting at 0400 hrs local time, since most of the aircraft departing on the previous day would have landed before 0400 hrs.The  The average trend of sector demand on different days can be observed in Fig. 1, which shows the variation of 15 min peak sector demand in September 2007.In this figure, each horizontal strip represents one day of 15 min peak sector demand, and each vertical strip represents the peak sector demand at the same time of day during the entire month.As shown, the horizontal strips on 1 September, 8 September, 15 September, 22 September, and 29 September, which are Saturdays, have lower demands than the others.The blue vertical regions on the left and right show the offpeak traffic in the early morning and the late night.A vertical light blue region at around 1200 hrs divides the sector demand into a morning rush left of the region and an afternoon peak right of it.The sector-demand prediction model presented in the next section captures these variations in the demand.
+B. Demand Prediction ModelSector demand, defined as the number of aircraft in a sector, is the result of planned inflow and outflow and TFM actions.Figure 2a shows the block diagram of the current sector-demand system, where d k is the sector demand at the kth time step and d kp is the sector demand at the (k p)th time step.In the system, the traffic flow manager monitors the sector-demand prediction based on enhanced traffic management system (ETMS), denoted as dETMS kp ; if the prediction is high, TFM is activated to reduce the demand in the sector.The top half of the diagram, shown in the dashed box, is considered as an open loop; the bottom half, with the TFM action, is considered as a feedforward loop with negative gain.In the sectordemand prediction model, shown in Fig. 2b,fd open kp f open k;p d 1 . . . d k e open k d kp d open kp f TFM k;p d open kp e TFM k (1)To implement the prediction model in Eq. ( 1), f open k;p and f TFM k;p need to be identified using historical data.In reality, it is not possible to identify the open-loop sector demand when TFM is in action because of the absence of data to verify the validity of the models during high demand.However, the open-loop model can be identified using data during low demand, since no TFM action is involved.With the assumption that the behavior of open-loop models are similar during low-and high-demand periods, the open-loop prediction model validated for low demand is also used during high demand.
+C. Periodic Autoregressive Sector-Demand ModelAutoregressive models have been used for short-term hourly air traffic delay prediction [9,10].This research extends the delay prediction approach to open-loop sector-demand prediction.The TFM action model is incorporated in the prediction model and can be identified once the open-loop model is identified.A 24 h period, starting at 0400 hrs local time, is divided into 96 15 min intervals.Given the observed 15 min peak sector demands for n days, the sector-demand data matrix is defined as   can then be solved explicitly [11].For high-demand cases, TFM action is active.The action is modeled as a negative linear feedforward gain based on the open-loop prediction and the threshold, formulated asD d 1;d kp k;p d k k;p k;p k;p d k k;p d threshold e k (4)where k;p and k;p are the least-squares solution of Eq. ( 3) using low-demand data, k;p is the feedforward gain, and e k is the error of the model.Note that k;p is equal to zero for low-demand cases.With k;p and k;p known, the least-squares solution of k;p for highdemand cases, denoted as k;p , can be solved explicitly using highdemand data.On a day m other than the n days in the data set, the p-step prediction of the sector demand at the kth time step, dkp;m , based on the observed sector demand, d k;m , can then be expressed as  4) and ( 5) is referred to as the periodic autoregressive (PAR) sector-demand prediction model.dopenAs an example, peak sector-demand data in August 2007 were used to construct the data matrix in Eq. ( 2).Equations ( 3) and ( 4) were used to identify the model parameters k;p , k;p , and k;p , where k 1; . . .; 96 and p 1 for one step, or 15-min-ahead prediction.The peak sector demands on 3 September 2007 were predicted using Eq. ( 5).The prediction results for sector ZID93 are shown in Fig. 3.The black dots represent the sector demand in a 1 min interval, the blue line represents the 15 min peak sector demand, the green line represents the 15-min-ahead sector-demand prediction, and the red line is the MAP value.The root-mean-squared (rms) error between the actual peak sector demand and the 15 min demand prediction for the day is 1.96.The rms error during the hours that most aircraft fly, 0700 to 2300 hrs EDT, is 2.23.The rms error when the demand is great than 50% of the MAP value is 2.63.The model can be extended by using the cumulative sum of the past sector demands as an observation instead of using a single observation d k;m in Eq. ( 5), since the sum includes more information than a single observation and has less noise compared with the single peak sector demand.Following the definition of the sector-demand matrix D in Eq. ( 2), where d k is the kth column of D, the cumulative p-step-ahead sector-demand model at time step k can be described in terms of the cumulative sum of q past sector demands asd
+III. Weather FactorWeather has a big influence on air traffic sector demand and the uncertainty in weather may cause error in the predictions [5,12].If a severe storm blocks a sector or regions near it, the sector capacity may drop dramatically, causing the TFM in action to reduce the sector demand [13,14].A weather factor that models the TFM action on severe weather days in the sector-demand prediction is derived in this section.To model the weather impact on TFM action, an accurate weather forecast product with a high update rate is required.In addition, to capture the impact on all low, high, and superhigh sectors, the storm echo tops information is useful.The weather data used in this paper was provided by the Corridor Integrated Weather System (CIWS) [15], which provides both accurate precipitation and echo tops data and is updated every 5 min.In addition, CIWS provides precipitation and echo tops forecasts at 5 min intervals up to 2 h in the future.The weather factor used to model the TFM action was chosen to be the sector weather index, defined as the percentage of area covered by the storm with precipitation vertically integrated liquid (VIL) level 3 and above.Only storms with the echo tops above the lower boundary of the sector are considered.The sector weather index at time k is formulated asw k A w k A (8)where A is the area of the sector and A w k is the area of the sector covered by storms with the echo tops at or above the lower bound of the sector.The sector weather index is a number between 0 and 1 and is often expressed in terms of a percent in the figures in this paper.Note that if time k is a future time, the weather forecast is used to determine A w k .It is possible to use other definitions of a sector weather index [13,14].Figure 4a shows a snap shot of the CIWS data for the high-altitude sectors at Indianapolis center (ZID) on a severe weather day.The red spots indicate the storms with VIL level 3 and above, and the echo tops at 35,000 ft.As shown in this figure, most of the sector ZID93 is covered by the storm.The sector weather index for ZID93 on 16 August 2007 is shown in the red curve in Fig. 4b.Also shown is the actual sector demand on the same day in the blue curve.Note that the sector weather index is greater than 30% from 1800 to 2000 hrs Eastern Daylight Time (EDT), and the sector demand clearly drops during the same period.Traffic reduction due to weather impact can be modeled in many different ways [16].In this research, the open-loop prediction was first estimated, and then the prediction was adjusted by the TFM action based on the sector weather index.Assume that the TFM action is active when the sector weather factor exceeds w low , and TFM blocks out the entire sector when the weather factor reaches w high .The sector-demand reduction rate is modeled as the power law distribution, 1 w k w low =w high w low , where is the power of the distribution.To reflect the thresholds, the sector weather index in Eq. ( 8) is redefined asw k 8 < : w low if A w k =A w low A w k =A if w low < A w k =A < w high w high if w high A w k =A (9)To model the TFM action on the sector-demand prediction model, the weather forecast is used to compute the predicted sector weather index.Assume at time k, the predicted sector weather index at time k p is w kp , the PAR sector-demand prediction model in Eq. ( 7) can be rewritten as    (10) Using the echo tops information provides a more representative weather index, especially for the high sectors.If there are storms with low echo tops located at some high sectors, the weather might have minimal impact on the sector demand.The sector demand and weather index for sector ZID93 on two different days is shown in Fig. 5.Both days have severe storms, but one has high echo tops, while the other has low echo tops.The sector demands on severe weather days were compared with the average sector demand on the rest of the days in the same month.In Fig. 5a, the sector demand on 16 August 2007 is lower than the average between 1800 and 2000 hrs EDT because of the high weather index during the period, as indicated in Fig. 5c.The blue line in Fig. 5c shows the weather index considering the area covered by storms without the echo tops information, and the red line is the weather index considering the echo tops at 35,000 ft and above.In this case, the two lines are close.This suggests that there are severe storms in the area and most of the echo tops are higher than the lower bound of sector ZID93.On the other hand, on 23 October 2007, there is no demand reduction compared to the average of the other days in October 2007 during 1800 and 2000 hrs EDT, shown in Fig. 5b, even though there are storms in the sector during the period, as shown in Fig. 5d.The red line in Fig. 5d is substantially lower than the blue line, which means even though there are storms in the sector, most the echo tops are lower than the low boundary of the sector and have minor impact on the sector demand.In the next section, the sector weather index refers to the index with the echo tops information.
+IV. ResultsThe sector demands of 25 high and superhigh sectors in ZID were investigated in this research.The sector demands for the month of August 2007 were used to build the PAR sector-demand prediction model, described in Eqs. ( 2) and (4).The time step used in the models is 15 min.Once the parameters were identified, Eq. ( 7) was used to perform the sector-demand prediction for the month of September 2007.Starting from the 15 min prediction model, up to 2 h prediction model were built and evaluated.The results of four superhigh sectors ZID91, ZID92, ZID93, and ZID94, and four high sectors ZID81, ZID82, ZID83, and ZID84 in the southwest region of ZID were presented.The prediction results for the eight sectors are summarized in Table 1.Only the errors from 0700 to 2300 hrs were computed.The results include open-loop predictions on low-demand days, when TFM is inactive, and closed-loop predictions when TFM is activated.Note that the errors of the PAR model are not sensitive to the look ahead time.In general, the errors are larger with longer look ahead time, but only slightly.The errors of the 120 min prediction is 2.97% larger than the 15 min prediction on average.For all the high and superhigh sector in ZID, the results are similar.The errors are between 1.77 and 2.44 for the 15 min prediction, and between 1.82 and 2.56 for the 120 min prediction.Even though the differences between the errors are small, the same trends hold for the majority of sectors tested.When the predicted sector-demands are lower than the demand threshold d threshold , defined as sector MAP value subtracted by 4, the TFM actions are inactive so the model is open-loop.When the predicted demand is higher than d threshold , TFM actions are activated so the closed-loop predictions are computed.Among the sectors tested, the TFM actions in the model are more active in ZID81 and ZID93, as more occurrences of TFM actions were triggered.The prediction errors of open-loop predictions on low-demand days, and closed-loop prediction at ZID81 and ZID93 are summarized in Table 2.The sector-demand prediction for bad weather days uses the weather factor described in the previous section to model the TFM action, formulated in Eqs. ( 9) and ( 10), with w low 0, w high 1, and 1.The days with peak weather factors greater than 30% are considered bad weather days.For the days and sectors tested, there are four cases of severe weather periods: ZID83 on 16 August 2007 between 1600-2200 hrs EDT, ZID93 on 16 August 2007 between 1600-2200 hrs EDT, ZID82 on 21 August 2007 between 0600-1400 hrs EDT, and ZID92 on 21 August 2007 between 0800-1400 EDT, shown in Fig. 6.Since all these cases happened in August 2007, the model is built using data for July 2007.Two types of weatherimpacted TFM action models are built: one uses the actual weather information and the other uses the forecast weather information.Using the actual weather information represents the cases with perfect weather forecast.It is used to evaluate the impact of weather forecast accuracy on the model.The average closed-loop prediction errors of the four severe weather periods in August 2007 are shown in Fig. 7.It is noted that in all four cases, both the model using actual weather information (red dashed line) and the model using forecast weather (green dasheddotted line) produce smaller errors than the open-loop model (blue solid line).The model using forecast weather performs as well as the model using actual weather when the prediction time is small (less than 30 min).However, with longer prediction time (more than 60 min), the performance starts to decay and the errors are closer to the open-loop model.As an example, in Fig. 7b, the closed-loop sector-demand prediction model using actual weather information improves the 15 min prediction over the open-loop model by 36%, the 60 min prediction by 43%, and the 120 min prediction by 41%.For the model using forecast weather, the improvement is 37% for the 15 min prediction, 44% for the 60 min prediction, and down to 23% for the 120 min prediction.This suggests that with longer prediction time, the forecast inaccuracy might effect the performance of the TFM action model, resulting in larger error in the prediction model.open k;p is the open-loop prediction model and dopen kp is the open-loop prediction, which is used to determine whether to activate the TFM action.When dopen kp is high, TFM is active.f TFM k;p is the model of the TFM action based on the open-loop prediction.d open k is the actual open-loop sector demand, which is the sum of dopen kp and the open-loop prediction error,e open k .e TFM k is the error of the TFM action model.The model in Fig. 2b can be formulated as
+Fig. 1 Fig. 212Fig. 1 Fifteen-minute peak sector demand at sector ZID93 in September 2007.
+solution of k;p and k;p that minimizes e open k T e open k
+Fig. 33Fig.3Sector demand and peak sector demand at sector ZID93 on 3 September 2007.
+Fig. 44Weather data, sector demand, and weather index on a severe weather day.
+Fig. 55Fig. 5 Sector demand and weather indices with and without echo tops information on 16 August and 23 October 2007.
+15 min peak sector demand, denoted as d k , where k 1; . . .; 96.Presented as Paper 2009-6195 at the AIAA Guidance, Navigation, andControl Conference, Chicago, IL, 10-13 August 2009; received 27 August2009; revision received 2 August 2010; accepted for publication 5 August2010. This material is declared a work of the U.S. Government and is notsubject to copyright protection in the United States. Copies of this paper maybe made for personal or internal use, on condition that the copier pay the$10.00 per-copy fee to the Copyright Clearance Center, Inc., 222 RosewoodDrive, Danvers, MA 01923; include the code 0731-5090/10 and $10.00 incorrespondence with the CCC.Senior Scientist for Air Transportation Systems, Aviation SystemsDivision, Mail Stop 210-10. Fellow AIAA.* Research Aerospace Engineer, Systems Modeling and Optimization Branch, Mail Stop 210-10.Member AIAA.†
+in 2007 Hour of the dayFor lowdemand time periods, TFM is inactive; therefore, open-loop demand is the same as actual demand.A sector-demand threshold d threshold , usually a small number lower than the sector MAP value, is used to define whether the demand is high or low.The demand is classified as high when d threshold > 0 and low when d threshold 0. Consider the sector demands that satisfy d kp;j d threshold , the least-squaresDate ← MAP4567892 6 4. . .1 d 2;1 . . . d 96;1 . . . . . . . . .3 7 5(2)d 1;n d 2;n . . . d 96;nwhere d i;j represents the 15 min peak sector demand at time step i on day j.For September 2007, D has a dimension of 30 by 96, and Fig.1shows the image of the matrix D. Letting d k be the kth column of D, the p-step-ahead open-loop sector-demand prediction model at the kth time step can be described in the form of a first-order periodic autoregressive model:d open kp k;p d k k;p e open k (3)where k;p and k;p are the coefficients that map the sector demand at time k to the open-loop sector demand at time k p.
+10 11 12 13 14 15 16 17 18 19 20 21 22 239/19/3209/59/79/9 9/11159/139/159/17109/199/219/23 9/2559/279/2901230
+and k;p are the coefficients that map the cumulative sector demand at the kth time step to the sector demand at the (k p) th time step, and k;p is the TFM action gain.Once the least-squares solution of coefficients k;p , k;p , and k;p are identified, the p-step prediction of the sector demand at the kth time step for a day m, dkp;m , based on the observed cumulative sector demand,kp k;pX kd i k;pik q1k;pk;pX kd i k;p d thresholde k(6)ik q1where k;p X kd i;mik q1can be expressed asdopen kp;m k;pX kd i;m k;pik q1dkp;mdopen kp;mk;pdopen kp;md threshold(7)
+Table 11Sector-demand prediction errors of the PAR model in September 2007 (the unit is the number of aircraft)SectorAverage prediction rms error from 0700 to 2300 hrs EDTName MAP 15 min 30 min 45 min 60 min 90 min 120 minZID81172.202.302.312.312.292.31ZID82161.771.821.841.771.801.82ZID83161.811.831.841.831.841.85ZID84162.092.132.102.122.102.07ZID91192.342.422.432.392.432.46ZID92171.921.981.951.961.981.99ZID93192.442.552.542.522.592.56ZID94172.192.262.272.232.242.23
+Table 22Open-and closed-loop sector-demand prediction errors of the PAR model in September 2007 (the unit is the number of aircraft)SectorAverage prediction rms error from 0700 to 2300 hrs EDTNameMAPType15 min30 min45 min60 min90 min120 minZID8117Open2.202.302.332.322.302.25ZID8117Closed2.202.302.312.312.292.31ZID9319Open2.392.502.502.502.532.56ZID9319Closed2.442.552.542.522.592.56
+		
+		
+			
+V. ConclusionsA class of periodic autoregressive (PAR) models with management-action-embedded for sector-demand prediction is used for predicting air traffic demand in a sector between 15 min and 2 h in the future.The open-loop model was first identified using lowdemand data, assuming no traffic flow management (TFM) action, then the TFM action model was identified using high-demand data.The closed-loop model is the net result of the open-loop and the TFM action models.The proposed PAR model captures both the midterm trend based on the historical data and the short-term transient response based on the near-past observation.For the sectors tested, the model provides the demand predictions with an average rootmean-squared (rms) error between 1.77 and 2.44 in the 15 min prediction and between 1.82 and 2.56 in the 120 min prediction.The performance of the prediction only decays slightly as the prediction interval is increased from 15 min to 2 h, with an error increase of 2.97%.For the sector-demand prediction in the presence of severe weather, the paper introduced the concept of a weather factor to model the TFM actions.For severe weather days, the model uses the storm precipitation and echo tops to form the TFM action model using the weather factor and then adjusts the open-loop prediction.The model improves the closed-loop sector-demand prediction by as much as 37% for the 15 min prediction, 44% for the 60 min prediction, and 23% for the 120 min prediction on the days and sectors tested.In addition to traditional trajectory-based sector-demand prediction methods that predict only the open-loop behavior of the National Airspace System adequately for short durations of up to 20 min and are vulnerable to weather uncertainties, the managementembedded PAR models provide a reliable short-to midterm (both open-and closed-loop) sector-demand prediction that accounts for non-weather-and weather-impacted TFM actions.A combination of closed-loop and open-loop models provide decision-makers with the full range of traffic behavior.			
+			
+
+				
+
+
+	
+		A New Model to Improve Aggregate Air Traffic Demand Predictions
+		
+			EugeneGilbo
+		
+		
+			ScottSmith
+		
+		10.2514/6.2007-6450
+	
+	
+		AIAA Guidance, Navigation and Control Conference and Exhibit
+		Hilton Head, SC
+		
+			American Institute of Aeronautics and Astronautics
+			Aug. 2007
+		
+	
+	A New Model to Improve Aggregate Air Traffic Demand Predictions
+	Gilbo, E., and Smith, S., "A New Model to Improve Aggregate Air Traffic Demand Predictions," AIAA Guidance, Navigation and Control Conference, AIAA Paper 2007-6450, Hilton Head, SC, Aug. 2007.
+
+
+
+	
+		FACET: Future ATM Concepts Evaluation Tool
+		
+			KarlDBilimoria
+		
+		
+			BanavarSridhar
+		
+		
+			ShonRGrabbe
+		
+		
+			GanoBChatterji
+		
+		
+			KapilSSheth
+		
+		10.2514/atcq.9.1.1
+	
+	
+		Air Traffic Control Quarterly
+		Air Traffic Control Quarterly
+		1064-3818
+		2472-5757
+		
+			9
+			1
+			
+			2001
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	Bilimoria, K., Sridhar, B., Chatterji, G. B., Sheth, K., and Grabbe, S., "FACET: Future ATM Concepts Evaluation Tool," Air Traffic Control Quarterly, Vol. 9, No. 1, 2001, pp. 1-20.
+
+
+
+	
+		Methods for Establishing Confidence Bounds on Sector Demand Forecasts
+		
+			GanoChatterji
+		
+		
+			BanavarSridhar
+		
+		
+			KapilSheth
+		
+		
+			DouglasKim
+		
+		
+			DanielMulfinger
+		
+		10.2514/6.2004-5232
+		AIAA Paper 2004-5232
+	
+	
+		AIAA Guidance, Navigation, and Control Conference and Exhibit
+		Providence, RI
+		
+			American Institute of Aeronautics and Astronautics
+			Aug. 2004
+		
+	
+	Chatterji, G. B., Sridhar, B., Sheth, K., Kim, D., and Mulfinger, D., "Methods for Establishing Confidence Bounds on Sector Demand Forecasts," AIAA Guidance, Navigation and Control Conference, AIAA Paper 2004-5232, Providence, RI, Aug. 2004.
+
+
+
+	
+		Compressive Representations of Weather Scenes for Strategic Air Traffic Flow Management
+		
+			JEEvans
+		
+		10.2514/6.2022-4079.vid
+		
+			Dec. 2001
+			American Institute of Aeronautics and Astronautics (AIAA)
+			Santa Fe, NM
+		
+	
+	4th USA/ Europe Air Traffic Management R&D Seminar
+	Evans, J. E., "Tactical Weather Decision Support to Complement Strategic Traffic Flow Management for Convective Weather," 4th USA/ Europe Air Traffic Management R&D Seminar, Santa Fe, NM, Dec. 2001.
+
+
+
+	
+		Measuring Uncertainty in Airspace Demand Predictions for Traffic Flow Management Applications
+		
+			CraigWanke
+		
+		
+			MichaelCallaham
+		
+		
+			DanielGreenbaum
+		
+		
+			AnthonyMasalonis
+		
+		10.2514/6.2003-5708
+		AIAA Paper 2003-5708
+	
+	
+		AIAA Guidance, Navigation, and Control Conference and Exhibit
+		Austin, TX
+		
+			American Institute of Aeronautics and Astronautics
+			Aug. 2003
+		
+	
+	Wanke, C. R., Callaham, M. B., Greenbaum, D. P., and Masalonis, A. J., "Measuring Uncertainty in Airspace Demand Predictions for Traffic Flow Management Applications," AIAA Guidance, Navigation and Control Conference, AIAA Paper 2003-5708, Austin, TX, Aug. 2003.
+
+
+
+	
+		Modeling Traffic Prediction Uncertainty for Traffic Management Decision Support
+		
+			CraigWanke
+		
+		
+			SandeepMulgund
+		
+		
+			DanielGreenbaum
+		
+		
+			LixiaSong
+		
+		10.2514/6.2004-5230
+		AIAA Paper 2004- 5230
+	
+	
+		AIAA Guidance, Navigation, and Control Conference and Exhibit
+		Providence, RI
+		
+			American Institute of Aeronautics and Astronautics
+			Aug. 2004
+		
+	
+	Wanke, C. R., Mulgund, S., and Song, L., "Modeling Traffic Prediction Uncertainty for Traffic Management Decision Support," AIAA Guidance, Navigation and Control Conference, AIAA Paper 2004- 5230, Providence, RI, Aug. 2004.
+
+
+
+	
+		
+			LLjung
+		
+		System Identification: Theory for the User
+		Englewood Cliffs, NJ
+		
+			Prentice Hall
+			1999
+			2
+			
+		
+	
+	nd ed.
+	Ljung, L., System Identification: Theory for the User, 2nd ed., Prentice Hall, Englewood Cliffs, NJ, 1999, pp. 79-93.
+
+
+
+	
+		
+			PFranses
+		
+		
+			RPapp
+		
+		Periodic Time Series Models
+		London, UK
+		
+			Oxford Univ. Press
+			2003
+			
+		
+	
+	Franses, P., and Papp, R., Periodic Time Series Models, Oxford Univ. Press, London, UK, 2003, pp. 27-60.
+
+
+
+	
+		Short Term National Airspace System Delay Prediction Using Weather Impacted Traffic Index
+		
+			BanavarSridhar
+		
+		
+			NeilChen
+		
+		10.2514/6.2008-7395
+		AIAA Paper 2008-7395
+	
+	
+		AIAA Guidance, Navigation and Control Conference and Exhibit
+		Honolulu, HI
+		
+			American Institute of Aeronautics and Astronautics
+			Aug. 2008
+		
+	
+	Sridhar, B., and Chen, N., "Short Term National Airspace System Delay Prediction Using Weather Impacted Traffic Index," AIAA Guidance, Navigation and Control Conference, AIAA Paper 2008-7395, AIAA, Honolulu, HI, Aug. 2008.
+
+
+
+	
+		Estimation of Air Traffic Delay Using Three Dimensional Weather Information
+		
+			NeilChen
+		
+		
+			BanavarSridhar
+		
+		10.2514/6.2008-8916
+		AIAA Paper 2008- 8916
+	
+	
+		The 26th Congress of ICAS and 8th AIAA ATIO
+		Anchorage, AK
+		
+			American Institute of Aeronautics and Astronautics
+			Sept. 2008
+		
+	
+	Chen, N., and Sridhar, B., "Estimation of Air Traffic Delay Using Three Dimensional Weather Information," The 8th AIAA Aviation Tech- nology, Integration, and Operations Conference, AIAA Paper 2008- 8916, AIAA, Anchorage, AK, Sept. 2008.
+
+
+
+	
+		Weather-Weighted Periodic Auto Regressive Models for Sector Demand Prediction
+		
+			NeilChen
+		
+		
+			BanavarSridhar
+		
+		10.2514/6.2009-6195
+		AIAA Paper 2009-6195
+	
+	
+		AIAA Guidance, Navigation, and Control Conference
+		Chicago, IL
+		
+			American Institute of Aeronautics and Astronautics
+			Aug. 2009
+		
+	
+	Chen, N., and Sridhar, B., "Weather-Weighted Periodic Auto Regressive Models for Sector Demand Prediction," AIAA Guidance, Navigation and Control Conference, AIAA Paper 2009-6195, Chicago, IL, Aug. 2009.
+
+
+
+	
+		Analysis of En Route Sector Demand Error Sources
+		
+			JimmyKrozel
+		
+		
+			DanRosman
+		
+		
+			ShonGrabbe
+		
+		10.2514/6.2002-5016
+		AIAA Paper 2002-5016
+	
+	
+		AIAA Guidance, Navigation, and Control Conference and Exhibit
+		Monterey, CA
+		
+			American Institute of Aeronautics and Astronautics
+			Aug. 2002
+		
+	
+	Krozel, J., Rosman, D., and Grabbe, S., "Analysis of En Route Sector Demand Error Sources," AIAA Guidance, Navigation and Control Conference, AIAA Paper 2002-5016, Monterey, CA, Aug. 2002.
+
+
+
+	
+		Predicting Sector Capacity under Severe Weather Impact for Traffic Flow Management
+		
+			LixiaSong
+		
+		
+			CraigWanke
+		
+		
+			DanielGreenbaum
+		
+		
+			DavidCallner
+		
+		10.2514/6.2007-7887
+	
+	
+		7th AIAA ATIO Conf, 2nd CEIAT Int'l Conf on Innov and Integr in Aero Sciences,17th LTA Systems Tech Conf; followed by 2nd TEOS Forum
+		Belfast, Northern Ireland
+		
+			American Institute of Aeronautics and Astronautics
+			Sept. 2007
+		
+	
+	Song, L., Wanke, C., Greenbaum, D., and Callner, D., "Predicting Sector Capacity under Severe Weather Impact for Traffic Flow Manage- ment," 7th AIAA Aviation Technology, Integration, and Operations Conference, AIAA Paper 2007-7887, Belfast, Northern Ireland, Sept. 2007.
+
+
+
+	
+		Methodologies of Estimating the Impact of Severe Weather on Airspace Capacity
+		
+			LixiaSong
+		
+		
+			CraigWanke
+		
+		
+			StephenZobell
+		
+		
+			DanielGreenbaum
+		
+		
+			ClaudeJackson
+		
+		10.2514/6.2008-8917
+		AIAA Paper 2008-8917
+	
+	
+		The 26th Congress of ICAS and 8th AIAA ATIO
+		Anchorage, AK
+		
+			American Institute of Aeronautics and Astronautics
+			Sept. 2008
+		
+	
+	26th Congress of International Council of the Aeronautical Sciences
+	Song, L., Wanke, C., Greenbaum, D., Zobell, S., and Jackson, C., "Methodologies of Estimating the Impact of Severe Weather on Airspace Capacity," 26th Congress of International Council of the Aeronautical Sciences, AIAA Paper 2008-8917, Anchorage, AK, Sept. 2008.
+
+
+
+	
+		Description of the Corridor Integrated Weather System (CIWS) Weather Products
+		
+			JEvans
+		
+		
+			DKlingle-Wilson
+		
+	
+	
+		MIT Lincoln Lab., Rept. ATC-317
+		
+			2005
+			Cambridge, MA
+		
+	
+	Evans, J., and Klingle-Wilson, D., "Description of the Corridor Integrated Weather System (CIWS) Weather Products," MIT Lincoln Lab., Rept. ATC-317, Cambridge, MA, 2005.
+
+
+
+	
+		Model Estimates of Traffic Reduction in Storm Impacted En Route Airspace
+		
+			BrianMartin
+		
+		10.2514/6.2007-7889
+	
+	
+		7th AIAA ATIO Conf, 2nd CEIAT Int'l Conf on Innov and Integr in Aero Sciences,17th LTA Systems Tech Conf; followed by 2nd TEOS Forum
+		Belfast, Northern Ireland
+		
+			American Institute of Aeronautics and Astronautics
+			Sept. 2007
+		
+	
+	Martin, B., "Model Estimates of Traffic Reduction in Storm Impacted En Route Airspace," 7th AIAA Aviation Technology, Integration, and Operations Conference, AIAA Paper 2007-7889, Belfast, Northern Ireland, Sept. 2007.
+
+
+				
+			
+		
+	

file134.txt

@@ -0,0 +1,630 @@
+
+	
+	
+		
+I. IntroductionC ontrails are clouds that are visible trails of water vapor made by the exhaust of aircraft engines.They appear and persist if the aircraft is flying in certain atmospheric conditions.The environmental impact of aircraft-induced persistent contrails has drawn attention in recent years. 1 Persistent contrails reduce incoming solar radiation and outgoing thermal radiation in a way that accumulates heat. 2 The global mean contrail cover observed in 1992 is estimated to double by 2015 and to quadruple by 2050 due to the increase in air traffic. 3Studies suggest that the environmental impact from persistent contrails may be three to four times, 4 or even ten times, 5 larger than that from aviation emissions.Therefore, methods to reduce aircraft-induced persistent contrails need to be explored to minimize the impact on the global environment.Efforts have been made in the past years to identify and reduce persistent contrail production.Gierens 6 and Noppel 7 reviewed various strategies for contrail avoidance including changing engine architecture, enhancing airframe and engine integration, using alternate fuels, and modifying traffic flow management procedures.Among the traffic flow management solutions, Mannstein 8 proposed a strategy to reduce the climate impact of contrails significantly by small changes to each aircraft's flight altitude.Campbell 9 presented a mixed integer programming methodology to optimally reroute aircraft trajectories to avoid the formation of persistent contrails.Both methodologies require a flexible flight plan and onboard contrail detection system.Fichter 10 showed that the global annual mean contrail coverage could be reduced by reducing the cruise altitude.Williams 11,12 proposed strategies for contrail reduction by identifying fixed and varying maximum altitude restrictions.These restrictions generally require more fuel burn and add congestion to the already crowded airspace at lower altitudes.The objective of this paper is to derive a class of indices that can identify and predict, up to six hours in advance, regions of airspace with high potential for contrail formation.Traffic and weather forecasts were used to generate the predicted contrail frequency index.The indices are used to identify air traffic control centers and altitudes with high contrail formation activities over the next one to six hours.The method uses actual air traffic data and provides a one-hour temporal resolution of predicted contrail frequency.The results show that the predicted indices are highly correlated with the actual contrail frequencies and have a high success rate in identifying the centers and flight levels with high contrail frequencies over the next one to three hours.The remainder of the paper is organized as follows.Section II provides the descriptions of atmospheric and aircraft data and the contrail model used in this paper.Section III describes contrail frequency index, predicted contrail frequency index, and their use for contrail reduction strategies.The results are demonstrated in Section IV.Finally, a summary and conclusions are presented in Section V.
+II. Atmospheric and Aircraft Data and Contrail ModelThe atmospheric data, contrail model, and aircraft data used to generate the contrail formation frequency are described in this section.
+A. Atmospheric DataContrails can be observed from surface observation data 13 and detected by satellite data. 14Duda 15 has related the observations to numerical weather analysis output and demonstrated that persistent contrail formation can be computed using atmospheric temperature and humidity data retrieved from the Rapid Updated Cycle (RUC) data, provided by the National Oceanic and Atmospheric Administration (NOAA).Contrails can persist when ambient air is supersaturated with respect to ice, which means the environmental relative humidity with respect to ice (RHi) is greater than one hundred percent. 16The RHi can be computed from relative humidity with respect to water (RHw) and temperature, which are available in the RUC data.The one-hour, two-hour, three-hour, and six-hour forecasts and the forty-kilometer (40km) resolution RUC data were used in this paper.The data have a temporal resolution of one hour, a horizontal resolution of 40 kilometers, and isobaric pressure levels from 100 to 1000 hectopascals (hPa) in 25 hPa increments.The vertical range is from 150 hPa to 400 hPa, which are equivalent to 23, 600 feet to 44, 400 feet in the standard atmosphere.As an example, snap shots of temperature and RHw contours at 8AM eastern daylight time (EDT) on August 1, 2007 at a pressure altitude of 250 hPa, or about 34, 100 feet, are shown in Fig. 1.
+B. Contrail ModelContrails are clouds produced by aircraft operating at high altitudes.Persistent contrail formation areas are defined as areas with RHi greater than or equal to 100%.RHi can be computed from RHw and temperature using the saturation vapor pressure coefficients of Alduchov, 17 formulated as RHi = RHw × 6.0612e 18.102T /(249.52+T ) 6.1162e 22.577T /(237.78+T ) ,where T is the temperature in Celsius.The temperature and relative humidity shown in Fig. 1 can be translated to RHi.The 40-km RUC data consist of a grid of (113 × 151) data points at each isobaric pressure level.The altitude level index, l, is defined as l = 1 . . .11 corresponding to isobaric pressure levels at 400, 375, . . ., 150 hPa.The level index, isobaric pressure level, and approximate corresponding flight levels are listed in Table 1.The potential persistent contrail formation matrix (contrail matrix) at hour h at level l is defined asR l h =     r 1,1 r 1,2 . . .    ,(2)where r i,j is 1 if RHi ≥ 100% at grid (i, j), and 0 if RHi < 100%.To indicate the location of the twenty U.S. air traffic control centers in the grid, the center grid matrix is defined asC k =     c 1,    ,(3)where k is the center index corresponding to the twenty continental U.S. air space control centers (see Table 2), and c i,j is one if the grid point is within the center and zero if not.The potential persistent contrail formation coverage ratio (contrail coverage ratio) of one center can be defined by the total area of the contrail regions in the center divided by the area of the center.Assuming all of the grid points have the same area, the contrail coverage ratio of center k at time t at altitude level l can be defined as113 i=1 151 j=1 r i,j c i,j 113 i=1 151 j=1 c i,j ,(4)where r i,j is an element of R l t , and c i,j is an element of C k .As an example, the contrail area at flight level 341 at 8AM EDT on August 1, 2007 is shown in Fig. 2a.The corresponding center contrail coverage ratio, computed from Eq. ( 4), is shown in Fig. 2b.The center contrail coverage ratio indicates the portion of center airspace that would form contrails when aircraft fly through it.When the ratio is zero, there will be no contrail formed in the center.
+C. Aircraft DataContrails form when aircraft fly through a potential contrail formation area.Aircraft locations are needed to determine the contrail formation frequency.The aircraft data used in this paper were extracted from the Federal Aviation Administration's Aircraft Situation Display to Industry (ASDI) data.The ASDI has a sampling rate of one minute.The same grid used in the RUC data was used to generate the aircraft position matrix.The aircraft position matrix is defined asA l t =     a 1,    ,(5)where a i,j is the number of aircraft within grid (i, j) flying closest to altitude level l at time t.The aircraft position matrix indicates the air traffic density in the grid scale at different altitudes.
+III. MethodologyThe concept of contrail frequency index, predicted contrail frequency index, and its use for contrail reduction strategies are described in this section.
+A. Contrail Frequency IndexAs discussed in the previous section, the size and coverage ratio of the persistent contrail formation areas are not sufficient indications of severity of contrail activities.The center contrail frequency index consists of both potential contrail formation area and air traffic information.It is defined as the number of aircraft flying through an area that would form persistent contrails at time t at level l.It is formulated as113 i=1 151 j=1 r i,j c i,j a i,j ,(6)where r i,j , c i,j , and a i,j are defined in Eq. ( 2), (3), and (5).As an example, the center contrail frequency indices at flight level 341 were computed at 8AM EDT on August 1, 2007 and are shown in Fig. 3.Even though in Fig. 2b the contrail coverage ratio of Houston Center is higher than Atlanta Center, in Fig. 3 the contrail frequency of Houston Center is zero.Figure 4a
+B. Predicted Contrail Frequency IndexThe contrail frequency index derived in the previous section indicates the actual contrail activities.For strategic planning, prediction of the contrail frequency for the next few hours is needed.The predicted contrail frequency index is defined as a convolution of traffic data and atmospheric conditions.They are similar to the concept of Weather Impacted Traffic Index (WITI) introduced by Callaham et al. 18 and Sridhar, 19 and the three dimensional index derived by Chen. 20The index consists of the RUC forecast data and the predicted aircraft locations.The center predicted contrail frequency index is defined as the predicted number of aircraft flying through the forecasted potential contrail area at time t at level l in center k.It is formulated as113 i=1 151 j=1 r i,j c i,j a i,j ,(7)where r i,j is defined in Eq. ( 2) with RUC forecast data, c i,j is defined in (3), and a i,j is defined in (5). a i,j is based on the historic air traffic data during the planning period.As in the case of WITI, the index is affected more by the changing atmospheric conditions than by small daily variations to the nominal traffic plan. 19In Eq. ( 7) the coefficient a i,j can be thought of as an air traffic weighting coefficient.
+C. Contrail Reduction StrategyThe feasibility of using predicted contrail frequency index for contrail reduction is investigated.The center predicted contrail frequency index can be used to identify the flight level that would have formed the most contrails and find an alternate altitude with less contrail activities.The contrail frequency index after the contrail reduction strategy has been applied is formulated as113 i=1 151 j=1 r i,j c i,j âi,j ,(8)where r i,j and c i,j are defined in Eqs. ( 2) and ( 3), and âi,j is defined in Eq, ( 5) with the aircraft location after the contrail reduction strategy is applied.The contrail reduction strategies need to consider extra fuel burn to minimize overall environmental impact and not to add congestions in the center.The strategy in Ref. 21 uses the predicted contrail frequency index to identify the area that would have formed the most contrail activities, and change the cruise altitudes of a group of aircraft to reduce contrails.The changes need to have minimal extra fuel utilization and maintain the air traffic density below airspace capacity.In general, changing the cruise altitude of a group of aircraft will not increase the air traffic density within the center and sectors.
+IV. ResultsThe temperature and relative humidity from RUC data and aircraft position from ASDI data in 2007 were processed and analyzed.Figure 5 shows four average hourly indices at each of the twenty continental U.S. centers at different altitudes in 2007.They are the contrail coverage ratio derived in Section II.B, the aircraft position matrix derived in Section II.C, the contrail frequency index derived in Section III.A, and the contrail frequency density derived later in this section.As shown in the figure, most of the contrails were formed between flight level 301 and 387.These flight levels account for 78% of contrail frequency over all centers and altitudes.Seattle Center (ZSE), Oakland Center (ZOA), Los Angeles Center (ZLA), and New York Center (ZNY) have lower contrail activities.The reasons are there are less flight activities at ZSE, ZOA, and ZLA, and the size of ZNY is small bringing the index low.To observe the density of the center contrail frequency index, the center contrail frequency density is defined by the center contrail frequency index divided by the number of grid points in the center, formulated as113 i=1 151 j=1 r i,j c i,j a i,j 113 i=1 151 j=1 c i,j ,(9)where r i,j , c i,j , and a i,j are defined in Eq. ( 2), (3), and (5). Figure 5d shows the average hourly center contrail frequency density at each of the eleven flight levels and twenty centers.It is shown that there are high contrail activities between flight level 320 and 363, having some centers with density higher than 0.5.The highest contrail density is 0.69 at flight level 363 at Indianapolis Center (ZID).Figure 6 shows the contrail frequency density at flight level 363 on a U.S. map.As shown in the figures, Indianapolis Center (ZID) has the highest contrail density, and its surrounding five centers, Kansas City Center (ZKC), Chicago Center (ZAU), Memphis Center (ZME), Cleveland Center (ZOB), and Atlanta Center (ZTL), also have high contrail density ranged from 0.45 to 0.57.The seasonal variation can be observed by the monthly average center contrail frequency at ZID in 2007, as shown in Fig. 7.In general, there are less contrail activities in summer.The contrail frequency index provides a way to quantify the contrail activities.Next, for the predicted contrail frequency index, one-hour, two-hour, three-hour, and six-hour predicted indices in August 2007 were generated using Eq. ( 7) and analyzed.a i,j in Eq. ( 7) was based on the air traffic data on the same day of week of July 15-21.For example, to generate the predicted contrail frequency indices on August 1, 8, 15, 22, and 29, the air traffic data on July 18 was used since they are all Wednesdays.As an example, actual and one-hour predicted contrail frequency indices at flight level 363 at Indianapolis center are shown in Fig. 8a.As shown in the figure, the one-hour predicted contrail index is highly correlated with actual index, with a correlation coefficient of 0.94.It is mentioned in Ref. 19 that the index is affected more by the changing atmospheric conditions than by small daily variations to the nominal traffic plan.To show the effect on different choices of the historical air traffic data used, the predicted contrail frequency index was regenerated using a i,j computed by the average air traffic in July 2007.The indices at flight level 363 at Indianapolis Center using the average traffic in July 2007 are shown in Fig. 8b with a correlation coefficient of 0.94.The result is very similar to using the air traffic data on July 15-21.Table 3 shows the average correlation coefficients between actual and predicted indices in twenty centers at different flight levels using historical data of July 15-21 and the average of July 2007.Note that there is no significant difference between the two types of historical data.The accuracy of the prediction decays with longer prediction time.The correlations at flight level 414 and 444 are small mainly because of the lower frequency of contrail formation and the resulting higher sensitivity to the noise.The mean correlation coefficients between actual index and one-hour, two-hour, three-hour, and six-hour predicted index are 0.85, 0.72, 0.64, and 0.52 using July 15-21 data, and 0.84, 0.72, 0.63, and 0.51 respectively.To use the prediction contrail frequency index for contrail reduction strategies, center prediction indices were generated and analyzed.Figure 9  In general, when the actual contrail frequency is high, the predicted contrail frequency is high for prediction up to three hours.The six-hour predicted index is under-predicted, most likely due to the prediction inaccuracy.For implementing a contrail reduction strategy, the centers with high contrail frequency indices need to be identified.As an example, the contrail reduction strategy may be enabled when the centers have indices higher than 100.This would affect seven centers including Los Angeles Center (ZLA), Salt Lake City Center (ZLC), Albuquerque Center (ZAB), Dallas/Fort Worth Center (ZFW), Houston Center (ZHU), Jacksonville Center (ZJX), and Miami Center (ZMA).All of the one-hour, two-hour, and three-hour prediction indices are able to correctly identify the centers that need a reduction strategy with the exception that the threehour predicted index fails to identify ZHU.There is one case that the one-hour prediction falsely identifying Denver Center (ZDV) with low contrail activity as having an index greater than 100.The success rate of the identification is defined as the rate of the predicted contrail index correctly identifying the center with high or low contrail activities.In this example, the one-hour, two-hour, three-hour, and six-hour predicted indices have success rates of 95%, 100% , 95% and 65% for identifying the correct centers respectively.The performance of predicted indices for identifying centers with high contrail frequency index is shown in Table 4.As expected, the success rate decays with longer prediction time due to the prediction inaccuracy.Also noticeable is that the success rate decays with higher threshold.There is a 83.47% success rate to identify centers with contrail frequency index greater than 100 using one-hour predicted index, 69.24% using two-hour index, 58.31% using three-hour index, and down to 38.92% using six-hour index.It is harder to successfully identify centers with index greater than 400.There is a 76.19% success rate using one-hour index, and down to 21.99% using six-hour index.
+V. ConclusionsThis paper described a methodology to predict contrail frequency index for contrail reduction.A class of predicted indices that reflects the severity of airspace contrail formation frequency was derived.The indices consist of weather forecast and actual and historical air traffic data.The results show that the predicted indices are affected more by changing atmospheric conditions than by small daily variations of traffic.For the data tested, the one-hour predicted contrail index is highly correlated with the actual index, resulting in an average correlation coefficient of 0.85 and is lower with longer prediction time.The average correlation coefficient between the actual index and the two-hour, three-hour, and six-hour predicted index are 0.72, 0.64, and 0.52, respectively.In terms of developing strategies for contrail reduction, there is a 83.47% success rate to identify centers with contrail frequency index greater than a threshold, 69.24% using two-hour index, 58.31% using three-hour index, and 38.92% using six-hour index.The method of using predicted contrail frequency index for contrail reduction shows promise but requires detailed future evaluation in a fast-time traffic flow management simulation environment.Relative humidity with respect to water
+Figure 1 .1Figure 1.Contours of temperature and RHw at 34,100 feet at 8AM EDT on August 1, 2007.
+Figure 2 .2Figure 2. Potential persistent contrail formation area and coverage ratio at flight level 341 at 8AM EDT on August 1, 2007.
+Figure 3 .3Figure 3. Center contrail frequencies at flight level 341 at 8AM EDT on August 1, 2007.
+Figure 4 .4Figure 4. Aircraft location and persistent contrail formation areas at 8AM EDT on August 1, 2007.
+Figure 5 .Figure 6 .Figure 7 .567Figure 5. Average hourly center contrail coverage ratio, air traffic density, contrail frequency and contrail frequency density in 2007.
+Use of average air traffic data in July 2007
+Figure 8 .8Figure 8. Actual and predicted center contrail frequency index at flight level 363 at Indianapolis Center in August 2007.
+shows the actual and predicted center contrail frequency indices at flight level 363 at 8PM EDT on August 1, 2007.The blue bars are the actual, and the light blue, green, orange, and red color bar are the one-hour, two-hour, three-hour, and six-hour predicted contrail frequency indices computed by Eq. (7) using traffic data on July 18, 2007.As shown in the figure, the actual contrail frequency index and the one-hour, two-hour, and three-hour predicted contrail frequency indices are correlated.
+Table 1 .1Altitude level index, isobaric pressure level, and pressure altitude.level index1234567891011pressure level (hPa ) 400 375 350 325 300 275 250 225 200 175 150flight level (100 feet) 236 251 267 283 301 320 341 363 387 414 444
+Table 2 .2Center index of twenty continental U.S. air traffic control centers.IndexNameIndexName1Seattle Center (ZSE)11Chicago Center (ZAU)2Oakland Center (ZOA)12Indianapolis Center (ZID)3Los Angeles Center (ZLA)13Memphis Center (ZME)4Salt Lake City Center (ZLC)14Cleveland Center (ZOB)5Denver Center (ZDV)15Washington D. C. Center (ZDC)6Albuquerque Center (ZAB)16Atlanta Center (ZTL)7Minneapolis Center (ZMP)17Jacksonville Center (ZJX)8Kansas City Center (ZKC)18Miami Center (ZMA)9Dallas/Fort Worth Center (ZFW)19Boston Center (ZBW)10Houston Center (ZHU)20New York Center (ZNY)
+Table 3 .3Average correlation coefficient between actual and predicted contrail frequency index over twenty U.S. centers in August 2007.Two types of historical data are used, air traffic on July 15-21, 2007 (ref. 1) and the average air traffic in July 2007 (ref.2).prediction timeflight levelone-hourtwo-hourthree-hoursix-hourref. 1 ref. 2 ref. 1 ref. 2 ref. 1 ref. 2 ref. 1 ref. 24440.640.630.580.570.530.530.460.464140.730.690.660.630.600.570.540.503870.890.900.810.820.730.740.570.583630.920.920.790.790.700.690.550.543410.910.910.770.770.680.680.530.523200.910.900.770.750.680.670.530.533010.880.880.730.740.650.650.530.532830.860.860.710.710.620.610.500.482670.860.840.700.700.620.610.490.482510.860.850.700.700.630.620.500.502360.870.860.710.710.630.610.500.48350Actual300one-hourContrail frequency index100 150 200 250two-hour three-hour six-hour500ZSEZOAZLAZLCZDVZABZMPZKCZFWZHUZAUZIDZMEZOBZDCZTLZJXZMAZBWZNYFigure 9. Actual and predicted center contrail frequency index at flight level 363 at 8PM EDT on August 1, 2007.
+Table 4 .4Success percentage of identifying the center with index greater than the threshold index for twenty U.S. centers in August 2007.thresholdprediction time one-hour two-hour three-hour six-hour10083.47%69.24%58.31%38.92%20081.43%64.37%51.27%30.69%30079.14%60.75%46.36%26.18%40076.19%57.28%41.69%21.99%
+		
+		
+			
+
+				
+
+
+	
+		
+			IWaitz
+		
+		
+			JTownsend
+		
+		
+			JCutcher-Gershenfeld
+		
+		
+			EGreitzer
+		
+		
+			JKerrebrock
+		
+		Report to the United States Congress: Aviation and the Environment, A National Vision, Framework for Goals and Recommended Actions
+		London, UK
+		
+			December 2004
+		
+	
+	Tech. rep
+	Partnership for AiR Transportation Noise and Emissions Reduction
+	Waitz, I., Townsend, J., Cutcher-Gershenfeld, J., Greitzer, E., and Kerrebrock, J., "Report to the United States Congress: Aviation and the Environment, A National Vision, Framework for Goals and Recommended Actions," Tech. rep., Partnership for AiR Transportation Noise and Emissions Reduction, London, UK, December 2004.
+
+
+
+	
+		Radiative forcing by contrails
+		
+			RMeerkötter
+		
+		
+			USchumann
+		
+		
+			DRDoelling
+		
+		
+			PMinnis
+		
+		
+			TNakajima
+		
+		
+			YTsushima
+		
+		10.1007/s00585-999-1080-7
+	
+	
+		Annales Geophysicae
+		Ann. Geophys.
+		1432-0576
+		
+			17
+			8
+			
+			1999
+			Copernicus GmbH
+		
+	
+	Meerkotter, R., Schumann, U., Doelling, D. R., Minnis, P., Nakajima, T., and Tsushima, Y., "Radiative forcing by contrails," Annales Geophysicae, Vol. 17, 1999, pp. 1080-1094.
+
+
+
+	
+		Future Development of Contrail Cover, Optical Depth, and Radiative Forcing: Impacts of Increasing Air Traffic and Climate Change
+		
+			SMarquart
+		
+		
+			MPonater
+		
+		
+			FMager
+		
+		
+			RSausen
+		
+		10.1175/1520-0442(2003)016<2890:fdocco>2.0.co;2
+	
+	
+		Journal of Climate
+		J. Climate
+		0894-8755
+		1520-0442
+		
+			16
+			17
+			
+			September 2003
+			American Meteorological Society
+		
+	
+	Marquart, S., Ponater, M., Mager, F., and Sausen, R., "Future Development of Contrail Cover, Optical Depth, and Radiative Forcing: Impacts of Increasing Air Traffic and Climate Change," Journal of Climate, Vol. 16, September 2003, pp. 2890-2904.
+
+
+
+	
+		A Standing Royal Commission
+		
+			SusanOwens
+		
+		10.1093/acprof:oso/9780198294658.003.0003
+	
+	
+		Knowledge, Policy, and Expertise
+		London, UK
+		
+			Oxford University Press
+			2002
+			
+		
+	
+	"The Environmental Effects of Civil Aircraft in Flight," Tech. rep., Royal Commission on Environmental Pollution, London, UK, 2002.
+
+
+
+	
+		Aircraft induced contrail cirrus over Europe
+		
+			HermannMannstein
+		
+		
+			UlrichSchumann
+		
+		10.1127/0941-2948/2005/0058
+	
+	
+		Meteorologische Zeitschrift
+		metz
+		0941-2948
+		
+			14
+			4
+			
+			2005
+			Schweizerbart
+		
+	
+	Mannstein, H. and Schumann, U., "Aircraft induced contrail cirrus over Europe," Meteorologische Zeitschrift, Vol. 14, No. 4, 2005, pp. 549-554.
+
+
+
+	
+		A Review of Various Strategies for Contrail Avoidance
+		
+			KlausGierens
+		
+		
+			LingLim
+		
+		
+			KostasEleftheratos
+		
+		10.2174/1874282300802010001
+	
+	
+		The Open Atmospheric Science Journal
+		TOASCJ
+		1874-2823
+		
+			2
+			1
+			
+			2008
+			Bentham Science Publishers Ltd.
+		
+	
+	The Open Atmospheric
+	Gierens, K., Limb, L., and Eleftheratos, K., "A Review of Various Strategies for Contrail Avoidance," The Open Atmo- spheric Science Journal, Vol. 2, 2008, pp. 1-7.
+
+
+
+	
+		Overview on Contrail and Cirrus Cloud Avoidance Technology
+		
+			FNoppel
+		
+		
+			RSingh
+		
+		10.2514/1.28655
+	
+	
+		Journal of Aircraft
+		Journal of Aircraft
+		0021-8669
+		1533-3868
+		
+			44
+			5
+			
+			2007
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	Noppel., F. and Singh, R., "Overview on Contrail and Cirrus Cloud Avoidance Technology," Journal of Aircraft, Vol. 44, No. 5, 2007, pp. 1721-1726.
+
+
+
+	
+		A note on how to avoid contrail cirrus
+		
+			HermannMannstein
+		
+		
+			PeterSpichtinger
+		
+		
+			KlausGierens
+		
+		10.1016/j.trd.2005.04.012
+	
+	
+		Transportation Research Part D: Transport and Environment
+		Transportation Research Part D: Transport and Environment
+		1361-9209
+		
+			10
+			5
+			
+			September 2005
+			Elsevier BV
+		
+	
+	Mannstein, H., Spichtinger, P., and Gierens, K., "A note on how to avoid contrail cirrus," Transportation Research. Part D, Transport and environment, Vol. 10, No. 5, September 2005, pp. 421-426.
+
+
+
+	
+		An Optimal Strategy for Persistent Contrail Avoidance
+		
+			ScotCampbell
+		
+		
+			NatashaNeogi
+		
+		
+			MichaelBragg
+		
+		10.2514/6.2008-6515
+		AIAA-2008-6515
+	
+	
+		AIAA Guidance, Navigation and Control Conference and Exhibit
+		Honolulu, HI
+		
+			American Institute of Aeronautics and Astronautics
+			August 2008
+		
+	
+	Campbell1, S. E., Neogi, N. A., and Bragg, M. B., "An Optimal Strategy for Persistent Contrail Avoidance," AIAA Guidance, Navigation and Control Conference, AIAA-2008-6515, AIAA, Honolulu, HI, August 2008.
+
+
+
+	
+		The impact of cruise altitude on contrails and related radiative forcing
+		
+			ChristineFichter
+		
+		
+			SusanneMarquart
+		
+		
+			RobertSausen
+		
+		
+			DavidSLee
+		
+		10.1127/0941-2948/2005/0048
+	
+	
+		Meteorologische Zeitschrift
+		metz
+		0941-2948
+		
+			14
+			4
+			
+			August 2005
+			Schweizerbart
+		
+	
+	Fichter, C., Marquart, S., Sausen, R., and Lee, D. S., "The impact of cruise altitude on contrails and related radiative forcing," Meteorologische Zeitschrift, Vol. 14, No. 4, August 2005, pp. 563-572.
+
+
+
+	
+		Reducing the climate change impacts of aviation by restricting cruise altitudes
+		
+			VictoriaWilliams
+		
+		
+			RobertBNoland
+		
+		
+			RalfToumi
+		
+		10.1016/s1361-9209(02)00013-5
+	
+	
+		Transportation Research Part D: Transport and Environment
+		Transportation Research Part D: Transport and Environment
+		1361-9209
+		
+			7
+			6
+			
+			November 2002
+			Elsevier BV
+		
+	
+	Williams, V., Noland, R. B., and Toumi, R., "Reducing the climate change impacts of aviation by restricting cruise altitudes," Transportation Research. Part D, Transport and environment, Vol. 7, No. 5, November 2002, pp. 451-464.
+
+
+
+	
+		Variability of contrail formation conditions and the implications for policies to reduce the climate impacts of aviation
+		
+			VictoriaWilliams
+		
+		
+			RobertBNoland
+		
+		10.1016/j.trd.2005.04.003
+	
+	
+		Transportation Research Part D: Transport and Environment
+		Transportation Research Part D: Transport and Environment
+		1361-9209
+		
+			10
+			4
+			
+			July 2005
+			Elsevier BV
+		
+	
+	Williams, V. and Noland, R. B., "Variability of contrail formation conditions and the implications for policies to reduce the climate impacts of aviation," Transportation Research. Part D, Transport and environment, Vol. 10, No. 4, July 2005, pp. 269-280.
+
+
+
+	
+		Contrail Frequency over the United States from Surface Observations
+		
+			PatrickMinnis
+		
+		
+			JKirkAyers
+		
+		
+			MicheleLNordeen
+		
+		
+			StevenPWeaver
+		
+		10.1175/1520-0442(2003)016<3447:cfotus>2.0.co;2
+	
+	
+		Journal of Climate
+		J. Climate
+		0894-8755
+		1520-0442
+		
+			16
+			21
+			
+			November 2003
+			American Meteorological Society
+		
+	
+	Minnis, P., Ayers, J. K., Nordeen, M. L., and Weaver, S. P., "Contrail Frequency over the United States from Surface Observations," Journal of Climate, Vol. 16, No. 21, November 2003, pp. 34473462.
+
+
+
+	
+		Contrail coverage derived from 2001 AVHRR data over the continental United States of America and surrounding areas
+		
+			RabindraPalikonda
+		
+		
+			PatrickMinnis
+		
+		
+			DavidPDuda
+		
+		
+			HermannMannstein
+		
+		10.1127/0941-2948/2005/0051
+	
+	
+		Meteorologische Zeitschrift
+		metz
+		0941-2948
+		
+			14
+			4
+			
+			August 2005
+			Schweizerbart
+		
+	
+	Palikonda, R., Minnis, P., Duda, D. P., and Mannstein, H., "Contrail coverage derived from 2001 AVHRR data over the continental United States of America and surrounding areas," Meteorologische Zeitschrift, Vol. 14, No. 4, August 2005, pp. 525-536.
+
+
+
+	
+		Relating observations of contrail persistence to numerical weather analysis output
+		
+			RPalikonda
+		
+		
+			PMinnis
+		
+		
+			DPDuda
+		
+		
+			HMannstein
+		
+	
+	
+		Atmospheric Chemistry and Physics
+		
+			9
+			4
+			
+			February 2009
+		
+	
+	Palikonda, R., Minnis, P., Duda, D. P., and Mannstein, H., "Relating observations of contrail persistence to numerical weather analysis output," Atmospheric Chemistry and Physics, Vol. 9, No. 4, February 2009, pp. 1357-1364.
+
+
+
+	
+		Estimated contrail frequency and coverage over the contiguous United States from numerical weather prediction analyses and flight track data
+		
+			DavidPDuda
+		
+		
+			PatrickMinnis
+		
+		
+			RabindraPalikonda
+		
+		10.1127/0941-2948/2005/0050
+	
+	
+		Meteorologische Zeitschrift
+		metz
+		0941-2948
+		
+			14
+			4
+			
+			June-July 2003
+			Schweizerbart
+			Friedrichshafen at Lake Constance, Germany
+		
+	
+	Duda, D. P., Minnis, P., Costulis, P. K., and Palikonda, R., "CONUS Contrail Frequency Estimated from RUC and Flight Track Data," European Conference on Aviation, Atmosphere, and Climate, Friedrichshafen at Lake Constance, Germany, June- July 2003.
+
+
+
+	
+		Improved Magnus Form Approximation of Saturation Vapor Pressure
+		
+			OlegAAlduchov
+		
+		
+			RobertEEskridge
+		
+		10.1175/1520-0450(1996)035<0601:imfaos>2.0.co;2
+	
+	
+		Journal of Applied Meteorology
+		J. Appl. Meteor.
+		0894-8763
+		1520-0450
+		
+			35
+			4
+			
+			April 1996
+			American Meteorological Society
+		
+	
+	Alduchov, O. A. and Eskridge, R. E., "Improved Magnus Form Approximation of Saturation Vapor Pressure," Journal of Applied Meteorology, Vol. 35, No. 4, April 1996, pp. 601-609.
+
+
+
+	
+		Assessing NAS Performance: Normalizing for the Effects of Weather
+		
+			MBCallaham
+		
+		
+			JSDearmon
+		
+		
+			ACooper
+		
+		
+			JHGoodfriend
+		
+		
+			DMoch-Mooney
+		
+		
+			GSolomos
+		
+	
+	
+		4th USA/Europe Air Traffic Management R&D Symposium
+		Santa Fe, NM
+		
+			December 2001
+		
+	
+	Callaham, M. B., DeArmon, J. S., Cooper, A., Goodfriend, J. H., Moch-Mooney, D., and Solomos, G., "Assessing NAS Performance: Normalizing for the Effects of Weather," 4th USA/Europe Air Traffic Management R&D Symposium, Santa Fe, NM, December 2001.
+
+
+
+	
+		Relationship Between Weather, Traffic and Delay Based on Empirical Methods
+		
+			BanavarSridhar
+		
+		
+			SeanSwei
+		
+		10.2514/6.2006-7760
+	
+	
+		6th AIAA Aviation Technology, Integration and Operations Conference (ATIO)
+		Wichita, KS
+		
+			American Institute of Aeronautics and Astronautics
+			September 2006
+		
+	
+	Sridhar, B. and Swei, S., "Relationship between Weather, Traffic and Delay Based on Empirical Methods," 6th AIAA Aviation Technology, Integration and Operations Conference (ATIO), Wichita, KS, September 2006.
+
+
+
+	
+		Estimation of Air Traffic Delay Using Three Dimensional Weather Information
+		
+			NeilChen
+		
+		
+			BanavarSridhar
+		
+		10.2514/6.2008-8916
+	
+	
+		The 26th Congress of ICAS and 8th AIAA ATIO
+		Anchrorage, AK
+		
+			American Institute of Aeronautics and Astronautics
+			September 2008
+		
+	
+	Chen, N. Y. and Sridhar, B., "Estimation of Air Traffic Delay Using Three Dimensional Weather Information," 8th AIAA Aviation Technology, Integration and Operations Conference (ATIO), Anchrorage, AK, September 2008.
+
+
+
+	
+		Fuel efficient strategies for reducing contrail formations in United States airspace
+		
+			BanavarSridhar
+		
+		
+			NeilYChen
+		
+		10.1109/dasc.2010.5655533
+	
+	
+		29th Digital Avionics Systems Conference
+		Salt Lake City, UT
+		
+			IEEE
+			October 2010
+		
+	
+	to appear
+	Sridhar, B., Chen, N. Y., and Ng, H. K., "Fuel Efficient Strategies for Reducing Contrail Formations in United State National Air Space," 29th Digital Avionics Systems Conference, Salt Lake City, UT, October 2010, to appear.
+
+
+				
+			
+		
+	

file135.txt

@@ -0,0 +1,344 @@
+
+	
+	
+		
+I. IntroductionS tudies show that 70% of all delays are related to weather and 60% are caused by convective weather. 1 To guide flow control decisions and identify the strategies to reduce delays, cancellations, and costs during the day of operations in various weather conditions, it is useful to create a delay estimation model and provide accurate delay estimation based on weather information.Efforts have been made to identify the correlation between weather and delay both at the regional and national levels.The most promising concept is to use the Weather-Impacted Traffic Index (WITI), which was first introduced by Callaham et al. 2 Sridhar 3, 4 and Chatterji 5 expanded the concept and built daily delay estimation models by linear regression.Klein 6 developed objective measures of the combined impact of traffic demand and weather on the air traffic system by further combining en route WITI, terminal WITI and queuing delay to form a new metric, the National Airspace System Weather Index.Hansen 7 developed models involving the use of econometric concepts to understand the relationship between observed airline delay and several causal factors, including traffic, airport weather, en route convective weather, and weather forecast accuracy.All of these models are two dimensional, considering only the storm location, not the echo tops.As a result, previous research in delay estimation does not take into account the ability of some aircraft to fly above echo tops.The objective of this paper is to extend the WITI concept by adding aircraft altitude and the storm echo tops.The methodology of WITI generation in Ref. 3-5 is refined and a three dimensional WITI (3D-WITI) is generated based on data from the Corridor Integrated Weather System (CIWS). 8CIWS, developed and operated by MIT Lincoln Laboratory, provides both accurate precipitation and echo tops data.The relationships between CIWS WITI without echo tops information (2D-WITI), 3D-WITI, and Aviation System Performance Metrics (ASPM) delay were studied.The periodic linear models 9, 10 were used to evaluate the performance of the delay estimation.The remainder of the paper is organized as follows.Section II provides the definition of 2D-WITI, 3D-WITI, and delay.Next, the delay estimation models are described in Section III.The delay estimation models and the methods to compute the model parameters are formulated.The results and performance of the models are demonstrated in Section IV.Finally, Section V provides conclusions.
+II. WITI and Delay
+A. 2D-WITIWITI is an indicator of the number of aircraft affected by weather.At given time k, the computation of WITI consists of finding: 1) the weather contours of interest W i (k), 2) the aircraft location T j (k), and 3) if aircraft T j (k) is located inside contour W i (k).It should be noted that in the second step, the aircraft location is based on the air traffic on days unaffected by weather, as described in Ref. 3-5.Considering only the projected positions of aircraft location and storm location, the 2D-WITI is formulated as follows,W IT I 2D (k) = m(k) j=1 n(k) i=1    1 if T j (k) is inside W i (k) 0 if T j (k) is outside W i (k) ,(1)where n(k) is the number of weather contours of interest at time k, and m(k) is the number of aircraft of interest at time k.][5] Here CIWS is used for the WITI computation.CIWS, created by MIT Lincoln Laboratory, provides 2-hour convective forecasts updated every 5 minutes.Although the current CIWS does not cover the entire NAS, the coverage includes the major east airway and most high volume terminal areas.All or most of the Chicago Center (ZAU), New York Center (ZNY), Atlanta Center (ZTL), Houston Center (ZHU), Washington Center (ZDC), Boston Center (ZBW), Cleveland Center (ZOB), and Memphis Center (ZME) are covered by the current CIWS.The CIWS-WITI generation method has been integrated with the Future ATM Concepts Evaluation Tool (FACET). 11,12 gure 1a shows a FACET display with the CIWS weather and the air traffic.The grey rectangular bounding box indicates the CIWS-covered area.The WITI computation involves finding the number of aircraft within the weather contours at a certain level, as formulated in Eq. (1).As an example, in Fig. 1b, an arbitrary level 3 contour is shown in yellow, with a total of five aircraft within the contour.Therefore, the WITI count of the contour is 5.A day is defined as 24 hours starting at 0400 Eastern Standard Time (EST), since most of the aircraft departing on the previous day would have landed before 0400 (EST) and new aircraft are starting to depart after 0400 (EST).The WITI is generated at the sampling rate of 1 minute, as in previous studies. 3The CIWS data are updated every 5 minutes and are considered constant during the 5-minutes interval.
+B. 3D-WITIIn addition to the precipitation weather products, CIWS provides the echo tops information, which indicates where it is safe to fly over the storms.If an aircraft is planning to fly through the area affected by the storm but over the echo tops, it should be able to fly through the area safely, and thus is not affected by the weather.Based on this concept, the definition of WITI is extended by including the echo tops information and the altitudes of the aircraft.The echo tops products used in this study have vertical resolutions of 5,000 feet, up to 65,000 feet.Similar to the WITI defined in Eq. ( 1), the three dimensional WITI (3D-WITI), which considers not only the position of the aircraft but also its altitude, consists of one more element, E j i , which are the echo tops weather contours.The superscript i, which denotes the level of the echo tops, is defined as the vertical height divided by 5000 feet.For example, E 3 i means the i th echo tops contour is at 15,000 feet.There are 14 echo tops products available, E 0 i . . .E 13 i .The 3D-WITI is defined asW IT I 3D (k) = m(k) j=1   n(k) i=1    1 if T j (k) is inside W i (k) 0 if T j (k) is outside W i (k)   .   p(k) i=1    1 if T j (k) is inside E aj (k) i (k) 0 if T j (k) is outside E aj (k) i (k)   ,(2)where n(k) is the number of precipitation weather contours of interest at time k, p(k) is the number of echo tops weather contours of interest at time k, m(k) is the number of flying aircraft of interest at time k, and a j (k) is the altitude level defined as the aircraft altitude divided by 5000 feet rounding to the next integer.For example, if the altitude of aircraft T j (k) is 36000 feet, a j (k) is the next integer of 36000/5000 = 7.2, which is 8.Following the example in Fig. 1b, the echo tops at altitude level 30,000 feet in the same area are shown in Fig. 2a.There are two aircraft, indicated in yellow color, outside the echo tops contours.The aircraft flying outside the contours at the flight level over 30,000 feet means that they are flying over the storms, thus they are not affected by weather.The one in the north is at flight level 36,000 feet and the one in the south is at 34,000 feet, which means both are above the storms and should not contribute to the W IT I 3D .Therefore, the W IT I 3D count for this area is 3.The three dimensional view of the CIWS echo top products is shown in Fig. 2b.It can be seen that the two yellow aircraft are above the echo tops of the storms.Both 2D-WITI and 3D-WITI are processed using the data from June 4, 2007.Figure 3 shows a comparison between the 2D-WITI and the 3D-WITI.The time series values are shown in Fig. 3a.The hourly WITIs are defined as the sum of WITIs in every hour, and are shown in Fig. 3b.The discrepancy between the two suggests that some air traffic affected by the precipitation weather products could fly over the echo tops and would not contribute to delay in the NAS.Further analysis of the delay estimate models will be presented in the next section.in Fig. 4. As illustrated by this figure, high correlation among the three is clearly shown.Note that the ASPM delay in the figure is scaled down by 1/6 in order to have the same level of magnitude as the WITI counts.As a reference, the monthly average correlation coefficient between 2D-WITI and ASPM delay is 0.86.There is no improvement in the monthly average correlation coefficient between 3D-WITI and ASPM over 2D-WITI.However, looking at days where there is a large discrepancy between 2D-WITI and 3D-WITI, 3D-WITI is indeed better correlated with the delay.For example, on June 4, 2007, the 2D-WITI is high at 0900(EST) while 3D-WITI remains low, as shown in Fig. 5. On June 4, the 2D-WITI and ASPM delay has a correlation coefficient of 0.84, while 3D-WITI and ASPM delay has a correlation coefficient of 0.93.This suggests even though there might be many aircraft routes covered by the bad weather, some of them should have no problem flying over the storms as planned.Thus, these aircraft should not contribute to the NAS delay, and this fact is indicated by a lack of a corresponding peak in the ASPM delay plot.correlation of hourly air traffic delay with respect to hourly 2D-WITI and 3D-WITI within a day.Three classes of models are described in this section: 1) a periodic linear (PL) hourly delay model, 2) a periodic finite impulse response (PFIR) hourly delay model, and 3) a periodic linear autoregressive with exogenous inputs (PARX) hourly delay model.
+A. Periodic Linear (PL) Hourly Delay ModelFirst, given the observed daily WITI values for p days, w = [w 1 w 2 . . .w p ] T and the observed aggregate daily delay, d = [d 1 d 2 . . .d p ] T , the linear model for the daily delay can be formulated asd = α w + γ + e,(3)where α and γ are the model coefficients and e is the error estimate.The α and γ can be found by solving the least-square solution of Eq. ( 3).The delay estimate, d, can be expressed asd = α w + γ,(4)Next, as seen in Fig. 4, both the ASPM delay and WITI have a 24-hour period.Instead of using the aggregate daily delay and WITI, the hourly data can be used to build the delay model.The daily delay model can be divided into 24 individual hourly delay models.Given the observed hourly WITI and delay on p days, the WITI and delay data matrices are defined asW =     w 1,1 w 2,    ,(5)where w i,j and d i,j are the hourly WITI and delay at hour i on day j.Assume w h and d h are the h th columns of W and D, which represent the hourly WITI and delay at hour h of the observed days.Note that h = 1, 2, . . ., 24, and it starts at 0400(EST).The delay model at hour h is described asd h = α h w h + γ h + e h ,(6)where α's and γ's can be found by solving the least-square solution of Eq. ( 6).The estimate of the hourly delay dh can then be expressed asdh = α h w h + γ h .(7)The model in Eq. ( 6) and Eq. ( 7) is referred to as the periodic linear hourly delay model.
+B. Periodic Finite Impulse Response (PFIR) Hourly Delay ModelThe periodic linear delay model considers only the relationship between current delay and current WITI.In reality, the delay might be caused by not only the current weather but also the weather hours earlier.Assuming the current delay is correlated with the current WITI and the WITI in the previous hour, the model can be described asd h = α h,0 w h + α h,1 w h-1 + γ h + e h .(8)The α's and γ's can be found by solving the least-square solution of Eq. ( 8).The estimate of hourly delay dh can be formulated asdh = α h,0 w h + α h,1 w h-1 + γ h .(9)The model in Eq. ( 8) and Eq. ( 9) is referred to as the first-order PFIR model with direct feed-through.First-order means that data one time-step earlier was used and the direct feed-through means that current data are used to build the model.More generally, the n th -order PFIR model can be formulated asd h = n k=0 α h,k w h-k + γ h + e h , (10)dh = n k=0 α h,k w h-k + γ h . (11)Note that w h-k is defined as 0 for h ≤ k, which implies that the least-square solutions of α h,k are 0's for h ≤ k.Also, the PL model described in the previous subsection is a special case of the PFIR model when n = 0.
+C. Periodic Linear Autoregressive with Exogenous Inputs (PARX) Hourly Delay ModelAt a given hour h, in addition to the current and past WITI, the past delay might also be available in certain applications such as real-time delay prediction. 13Assuming the current delay is correlated with the current WITI, the WITI in the past n hours, and the delay in the past m hours, the model can be formulated asd h = n k=0 α h,k w h-k + m l=1 β h,k d h-l + γ h + e h ,(12) dh= n k=0 α h,k w h-k + m l=1 β h,k d h-l + γ h ,(13)where α's, β's, and γ's can be found by solving the least-square solution of Eq. ( 12).This model is referred to as the PARX model with order (n, m).Note that the PFIR model in the previous subsection is a subset of PARX model when m = 0. To be more explicit, Eq. ( 12) can be rewritten asd h = w h . . . w h-n d h-1 . . . d h-m 1               α h,0 . . . α h,n β h,1 . . . β h,m γ h               + e h .(14)The Moore-Penrose pseudo-inverse 14 is used to solve the equation.The solution is described asα h,0 . . . α h,n β h,1 . . . β h,m γ h T = w h . . . w h-n d h-1 . . . d h-m 1 † d h ,(15)where [•] † represents the pseudo-inverse of the matrix.
+IV. ResultsThe delay and 3D-WITI data for the month of June, 2007 were used as reference data to build a PARX model, described in Eq. ( 12), with model order (n, m) = (1, 1).In this model, there are a total of 96 model parameters to be identified, including α 1,0 . . .α 24,0 , α 1,1 . . .α 24,1 , β 1,1 . . .β 24,1 , and γ 1 . . .γ 24 .Once the parameters are identified, Eq. ( 13) is used to compute the estimate of the hourly delay on reference days, dh .Figure 6a shows the actual ASPM hourly delay d h versus the estimate of the hourly delay dh on all the reference days.The red line in the figure indicates the perfect estimates.As shown in the figure, all the dots lie around the red line which suggests d h and dh are close.The average daily root-mean-square (RMS) error between d h and dh , or e h , is 1714 minutes, which yields only 5.83% of the average RMS of daily ASPM delay in June, 2007, which is 29403 minutes.Next, a day not in the reference days was selected to evaluate the performance of the delay estimation model.For July 9, 2007, which has total ASPM delay of 324577 minutes, Fig. 6b shows the actual ASPM delay and the estimated delay.The RMS error between the actual ASPM delay and the delay estimate is 1573 minutes, only 5.54% of the RMS of the actual ASPM delay.Furthermore, the PARX models with different order (n, m) were used to evaluate the performance of air traffic delay estimates using 2D-WITI and 3D-WITI.The pair (n, m) is the order of the model, where n is the number of past WITI and m is the number of past delay used in the model.There are different variations of the models.For example, for m = 0, the delay estimates are related to the WITI and do not depend on past values of delay.These models are essentially PFIR models, and n = 0 represents the simple PL model.On the other hand, for n = 0, the delay estimates are only related to the past delays and do not depend on the WITI.The models are periodic autoregressive (PAR) models.The PAR models are used as the baseline to evaluate how much improvement can be achieved with the WITI information.The whole month of data from July, 2007 are used to validate the models.All PL, PFIR, PAR and PARX models using both 2D-WITI and 3D-WITI with different orders were tested.The results are summarized in Table 1 and2.As shown in Table 1, the PFIR models do not perform well because of the lack of past delay information.In Table 2, it shows that the PARX model with order (1, 1) using 3D-WITI is slightly better than the other models.It was noticed that higher order models do not provide better performance for this class of models.The reason might be higher order models tend to over fit the observed data and lose the generality for the validation data.The PARX model using 3D-WITI with order (1, 1) was selected as the best for this class of models.In this case, the average daily RMS error of the model is 1876 minutes, mean absolute error is 1382 minutes, and maximum error is 4762 minutes.It provides a small improvement (about 2%) in delay estimation over other methods.Figure 7 shows the correlation between the actual delay and the optimal estimated delay for each hour in July, 2007.The correlation coefficient between the two is 0.98.
+V. Conclusions and Future WorkIn this paper, a new three-dimensional weather-impacted traffic index was developed and presented.The new index uses the aircraft altitude and the storm echo tops information to discount an aircraft if it can fly over the weather-impacted area safely, thus incurring no delay.Both 2D-WITI and 3D-WITI were computed using CIWS weather product, which provides both accurate precipitation and echo tops weather information.The delay estimation methodology utilizes the hourly resolution of the ASPM data.The indices were used as exogenous inputs for periodic autoregressive models to perform the NAS delay estimation.Various linear hourly models using different combinations of past and current weather and traffic information were examined to determine the optimal delay estimation model.The models were built using traffic and weather data from June 2007, and were validated with the data from July 2007.The recursive models using WITI information outperform models using only delay information.Another result from the study is that using higher order models may not provide more accurate estimates due to overfitting of the data.No clear conclusions can be drawn about the additional benefits of using 3D-WITI information versus 2D-WITI information.The performance of 3D-WITI models need more examination.The result shows that the 3D-WITI provides a small improvement (about 2%) in delay estimation.The reason that using 3D-WITI does not provides significant superior performance over using 2D-WITI might be that the aircraft do not take full advantage of the echo tops information to fly over storm.Further studies, with larger datasets and a better understanding of how echo top information is used by pilots and air traffic controllers may lead to more accurate delay estimates.Accurate delay estimates will benefit the ATM in identifying the strategies to reduce delays, cancelations, and costs during operations in severe weather conditions.Figure 1 .1Figure 1.FACET is running with CIWS weather and air traffic loaded.
+Figure 2 .2Figure 2. FACET is running with CIWS echo tops weather and air traffic loaded.
+Figure 3 .3Figure 3. 2D-WITI and 3D-WITI on June 4, 2007.
+Figure 4 .Figure 5 .45Figure 4. Hourly 2D-WITI, 3D-WITI, and ASPM delay in June, 2007
+Figure 6 .6Figure 6.ASPM delay and estimated delay
+Figure 7 .7Figure 7. Hourly ASPM delay and estimated delay in July, 2007.
+Table 1 .1Validation for the PFIR models with different parameters using July, 2007 data.The numbers are in minutes.Model2D PL2D PFIR3D PL3D PFIR(n,m)(0,0)(1,0)(2,0)(0,0)(1,0)(2,0)RMS error479148014850481447844907Mean absolute error360235673692362935703761Maximum error10861 11011 10830 10940 11050 10971
+Table 2 .2Validation for the PARX models with different parameters using July, 2007 data.The numbers are in minutes.ModelPAR2D PARX3D PARX(n,m)(0,1) (0,2) (1,1) (2,2) (1,1) (2,2)RMS error1920 2050 1895 2086 1876 2070Mean absolute error 1388 1496 1399 1525 1382 1513Maximum error5074 5273 4800 5433 4762 5433
+		
+		
+			
+
+				
+
+
+	
+		Weather Forecasting Accuracy for FAA Traffic Flow Management
+		10.17226/10637
+	
+	
+		Weather Forecasting Accuracy for FAA Traffic Flow Management
+		Washington, DC
+		
+			National Academies Press
+			2003
+		
+		
+			National Research Council
+		
+	
+	National Research Council, Weather Forecasting Accuracy for FAA Traffic Flow Management, The National Academies Press, Washington, DC, 2003.
+
+
+
+	
+		Assessing NAS Performance: Normalizing for the Effects of Weather
+		
+			MBCallaham
+		
+		
+			JSDearmon
+		
+		
+			ACooper
+		
+		
+			JHGoodfriend
+		
+		
+			DMoch-Mooney
+		
+		
+			GSolomos
+		
+	
+	
+		4th USA/Europe Air Traffic Management R&D Symposium
+		Santa Fe, NM
+		
+			December 2001
+		
+	
+	Callaham, M. B., DeArmon, J. S., Cooper, A., Goodfriend, J. H., Moch-Mooney, D., and Solomos, G., "Assessing NAS Performance: Normalizing for the Effects of Weather," 4th USA/Europe Air Traffic Management R&D Symposium, Santa Fe, NM, December 2001.
+
+
+
+	
+		Relationship Between Weather, Traffic and Delay Based on Empirical Methods
+		
+			BanavarSridhar
+		
+		
+			SeanSwei
+		
+		10.2514/6.2006-7760
+	
+	
+		6th AIAA Aviation Technology, Integration and Operations Conference (ATIO)
+		Wichita, KS
+		
+			American Institute of Aeronautics and Astronautics
+			September 2006
+		
+	
+	Sridhar, B. and Swei, S., "Relationship between Weather, Traffic and Delay Based on Empirical Methods," 6th AIAA Aviation Technology, Integration and Operations Conference (ATIO), Wichita, KS, September 2006.
+
+
+
+	
+		Classification and Computation of Aggregate Delay Using Center-Based Weather Impacted Traffic Index
+		
+			BanavarSridhar
+		
+		
+			SeanSwei
+		
+		10.2514/6.2007-7890
+	
+	
+		7th AIAA ATIO Conf, 2nd CEIAT Int'l Conf on Innov and Integr in Aero Sciences,17th LTA Systems Tech Conf; followed by 2nd TEOS Forum
+		Belfast, Northern Ireland
+		
+			American Institute of Aeronautics and Astronautics
+			September 2007
+		
+	
+	Sridhar, B. and Swei, S., "Classification and Computation of Aggregate Delay Using Center-Based Weather Impacted Traffic Index," 7th AIAA Aviation Technology, Integration and Operations Conference (ATIO), Belfast, Northern Ireland, September 2007.
+
+
+
+	
+		National Airspace System Delay Estimation Using Weather Weighted Traffic Counts
+		
+			GanoChatterji
+		
+		
+			BanavarSridhar
+		
+		10.2514/6.2005-6278
+	
+	
+		AIAA Guidance, Navigation, and Control Conference and Exhibit
+		San Francisco, CA
+		
+			American Institute of Aeronautics and Astronautics
+			August 2005
+		
+	
+	Chatterji, G. and Sridhar, B., "National Airspace System Delay Estimation Using Weather Weighted Traffic Counts," AIAA Guidance, Navigation and Control Conference, San Francisco, CA, August 2005.
+
+
+
+	
+		"Airspace Playbook": Dynamic Airspace Reallocation Coordinated with the National Severe Weather Playbook
+		
+			AlexanderKlein
+		
+		
+			ParimalKopardekar
+		
+		
+			MarkRodgers
+		
+		
+			HongKaing
+		
+		10.2514/6.2007-7764
+	
+	
+		7th AIAA ATIO Conf, 2nd CEIAT Int'l Conf on Innov and Integr in Aero Sciences,17th LTA Systems Tech Conf; followed by 2nd TEOS Forum
+		Barcelona, Spain
+		
+			American Institute of Aeronautics and Astronautics
+			July 2007
+		
+	
+	Klein, A., Jehlen, R., and Liang, D., "Weather Index With Queuing Component For National Airspace System Perfor- mance Assessment," 7th USA-Europe ATM R&D Seminar , Barcelona, Spain, July 2007.
+
+
+
+	
+		Forecast and Real-time Status of Airspace Closures in the National Airspace System (NAS)
+		
+			MHansen
+		
+		
+			JXiong
+		
+		10.2514/6.2021-2362.vid
+	
+	
+		7th USA-Europe ATM R&D Seminar
+		Barcelona, Spain
+		
+			American Institute of Aeronautics and Astronautics (AIAA)
+			July 2007
+		
+	
+	Hansen, M. and Xiong, J., "Weather Normalization for Evaluating National Airspace System (NAS) Performance," 7th USA-Europe ATM R&D Seminar , Barcelona, Spain, July 2007.
+
+
+
+	
+		Description of the Corridor Integrated Weather System (CIWS) Weather Products
+		
+			JEvans
+		
+		
+			DKlingle-Wilson
+		
+		ATC-317
+		
+			August 2005
+		
+		
+			MIT Lincoln Laboratory
+		
+	
+	Project Report
+	Evans, J. and Klingle-Wilson, D., "Description of the Corridor Integrated Weather System (CIWS) Weather Products," Project Report ATC-317, MIT Lincoln Laboratory, August 2005.
+
+
+
+	
+		System Identification: Theory for the User
+		
+			LLjung
+		
+		
+			1999
+			Prentice Hall
+			Englewood Cliffs, NJ, 2nd ed.
+		
+	
+	Ljung, L., System Identification: Theory for the User , Prentice Hall, Englewood Cliffs, NJ, 2nd ed., 1999.
+
+
+
+	
+		Multivariate periodic time series models
+		
+			PhilipHansFranses
+		
+		
+			RichardPaap
+		
+		10.1093/019924202x.003.0005
+	
+	
+		Periodic Time Series Models
+		London, UK
+		
+			Oxford University PressOxford
+			2003
+			
+		
+	
+	Franses, P. and Papp, R., Periodic Time Series Models, Oxford Univ. Press, London, UK, 2003.
+
+
+
+	
+		FACET: Future ATM Concepts Evaluation Tool
+		
+			KarlDBilimoria
+		
+		
+			BanavarSridhar
+		
+		
+			ShonRGrabbe
+		
+		
+			GanoBChatterji
+		
+		
+			KapilSSheth
+		
+		10.2514/atcq.9.1.1
+	
+	
+		Air Traffic Control Quarterly
+		Air Traffic Control Quarterly
+		1064-3818
+		2472-5757
+		
+			9
+			1
+			
+			2001
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	Bilimoria, K., Sridhar, B., Chatterji, G. B., Sheth, K., and Grabbe, S., "FACET: Future ATM Concepts Evaluation Tool," Air Traffic Control Quarterly, Vol. 9, No. 1, 2001, pp. 1-20.
+
+
+
+	
+		Migration of Facet from Simulation Environment to Dispatcher Decision Support System
+		
+			BSridhar
+		
+		
+			KSheth
+		
+		
+			PSmith
+		
+		
+			WLeber
+		
+		10.1109/dasc.2005.1563359
+	
+	
+		24th Digital Avionics Systems Conference
+		
+			IEEE
+			November 2005
+			1
+			
+		
+	
+	Sridhar, B., Sheth, K., Smith, P., and Leber, W., "Migration of FACET from Simulation Environment to Dispatcher Decision Support System," 24th Digital Avionics Systems Conference, Vol. 1, November 2005, pp. 3.E.4-31-12.
+
+
+
+	
+		Short Term National Airspace System Delay Prediction Using Weather Impacted Traffic Index
+		
+			BanavarSridhar
+		
+		
+			NeilChen
+		
+		10.2514/6.2008-7395
+	
+	
+		AIAA Guidance, Navigation and Control Conference and Exhibit
+		Honolulu, HI; Baltimore, MD
+		
+			American Institute of Aeronautics and Astronautics
+			August 2008. 1996
+		
+	
+	Matrix Computations. rd ed.
+	Sridhar, B. and Chen, N., "Short Term National Airspace System Delay Prediction Using Weather Impacted Traffic Index," AIAA Guidance, Navigation and Control Conference, Honolulu, HI, August 2008, to appear. 14 Golub, G. H. and Van Loan, C. F., Matrix Computations, The Johns Hopkins University Press, Baltimore, MD, 3rd ed., 1996.
+
+
+				
+			
+		
+	

file136.txt

@@ -0,0 +1,404 @@
+
+	
+	
+		
+I. IntroductionD emand for air transportation has grown rapidly in recent years and is expected to grow in the future.In order to ensure smooth air traffic flow and safety in the presence of disruptions caused by convective weather, innovative modeling and design methods are needed in traffic flow management (TFM).One of the main functions of TFM is to predict and resolve demand-capacity imbalances at the sector level to avoid congestion.Thus an accurate sector prediction model that can account for traffic flow uncertainty and weather impact is an essential component of TFM.Efforts have been made in the past few years to perform sector demand predictions.Traditionally, models used in air traffic control and flow management are based on simulating the trajectories of individual aircraft.Deterministic forecasting of sector demand is routinely done within the Enhanced Traffic Management System (ETMS), which relies on the computation of each aircraft's entry and exit times at each sector along the path of flight.Gilbo 1 proposed a regression model for improving aggregate traffic demand prediction in ETMS, acknowledging the uncertainty in the predictions.A more recent TFM simulation tool, the Future ATM Concepts Evaluation Tool (FACET), 2 was used to propagate the trajectories of the proposed flights forward in time and use them to count the number of aircraft in each sector for demand forecasting and establish confidence bounds on the forecasts. 3][6] The objective of this paper is to develop an empirical sector prediction model that accounts for traffic flow uncertainty and weather impact on the prediction for both short-term (less than 30 minutes) and midterm (30 minutes to 2 hours) predictions.Unlike the traditional methods that use trajectory prediction to perform the sector demand prediction, the periodic autoregressive (PAR) model and its variants 7,8 were used to build the prediction model.The class of PAR models consider both the historical traffic flows to capture the mid-term trend, and the flows in the near past to capture the transient response.In addition, a weather component was embedded in the model to reflect weather impact on sector demand.The remainder of the paper is organized as follows.Section II provides the sector demand data used in the research and the description of the weather-free sector demand prediction model.Next, a weather factor is introduced and the prediction model that considers weather is described in Section III.The results and performance of the models are demonstrated in Section IV.Finally, a summary and conclusions are presented in Section V.
+II. Data and Model
+II.A. Sector Demand DataThe air traffic demand data used in this paper are provided by the recorded Aircraft Situation Display to Industry (ASDI) data generated by the Federal Aviation Administration's Enhanced Traffic Management System (ETMS).The ASDI data provide the locations of all aircraft at one-minute intervals.The sector demand, defined as the number of aircraft in each sector at a given time, can be computed using the ASDI data.In this research, the recorded ASDI data were processed using FACET to obtain the sector demand.Since traffic flow management decisions are made by comparing the peak number of aircraft in a sector during a fifteen-minute interval with the sector's Monitor Alert Parameter (MAP) value, the 15-minute peak sector demand, defined as the maximum sector demand every 15 minutes, was used to build the models.Figure 1 shows the sector demand at every minute and the 15-minute peak sector demand at sector ZID93 on September 3, 2007.Note that in the paper, a day is defined as a 24-hour interval starting at 4:00 AM local time since most of the aircraft departing on the previous day would have landed before 4:00 AM.The black line in Fig. 1 represents the sector demand in an one-minute intervals and the blue dots in Fig. 1 represent the 15-minute peak sector demand during a day, denoted as d k , where k = 1 . . .96.The mid-term trend of sector demand on different days can be observed in Fig. 2, which shows the variation of 15-minute peak sector demand in September 2007.In this figure, each horizontal strip represents one day of 15-minute peak sector demand, and each vertical strip represents the peak sector demand at the same time of day during the entire month.As shown in the figure, the horizontal strips on 9/1, 9/8, 9/15, 9/22 and 9/29, which are Saturdays, have lower demands than the others.The blue vertical regions on the left and right show the off-peak traffic in the early morning and the late night.A vertical light blue region at around 12 o'clock divides the sector demand into morning rush left of the region and the afternoon peak right of it.The sector demand prediction model presented in the next section captures these variations in the demand.
+II.B. Demand Prediction ModelAuto-regressive models have been used for short-term hourly air traffic delay prediction. 9,10 his research extends the delay prediction approach to sector demand prediction.For a given day, a 24-hour period, starting at 4:00 AM local time, is divided into 96 fifteen-minute intervals.Given the observed 15-minute peak sector demands for n days, the sector demand data matrix is defined asD =     d 1,1 d 2,1 . . . d 96,1 . . . . . . . . . . . . d 1,n d 2,n . . . d 96,n     ,(1)where d i,j represents the 15-minute peak sector demand at the i th time step on day j.For September 2007, D has a dimension of 30 by 96, and Fig. 2 shows the image of the matrix D. Assuming d k as the k th column of D, the p-step-ahead sector demand model at time step k in the form of a linear regression model is described asd k+p = α k,p d k + β k,p + e k ,(2)where α k,p and β k,p are coefficients that map the sector demand at the k th time step to the (k + p) th time step, and e k is the error of the model.The least-square solution of α k,p and β k,p that minimizes e T k e k in Eq. ( 2) can be written explicitly asαk,p = n i=1 (d k,i -dk )(d k+p,i -dk+p ) n i=1 (d k,i -dk ) 2 , (3) βk,p = dk+p -α k,p dk ,(4)In the model, αk,p and βk,p , identified from the historical data, capture the periodic features during a day, and the observed sector demand d k,m provides the transient information.The model in Eq. ( 2) and Eq. ( 5) is referred to as the periodic auto-regressive (PAR) sector demand prediction model.As an example, peak sector demand data in August 2007 were used to construct the data matrix in Eq. ( 1), and Eq. ( 2) was used to identify the model parameters αk,p and βk,p , where k = 1 . . .96 and p = 1 . . .8 for 1-step-to 8-step-ahead predictions.The 15-minute-ahead peak sector demand on September 3, 2007 was predicted using Eq. ( 5) with p = 1.The result is shown in Fig. 3a.The root-mean-squared (RMS) error between the actual peak sector demand and the 15-minute demand prediction is 1.93.For the 2-hour prediction, the prediction model is solved for p = 8, and the estimates in Eq. ( 5) are generated.The result is shown in Fig. 3b.The RMS error is 2.15.It is noticed that the PAR model yields larger error as the prediction interval increases.This suggests that using a single observation d k,m in Eq. ( 5) contains less information about dk+p,m when p is large.An alternate method to perform the demand prediction is to use the cumulative sum of the past sector demands as an observation, since the sum includes more information than a single observation and has less noise compared with the single peak sector demand.Following the definition of the sector demand matrix D in  Eq. ( 1), where d k is the k th column of D, the cumulative p-step-ahead sector demand model at time step k can be described in terms of the cumulative sum of q past sector demands asd k+p = α k,p k i=k-q+1 d i + β k,p + e k ,(6)where α k,p and β k,p are the coefficients that map the cumulative sector demand at the k th time step to the sector demand at the (k + p) th time step.Once the least-square solution of coefficients αk,p and βk,p are identified, the p-step prediction of the sector demand at the k th time step for a day m, dk+p,m , based on the observed cumulative sector demand,The model in Eq. ( 6) and Eq. ( 7) is referred to as the cumulative periodic auto-regressive (CPAR) sector demand prediction model.During the analysis, it is noticed that the CPAR model using the sum of the all demands in the past (q = k) works best overall.For the example used in the PAR model, the CPAR model with q = 8 has a RMS error of 1.68 for the 15-minute prediction, and 2.00 for the 2-hour prediction, compared with 1.93 and 2.15 respectively for the PAR model.It appears that the CPAR model performs a little better than the PAR model.More analysis is done in Section IV to evaluate this property.
+III. Weather FactorWeather has a big influence on air traffic sector demand and the uncertainty in weather may cause error in the predictions. 5,11 f a severe storm blocks a sector or regions near it, both the sector capacity and demand may drop dramatically. 12,13  weather factor that discounts the weather-free sector demand prediction is derived in this section.In order to model the weather impact on sector demand prediction, an accurate weather forecast product with high update rate is required.In addition, to capture the impact on all low, high, and super high sectors, the storm echo tops information is useful.The weather data used in this paper was provided by the Corridor Integrated Weather System (CIWS). 14CIWS, developed and operated by MIT Lincoln Laboratory, provides both accurate precipitation and echo tops data and is updated every 5 minutes.In addition, CIWS provides convective forecasts at 5-minute intervals up to 2 hours in the future.The weather factor used to discount the sector demand prediction was chosen to be the sector weather index, defined as the percentage of area covered by the storm with precipitation vertically integrated liquidw k = A w k A ,(8)where A is the area of the sector and A w k is the area of the sector covered by storms with the echo tops at or above the lower bound of the sector at time k.Note that if time k is a future time, the weather forecast is used to determined A w k .It is possible to use other definitions of a sector weather index. 12, 13Figure 4a shows a snap shot of the CIWS data for the high altitude sectors at Indianapolis center (ZID) on a severe weather day.The red spots indicate the storms with VIL level 3 and above, and the echo tops at 35,000 ft.As shown in this figure, most of the sector ZID93 is covered by the storm.The sector weather index for ZID93 on August 16, 2007 is shown in the red line in Fig. 4b.Also shown in the figure is the actual sector demand on the same day in blue line.Notice the sector weather index is greater than 30% from 18:00 to 20:00 Eastern Daylight Time (EDT), and clearly the sector demand drops during the same period.Traffic reduction due to weather impact can be modeled in many different ways. 15In this research, the weather-free prediction was first estimated, then the sector weather index was used to adjust the prediction.Assume that the sector demand starts to decay when the sector weather factor exceeds w low , and reaches 0 when the weather factor reaches w high .The sector demand reduction rate is modeled as the power law distribution, 1 -((w k -w low )/(w high -w low )) γ , where γ is the power of the distribution.To reflect the thresholds, the sector weather index in Eq. ( 8) is redefined asw k =      w low if A w k /A ≤ w low A w k /A if w low < A w k /A < w high w high if w high ≤ A w k /A .(9)In order to adjust the weather impact on the sector demand prediction model, the weather forecast is used to compute the predicted sector weather index.Assume at time k, the predicted sector weather index at time k + p is w k+p , the PAR sector demand prediction model in Eq. ( 5) can be rewritten asdk+p,m = 1 - w k+p -w low w high -w low γ (α k,p d k,m + βk,p ), (10)or the CPAR sector demand prediction model in Eq. ( 7) can be rewritten asdk+p,m = 1 - w k+p -w low w high -w low γ (α k,p k i=k-q+1 d i,m + βk,p ). (11)Using the echo tops information provides a more representative weather index, especially for the high sectors.If there are storms with low echo tops located at some high sectors, the weather might have minimal impact on the sector demand.The sector demand and weather index for sector ZID93 on two different days is shown in Fig. 5.Both days have severe storms, but one has high echo tops while the other has low echo tops.The sector demands on severe weather days were compared with the average sector demand on the rest of the days in the same month.In Fig. 5a, the sector demand on August 16, 2007 is lower than the average between 18:00 and 20:00 (EDT) because of the high weather index during the period, as indicated in Fig. 5c.The blue line in Fig. 5c shows the weather index considering the area covered by storms without the echo tops information, and the red line is the weather index considering the echo tops at 35,000ft and above.In this case, the two lines are close.This suggests that there are severe storms in the area and most of the echo tops are higher than the lower bound of sector ZID93.On the other hand, on October 23, 2007, there is no demand reduction compared to the average of the other days in October 2007 during 18:00 and 20:00 (EDT), shown in Fig. 5b, even though there are storms in the sector during the period, as shown in Fig. 5d.The red line in Fig. 5d is substantially lower than the blue line, which means even though there are storms in the sector, most the echo tops are lower than the low boundary of the sector and have minor impact on the sector demand.In the next section, the sector weather index refers the index with the echo tops information.
+IV. ResultsThe sector demands of 25 high and superhigh sectors in ZID were investigated in this research.The major flows of ZID include the arrivals to Philadelphia International Airport (PHL), Ronald Reagan Washington National Airport (DCA), Chicago O'Hare International Airport (ORD), Detroit Metropolitan Wayne County Airport (DTW), and Cleveland-Hopkins International Airport (CLE), the departures from ORD and DTW, the westbound traffic of airway J80 from New York Center (ZNY) and Boston Center (ZBW), and the traffic to New York Terminal Radar Approach Control (N90).The sector demands for the month of August, 2007 were used to build the sector demand prediction PAR and CPAR models, described in Eq. ( 1), Eq. ( 2), and Eq. ( 6) .The time step used in the models is 15 minutes.Once the parameters were identified, Eq. ( 5) and Eq. ( 7) were used to perform the sector demand prediction for the month of September, 2007.Starting  6, were presented.The behavior of both PAR and CPAR models are summarized in Table 1.Even though the performance of the two models are very close, CPAR prediction models perform equal to or better than the PAR models in all the cases with the exceptions of the 15-min prediction at ZID81 (PAR 1.95, CPAR 1.98) and at ZID82 (PAR 1.57, CPAR 1.58).Among the cases, the error in CPAR models is 2.46% smaller than the error in PAR models in average.Also notice the errors of both PAR and CPAR models are not sensitive to the look ahead time.In general, the errors are larger with longer look ahead time, but only slightly.The errors of 120-min prediction is 5.12% larger than the 15-min prediction in average for the PAR models, and 2.87% larger for the CPAR models.Consider all the high and super-high sector in ZID, the results are similar.The errors of the PAR models are between 1.57 and 2.11 in the 15-min prediction, and between 1.64 and 2.24 in the 120-min prediction, while the errors of the CPAR models are between 1.58 and 2.10 in the 15-min prediction, and between 1.61 and 2.15 in the 120-min prediction.The sector demand prediction for bad weather days uses the weather factor described in the previous section to adjust the weather-free prediction, formulated in Eq. ( 9), Eq. ( 10), and Eq. ( 11), with w low = 0, w high = 1, and γ = 1.The days with peak weather factors greater than 30% are considered bad weather days.For the days and sectors tested, there are four cases of severe weather periods: ZID83 on 08/16/07 between 1600-2200 (EDT), ZID93 on 08/16/07 between 1600-2200 (EDT), ZID82 on 08/21/07 between 0600-1400 (EDT), and ZID92 on 08/21/07 between 0800-1400 (EDT), shown in Fig. 7. Since all these cases happened in August 2007, the model is built using data for July 2007.Two types of weather-weighted models are built, one uses the actual weather information and the other uses the forecast weather information.Using the actual weather information to perform sector demand prediction represents the cases with perfect weather forecast.It is used to evaluate the performance of the weather-weighted prediction model and eliminate the error caused by forecast inaccuracy.The average prediction errors of the four severe weather periods in August 2007 are shown in Fig. 8.It is noticed that in all four cases, both the weather-weighted model using actual weather information (red dash line) and the model using forecast weather (green dash-dot line) produce smaller error than the weather-free model (blue solid line).The weather-weighted model using forecast weather performs as well as the model using actual weather when the prediction time is small (less than 30 minutes).However, with longer prediction time (more than 60 minutes), the performance starts to decay and the errors are closer to the weather-free model.As an example, in Fig. 8b, the weather-weighted sector demand prediction model using actual weather information improves the 15-minute prediction over the weather-free model by 36.38%, the 60-minute prediction by 42.92%, and the 120-minute prediction by 40.77%.For the weather-weighted model using forecast weather, the improvement is 34.73% for the 15-minute prediction, reduced to 27.81% for the 60-minute prediction, and down to 7.71% for the 120-minute prediction.This suggests that with longer prediction time, the forecast inaccuracy might effect the performance of the weather-weighted prediction model using forecast weather.
+ZID super high and high sectors
+V. ConclusionA class of auto-regressive models developed for sector delay estimation is used for predicting traffic demand in a sector between 15 minutes and two hours in the future.The PAR and CPAR models capture both the mid-term trend based on the historical data, and the short-term transient response based on the near past observation.For the sectors tested, the errors of CPAR models are 2.46% smaller than the PAR models.The CPAR model provides the demand predictions with an average RMS error between 1.58 and 2.10 in theFigure 1 .Figure 2 .12Figure 1.Sector demand and peak sector demand at sector ZID93 on September 3, 2007
+Figure 3 .3Figure 3. Peak sector demand and predicted peak sector demand on September 3, 2007
+k i=k-q+1 d i,m , can be expressed as dk+p,m = αk,p k i=k-q+1 d i,m + βk,p .
+CIWS at ZID at 19:30 (EDT) on August 16, 2007 The sector demand and weather index of ZID93 on August 16, 2007
+Figure 4 .4Figure 4.The weather data, the sector demand and weather index on a severe weather day
+Weather index on October 23, 2007
+Figure 5 .5Figure 5. Sector demand and weather indices with and without echo tops information on August 16 and October 23, 2007.
+Figure 6 .6Figure 6.Southwest region of superhigh (red) and high sector (blue) in ZID center
+ZID92 on August 21, 2007
+Figure 7 .7Figure 7. Sector weather indices on severe weather days in August 2007.
+Table 1 .1Sector demand prediction error of the PAR and CPAR models in September 2007.The model is built using August 2007 data.The smaller errors in each case are in bold.The unit is the number of aircraft.SectorAverage prediction RMS errorName MAP Model 15-min 30-min 45-min 60-min 75-min 90-min 105-min 120-minZID8117PAR CPAR1.95 1.982.05 2.012.07 2.012.09 2.022.06 2.022.08 2.032.09 2.052.11 2.06ZID8216PAR CPAR1.57 1.581.62 1.581.64 1.581.60 1.581.61 1.591.62 1.591.63 1.601.64 1.61ZID8316PAR CPAR1.63 1.581.67 1.591.67 1.611.70 1.621.71 1.631.71 1.641.71 1.651.72 1.65ZID8416PAR CPAR1.82 1.821.87 1.831.92 1.851.94 1.851.92 1.851.92 1.871.90 1.881.89 1.88ZID9119PAR CPAR2.06 2.042.13 2.052.13 2.062.10 2.062.12 2.072.14 2.082.11 2.092.16 2.09ZID9217PAR CPAR1.68 1.681.76 1.691.73 1.691.72 1.701.71 1.701.75 1.701.74 1.711.74 1.71ZID9319PAR CPAR2.11 2.102.20 2.112.21 2.122.19 2.122.23 2.132.24 2.152.23 2.142.24 2.15ZID9417PAR CPAR1.90 1.901.99 1.911.98 1.911.98 1.921.99 1.921.95 1.921.97 1.931.99 1.93
+		
+		
+			
+15-min prediction, and between 1.61 and 2.15 in the 120-min prediction.The performance of the prediction only decays slightly as the prediction interval is increased from 15-minute to 2-hour in both the PAR and CPAR models, as the error increases 5.12% in PAR models and 2.87% in the CPAR models.To improve the accuracy of sector demand prediction in the presence of severe weather, the paper introduced the concept of weather factor.For severe weather days, the model uses the three-dimensional weather information, considering both storm location and echo tops to form the weather factor and then adjusts the weather-free prediction.The weather-weighted model improves the sector demand prediction by as much as 34.73% for the 15-minute prediction, 27.81% for the 60-minute prediction, and 7.71% for the 120-minute prediction on the days and sectors tested.Unlike traditional trajectory-based sector demand prediction methods which predict the behavior of the National Airspace System adequately for short durations of up to 20 minutes and are vulnerable to weather uncertainties, the weather-weighted periodic auto-regressive models provide a reliable short-to mid-term sector demand prediction which accounts for weather uncertainty.			
+			
+
+				
+
+
+	
+		A New Model to Improve Aggregate Air Traffic Demand Predictions
+		
+			EugeneGilbo
+		
+		
+			ScottSmith
+		
+		10.2514/6.2007-6450
+	
+	
+		AIAA Guidance, Navigation and Control Conference and Exhibit
+		Hilton Head, SC
+		
+			American Institute of Aeronautics and Astronautics
+			Aug 2007
+		
+	
+	Gilbo, E. and Smith, S., "A New Model to Improve Aggregate Air Traffic Demand Predictions," AIAA Guidance, Navigation and Control Conference, Hilton Head, SC, Aug 2007.
+
+
+
+	
+		FACET: Future ATM Concepts Evaluation Tool
+		
+			KarlDBilimoria
+		
+		
+			BanavarSridhar
+		
+		
+			ShonRGrabbe
+		
+		
+			GanoBChatterji
+		
+		
+			KapilSSheth
+		
+		10.2514/atcq.9.1.1
+	
+	
+		Air Traffic Control Quarterly
+		Air Traffic Control Quarterly
+		1064-3818
+		2472-5757
+		
+			9
+			1
+			
+			2001
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	Bilimoria, K., Sridhar, B., Chatterji, G. B., Sheth, K., and Grabbe, S., "FACET: Future ATM Concepts Evaluation Tool," Air Traffic Control Quarterly, Vol. 9, No. 1, 2001, pp. 1-20.
+
+
+
+	
+		Methods for Establishing Confidence Bounds on Sector Demand Forecasts
+		
+			GanoChatterji
+		
+		
+			BanavarSridhar
+		
+		
+			KapilSheth
+		
+		
+			DouglasKim
+		
+		
+			DanielMulfinger
+		
+		10.2514/6.2004-5232
+	
+	
+		AIAA Guidance, Navigation, and Control Conference and Exhibit
+		Providence, RI
+		
+			American Institute of Aeronautics and Astronautics
+			August 2004
+		
+	
+	Chatterji, G. B., Sridhar, B., Sheth, K., Kim, D., and Mulfinger, D., "Methods for Establishing Confidence Bounds on Sector Demand Forecasts," AIAA Guidance, Navigation and Control Conference, Providence, RI, August 2004.
+
+
+
+	
+		Compressive Representations of Weather Scenes for Strategic Air Traffic Flow Management
+		
+			JEEvans
+		
+		10.2514/6.2022-4079.vid
+	
+	
+		Europe Air Traffic Management RD Seminar
+		
+			2001
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	Evans, J. E., "Tactical Weather Decision Support to Complement Strategic Traffic Flow Management for Convective Weather," Europe Air Traffic Management RD Seminar , 2001.
+
+
+
+	
+		Measuring Uncertainty in Airspace Demand Predictions for Traffic Flow Management Applications
+		
+			CraigWanke
+		
+		
+			MichaelCallaham
+		
+		
+			DanielGreenbaum
+		
+		
+			AnthonyMasalonis
+		
+		10.2514/6.2003-5708
+	
+	
+		AIAA Guidance, Navigation, and Control Conference and Exhibit
+		Austin, TX
+		
+			American Institute of Aeronautics and Astronautics
+			August 2003
+		
+	
+	Wanke, C. R., Callaham, M. B., Greenbaum, D. P., and Masalonis, A. J., "Measuring Uncertainty in Airspace Demand Predictions for Traffic Flow Management Applications," AIAA Guidance, Navigation and Control Conference, Austin, TX, August 2003.
+
+
+
+	
+		Modeling Traffic Prediction Uncertainty for Traffic Management Decision Support
+		
+			CraigWanke
+		
+		
+			SandeepMulgund
+		
+		
+			DanielGreenbaum
+		
+		
+			LixiaSong
+		
+		10.2514/6.2004-5230
+	
+	
+		AIAA Guidance, Navigation, and Control Conference and Exhibit
+		Providence, RI
+		
+			American Institute of Aeronautics and Astronautics
+			August 2004
+		
+	
+	Wanke, C. R., Mulgund, S., and Song, L., "Modeling Traffic Prediction Uncertainty for Traffic Management Decision Support," AIAA Guidance, Navigation and Control Conference, Providence, RI, August 2004.
+
+
+
+	
+		System Identification: Theory for the User
+		
+			LLjung
+		
+		
+			1999
+			Prentice Hall
+			Englewood Cliffs, NJ, 2nd ed.
+		
+	
+	Ljung, L., System Identification: Theory for the User , Prentice Hall, Englewood Cliffs, NJ, 2nd ed., 1999.
+
+
+
+	
+		Multivariate periodic time series models
+		
+			PhilipHansFranses
+		
+		
+			RichardPaap
+		
+		10.1093/019924202x.003.0005
+	
+	
+		Periodic Time Series Models
+		London, UK
+		
+			Oxford University PressOxford
+			2003
+			
+		
+	
+	Franses, P. and Papp, R., Periodic Time Series Models, Oxford Univ. Press, London, UK, 2003.
+
+
+
+	
+		Short Term National Airspace System Delay Prediction Using Weather Impacted Traffic Index
+		
+			BanavarSridhar
+		
+		
+			NeilChen
+		
+		10.2514/6.2008-7395
+	
+	
+		AIAA Guidance, Navigation and Control Conference and Exhibit
+		Honolulu, HI
+		
+			American Institute of Aeronautics and Astronautics
+			Aug 2008
+		
+	
+	Sridhar, B. and Chen, N., "Short Term National Airspace System Delay Prediction Using Weather Impacted Traffic Index," AIAA Guidance, Navigation and Control Conference, AIAA, Honolulu, HI, Aug 2008.
+
+
+
+	
+		Estimation of Air Traffic Delay Using Three Dimensional Weather Information
+		
+			NeilChen
+		
+		
+			BanavarSridhar
+		
+		10.2514/6.2008-8916
+	
+	
+		The 26th Congress of ICAS and 8th AIAA ATIO
+		Anchorage, AK
+		
+			American Institute of Aeronautics and Astronautics
+			Sep 2008. 11
+		
+	
+	Chen, N. and Sridhar, B., "Estimation of Air Traffic Delay Using Three Dimensional Weather Information," The 8th AIAA Aviation Technology, Integration, and Operations Conference, AIAA, Anchorage, AK, Sep 2008. 11
+
+
+
+	
+		Analysis of En Route Sector Demand Error Sources
+		
+			JimmyKrozel
+		
+		
+			DanRosman
+		
+		
+			ShonGrabbe
+		
+		10.2514/6.2002-5016
+	
+	
+		AIAA Guidance, Navigation, and Control Conference and Exhibit
+		Monterey, CA
+		
+			American Institute of Aeronautics and Astronautics
+			August 2002
+		
+	
+	Krozel, J., Rosman, D., and Grabbe, S., "Analysis of En Route Sector Demand Error Sources," AIAA Guidance, Navigation and Control Conference, Monterey, CA, August 2002.
+
+
+
+	
+		Predicting Sector Capacity under Severe Weather Impact for Traffic Flow Management
+		
+			LixiaSong
+		
+		
+			CraigWanke
+		
+		
+			DanielGreenbaum
+		
+		
+			DavidCallner
+		
+		10.2514/6.2007-7887
+	
+	
+		7th AIAA ATIO Conf, 2nd CEIAT Int'l Conf on Innov and Integr in Aero Sciences,17th LTA Systems Tech Conf; followed by 2nd TEOS Forum
+		Belfast, Northern Ireland
+		
+			American Institute of Aeronautics and Astronautics
+			September 2007
+		
+	
+	Song, L., Wanke, C., Greenbaum, D., and Callner, D., "Predicting Sector Capacity under Severe Weather Impact for Traffic Flow Management," The 7th AIAA Aviation Technology, Integration, and Operations Conference, Belfast, Northern Ireland, September 2007.
+
+
+
+	
+		Methodologies of Estimating the Impact of Severe Weather on Airspace Capacity
+		
+			LixiaSong
+		
+		
+			CraigWanke
+		
+		
+			StephenZobell
+		
+		
+			DanielGreenbaum
+		
+		
+			ClaudeJackson
+		
+		10.2514/6.2008-8917
+	
+	
+		The 26th Congress of ICAS and 8th AIAA ATIO
+		Anchorage,AK
+		
+			American Institute of Aeronautics and Astronautics
+			September 2008
+		
+	
+	The 26th Congress of International Council of the Aeronautical Sciences
+	Song, L., Wanke, C., Greenbaum, D., Zobell, S., and Jackson, C., "Methodologies of Estimating the Impact of Severe Weather on Airspace Capacity," The 26th Congress of International Council of the Aeronautical Sciences, Anchorage,AK, September 2008.
+
+
+
+	
+		Description of the Corridor Integrated Weather System (CIWS) Weather Products
+		
+			JEvans
+		
+		
+			DKlingle-Wilson
+		
+		ATC-317
+		
+			Aug 2005
+		
+		
+			MIT Lincoln Laboratory
+		
+	
+	Project Report
+	Evans, J. and Klingle-Wilson, D., "Description of the Corridor Integrated Weather System (CIWS) Weather Products," Project Report ATC-317, MIT Lincoln Laboratory, Aug 2005.
+
+
+
+	
+		Model Estimates of Traffic Reduction in Storm Impacted En Route Airspace
+		
+			BrianMartin
+		
+		10.2514/6.2007-7889
+	
+	
+		7th AIAA ATIO Conf, 2nd CEIAT Int'l Conf on Innov and Integr in Aero Sciences,17th LTA Systems Tech Conf; followed by 2nd TEOS Forum
+		Belfast, Northern Ireland
+		
+			American Institute of Aeronautics and Astronautics
+			September 2007
+		
+	
+	Martin, B., "Model Estimates of Traffic Reduction in Storm Impacted En Route Airspace," The 7th AIAA Aviation Technology, Integration, and Operations Conference, Belfast, Northern Ireland, September 2007.
+
+
+				
+			
+		
+	

file137.txt

@@ -0,0 +1,517 @@
+
+	
+	
+		
+I. IntroductionAircraft-induced environmental impact has drawn attention in recent years. 1 The three largest emission impacts include direct emissions of greenhouse gases such as CO 2 , emissions of NOx, and persistent contrails.Contrails are clouds that are visible trails of water vapor made by the exhaust of aircraft engines.Contrails form when a mixture of warm engine exhaust gases and cold ambient air reaches saturation with respect to water, forming liquid drops which quickly freeze.Contrails persist if aircraft are flying in certain atmospheric conditions.Persistent contrails reduce incoming solar radiation and outgoing thermal radiation in a way that accumulates heat. 2 The global mean contrail coverage in 1992 was estimated to double by 2015, and quadruple by 2050 due to predicted increase in air traffic. 3Studies suggest that the environmental impact from persistent contrails are estimated to range from three to four times, 4 to ten times 5 larger than from aviation-induced emissions.To address minimizing environmental impacts due to contrails, methods to reduce aircraft induced persistent contrails have been proposed.Various approaches have been proposed in the past to reduce the persistent contrail formation.The approach based on changing aircraft flight altitudes looks promising.Mannstein 6 proposed a strategy to reduce the climate impact of contrails significantly by only small changes in individual flight altitude.Fichter 7 showed that the global annual mean contrail coverage could be reduced by downshifting the cruise altitude.Williams 8,9 proposed strategies for contrail reduction by identifying fixed and varying maximum altitude restriction policy.These restrictions generally imply more fuel burn, thus more emissions, and add congestion to the already crowded airspace at lower altitudes.Sridhar 10 and Chen 11 proposed contrail reduction strategies by altering an aircraft's cruising altitude in a fuel-efficient way.The strategies were designed without increasing congestion in the airspace.However, none of the above strategies take into account the range and duration of an aircraft's flight.The objective of this paper is to evaluate contrail reduction strategies based on an aircraft's flight distance.Contrail reduction strategies have different effects on aircraft with different flight distances.In general, aircraft with longer flight distances cruise at the altitudes where contrails are more likely to form because of cold air temperatures.The concept of the contrail frequency index is used to quantify contrail formation.The strategy of reducing persistent contrail formation is to minimize the contrail frequency index by altering the aircraft's cruising altitude.A user-defined factor was applied to evaluate the tradeoff between contrail reduction and extra CO 2 emissions.A higher tradeoff factor results in more contrail reduction and extra CO 2 emissions.Contrail reduction strategies using different tradeoff factors behave differently for different flight distances.For this analysis, the flights during a day were divided into four groups: flights with flight distances less than 500 miles (short-distance flights), between 500 and 1,000 miles (medium-distance flights), between 1,000 and 1,500 miles (long-distance flights), and more than 1,500 miles (transcontinental flights).The remainder of the paper is organized as follows.Section II provides descriptions of the contrail model, definition of contrail frequency index, and the contrail reduction strategies.Next, Section III shows the results and analysis of contrail reduction strategies applied to different ranges of flights.Finally, Section IV presents a summary and conclusions.
+II. Models and Strategies
+II.A. Contrail Model and Contrail Frequency IndexThis paper follows the contrail models described in Ref. 11.The contrail models use atmospheric temperature and humidity data retrieved from the Rapid Updated Cycle (RUC) data, provided by the National Oceanic and Atmospheric Administration (NOAA).The horizontal resolution in RUC is 13-km with 37 vertical isobaric pressure levels ranging between 100 and 1000 millibar (mb) in 25 mb increments.Since the vertical isobaric pressure levels do not correspond to 2,000 feet increments, linear interpolation was used to convert the RUC data to a vertical range from 26,000 feet to 44,000 feet with increments of 2,000 feet.This range is chosen because it generally is too warm for contrails to form below 26,000 feet and most commercial aircraft fly below 44,000 feet.The 2,000 feet increment is chosen is because in general same direction of flights have a vertical separation range of 2,000 feet due to the standard in Reduced Vertical Separation Minima. 12These modifications result in dividing the U.S. national airspace into a three dimensional grid with 337 elements along the latitude, 451 elements along the longitude, and 10 altitudes ranging from 26,000 feet to 44,000 feet.Contrails form when a mixture of warm engine exhaust gases and cold ambient air reaches saturation with respect to water, forming liquid drops which quickly freeze.Contrails form in the regions of airspace that have ambient Relative Humidity with respect to Water (RHw) greater than a critical value r contr . 13Regions with RHw greater than or equal to 100% are excluded because clouds are already present. 14Contrails can persist when the environmental Relative Humidity with respect to Ice (RHi) is greater than 100%. 15In this paper, contrail favorable regions are defined as the regions of airspace that have r contr ≤ RHw < 100% and RHi ≥ 100%.The contrail frequency index (CFI) is used to quantify the severity of contrail activities and represents the number of aircraft in a volumetric element which meets conditions for persistent contrail formation.Air traffic in the U.S. can be mapped into the same volumetric grid as in the RUC data.The Contrail frequency index is zero for volumetric elements which do not meet the conditions for persistent contrail formation.Precise definitions of contrail frequency index are provided in Ref. 11.
+II.B. Contrail Reduction StrategiesThis paper uses the contrail reduction strategies described in Ref. 11.That strategy for reducing the persistent contrail formations is to minimize contrail frequency index by altering the aircraft's cruising altitude.Note that these altitude changes are subject to the cruise altitude limits of each aircraft.An additional constraint is added such that where an aircraft crosses a sector boundary and causes congestion, it will stay at the original cruise altitude.Additional conditions can be added to satisfy other operational procedures.An Air Route Traffic Control Center (or Center) is divided into sectors horizontally and vertically which is monitored by an air traffic controller to maintain separation between aircraft.The number of aircraft in a sector is kept below a maximum, referred to as the Monitor Alert Parameter (MAP), to keep the controllers workload within acceptable limits. 16Therefore the MAP can be used to define the airspace capacity.The contrail reduction altitude changes will not change the sector counts unless they cross the sector boundaries.To address this operational constraints, the strategies only allow the altitude changes such that the aircraft count in a sector does not exceed the sector capacity after the altitude changes.Consider the traffic situation at Kansas City Center at 8AM Eastern Daylight Time (EDT) on April 23, 2010.Kansas City Center has 15 high sectors and 11 super-high sectors.Among them, sector 31 has the highest sector count during the hour.Sector 31 has a lower bound of 37,000 feet and is on top of sector 28, 29 and 30, shown in Fig. 1.During the hour, altitude 35,000-36,999 feet has been identified as a high contrail area.The contrail reduction strategy suggests to increase the cruise altitude of those aircraft passing through this area by 2,000 feet.The move would reduce the contrail frequency index by 17.Now consider if the move would cause congestion by examining the sector counts and capacities.The aircraft cruise altitude changing from 35,000-36,999 feet to 37,000-38,999 feet would move some aircraft in sector 28, 29 and 30 to sector 31.Sector 28, 29, 30 and 31 have the MAP values of 18, 18, 19 and 21 respectively.Figure 2 shows the MAP values and the sector counts in sector 28, 29, 30 and 31 before and after the altitude changes.The aircraft counts in sector 28, 29 and 30 decrease because some aircraft have been moved up to sector 31; the sector count in sector 31 increases but is still lower than the sector capacity of 21.Thus the contrail reduction altitude changes are applied without exceeding the capacity of the airspace.In addition, contrail reduction altitude changes are only applied when the aircraft enter a new Center.The number of altitude changes is not expected to result in frequent climbs and descents to affect current operations.However, if needed, additional constraints can be imposed on the number of altitude changes.All flights over the United States National Airspace from a 24-hour period on April 23, 2010 were analyzed.The contrail reduction strategies were applied and the results are shown in Fig. 3.The CFIs for a Center before applying the maximum reduction strategy are shown as dark blue bars.When the aircraft altitudes are allowed to alter by 2,000 feet, the center CFIs after reduction are shown as light blue bars.The total CFI reduction among all centers is 62%.When the aircraft altitudes are allowed to alter by 4,000 feet, the total reduction is 88% as indicated as green bars.The strategies in this paper limit the altitude changes to 4,000 feet.Altering cruising altitudes changes the aircraft fuel consumption and emissions.In order to analyze the environmental impact of contrail reduction strategies, fuel consumption and emissions are considered in the strategies.Fuel burn and emissions computations are based on a prototype of the Aviation Environmental Design Tool (AEDT) developed by the Federal Aviation Administration (FAA). 17Considering the relative environmental impact of emissions and contrails, the aircraft altitudes are modified only if the contrail reduction benefits exceed the environmental impact of additional emissions.The strategy uses a user-defined trade-off factor α to determine whether the strategy should apply to an aircraft.It can be interpreted as the equivalent emissions in kg that the user is willing to trade off for a contrail frequency index of 1.In general, higher α would result in more contrail reduction and extra CO 2 emissions.Figure 4 shows the amount of contrail reductions versus extra CO 2 emissions using different α values when the aircraft altitudes are allowed to alter by 4,000 feet on April 23, 2010.In the figure, more contrail reduction takes place from left to right and more CO 2 emissions occurs from bottom to top.At the lower-left point, no reduction strategy (α = 0) is applied.The upper-right point is the maximal reduction strategy (α = ∞).As the value of α increases, the curve moves from lower-left to upper-right.The user-defined trade-off factor α provides a flexible way to trade off between contrail reduction and extra emissions.Better understanding of the trade-offs between contrails and emissions and impact on the climate needs to be developed to fully utilize this class of contrail reduction strategies.
+III. Analysis
+III.A. Contrail Frequency IndexThe relative climate impacts of long haul and short haul air travel were studied previously. 18For long distance flights, the fraction of the flight time spent in the high-thrust take-off and climb-out is smaller than the short distance flights.Therefore, long distance flights are more fuel efficient and generate less CO 2 emissions per unit distance than short distance flights.However, short distance flights generally cruise at lower altitudes where contrails are less likely to form because the temperature is too warm.Therefore, long distance flights create more contrail impact than short distance flights.The combined climate impacts of contrails and CO 2 emissions from long and short distance flights need further investigation.This section evaluates the effect of flight distances on contrail reduction strategies.This paper defines the flight distance as the great circle distance between the origin and destination of a flight plan.The flights during a day are divided into four groups: flights with flight distance less than 500 miles (short-distance flights), between 500 and 1,000 miles (medium-distance flights), between 1,000 and 1,500 miles (long-distance flights), and more than 1,500 miles (transcontinental flights).The reason for dividing flights into such groups is to have comparable total flight distances in each group.Based on the flight data on April 23, 2010, 43% of short-distance flights have cruise altitudes lower than 24, 000 feet, which are not likely to form contrails because of the warm temperature.On the other hand, most medium, long, and transcontinental flights cruise at high altitudes, and less than 2% of flights with flight distance more than 500 miles have cruise altitudes lower than 24, 000 feet.Data from all flights during the month of April 2010 were analyzed, and in that month, April 12, April 19 and April 3 had the highest CFI for all flights during a 24-hour period.Considering the contrail activities of different ranges of flights for these three days, the number of flights, total flight distance, total CFI and CFI per 1000 miles of range are summarized in Table 1.The group of short-distance flights has the most flights but the lowest CFI, less than 5% of the total CFI for all flights.The short-distance flights have only 0.6 to 1 CFI per flight on average, or 2.3 to 3.5 CFI per 1000 miles.Even excluding the flights with a cruise altitude lower than 24, 000 feet, the average CFI per flight is 1.1 to 1.7 and the CFI per 1000 miles is 3.1 to 4.8; they are still lower than the averages in other groups.The group of transcontinental flights has the fewest flights and the greatest CFI, on April 12, among all four groups.However, on April 3 and19, its total CFIs are less than the total of groups of medium-and long-distance flights.For the month of April, transcontinental flights have the greatest CFI on 15 days.It seems that the total CFIs of transcontinental flights are more sensitive to the locations of contrail areas than the other groups, therefore the CFIs for this group has larger variance.The group of medium-distance flights has the greatest total distance.Even though its total CFI is more than in the group of long distance flights, it is consistent that its CFI per flight and CFI per 1000 miles are lower than the CFI in long-distance flights.The groups of long-distance and transcontinental flights have the grestest CFI per flight and CFI per unit distance.However, the CFI per unit distance for the group of transcontinental flights drop below the group of long and medium flights on April 19 and April 3, 2010.It is consistent that longer range of flights has more CFI per flight and, with the exception of the group of transcontinental flights, per unit distance.
+III.B. Contrail Reduction StrategiesThe contrail reduction strategies using different values of trade-off factor α were applied to all four groups. .The group of short-distance flights has a much smaller CFI therefore the tradeoff curves (blue) in the figures are relatively short and are located at the lower left corner.The groups of medium-distance (green curves) and long-distance (red curves) flights have similar trends where the medium-distance flight is on the right because of higher CFI to be reduced.This is true for 29 days in April, with an exception on April 28.It is noticed that the locations of the curves (light blue) for the group of transcontinental flights are not consistent across the three figures.This is also true for the entire month, which suggests that the efficiency of contrail reduction strategies for the group of transcontinental flight has different characteristics than in the other groups.It can be interpreted that the transcontinental flights have longer flight distances and the contrail reduction efficiencies are sensitive to the locations of the contrail areas.Also note that in Fig. 5c, the second dot (α = 10) from the left of the light blue curve (transcontinental flights) has a negative value of extra CO 2 emissions.This indicates that the strategy found a way to reduce both contrails and CO 2 emissions for the transcontinental flights on April 3, 2010.Table 2 summarizes the amount of contrail reductions with different α values for all four groups.The numbers are the CFI reductions and the percentages are over total reductions during the day.The shortdistance flights contribute the least reductions, only 9.6% to 12.6% using different α values on April 12, 7.3% to 8.8% on April 19, and 3.9% to 4.3% on April 3.As described in the previous paragraph, the contrail reduction efficiencies for the group of transcontinental flights are sensitive to the locations of the contrail areas, therefore, there is no obvious trend for the reduction efficiencies compared with other groups.It contributes the most reductions, 36.3%, among all reductions for the maximum reduction strategy on April 12, but the reduction rates decay with smaller α values.Also, for April 3 and 19, the reduction rates are smaller than the groups of medium-and long-distance flights.Among the groups of short-, medium-, and long-distance flights, it is consistent that the group of medium flights contributes the most reduction rate, with 33.3% on April 12, 39.9% on April 19, and 39.5% on April 3 for the maximum reduction strategies.It is noticed that the contribution rates of medium-distance flights increase with smaller α values (smaller α value means less CO 2 emissions).For α = 10, the reduction rate increases to 38.1% on April 12, 43.8% on April 19, and 42.6% on April 3 for the maximum reduction strategies.The strategies are more efficient for transcontinental flights with larger α values and more efficient for short-and medium-distance flights with smaller α values.The percentages of the reductions are similar with different α values for long-distance flights.Table 3 summarizes the CFI per 1,000 miles after the contrail reduction strategies were applied.The CFI per 1,000 miles for medium, long, and transcontinental flights can be reduced, from 9.0, 10.7 and 20.0 to 2.4, 3.5 and 8.1, respectively, on April 12; from 12.5, 16.8 and 12.0 to 2.5, 4.1 and 2.6, respectively, on April 19; from 13.1, 16.4 and 11.1 to 2.5, 3.5 and 1.6, respectively, on April 3, an average reduction of 75%.The reduction rates of CFI per 1,000 miles in medium-distance flights are larger than the rates in long-distance flights but the absolute reduction values are smaller.The contrail reduction performance for the group of transcontinental flights varies with the days selected.There is no obvious trend compared with other groups.
+IV. ConclusionsDifferent concepts of contrail reduction strategies based on range of flights have been analyzed and evaluated.The concept of the contrail frequency index is used to quantify the contrail formations.The proposed strategy for reducing the persistent contrail formations is to minimize the contrail frequency index by altering the aircraft's cruising altitude within 4,000 feet.A user-defined tradeoff factor was used to trade off between contrail reductions and extra CO 2 emissions.A high value of tradeoff factor results in more contrail reduction but more CO 2 emissions.The results from an analysis of a month of data show that the groups of short distance flights (< 500 miles) contributes the least to contrail reduction when the strategy is applied.The contrail reduction performance for the group of transcontinental flights (> 1500 miles) varies with the days selected.Among the groups of short-distance, medium-distance (500 to 1000 miles), and long-distance (1000 to 1500 miles) flights, when the strategy is applied, it is consistent that the group of medium-distance flights contributes the most contrail reduction during a day.The strategies are more efficient for transcontinental flights with larger α values and more efficient for short-and medium-distance flights with smaller α values.The percentages of the reductions are similar with different α values for longdistance flights.For the top three contrail days in April, 2010, the contrail frequency index per 1,000 miles for medium-range, long-range, and transcontinental flights can be reduced by an average of 75%.In general, the short-distance flights are more frequent but contribute least to contrail reduction, therefore the group has the lowest priority when applying the contrail reduction strategy.The group of medium-distance flights has a higher priority if the goal is to achieve maximum contrail reduction, the group of long-distance flights has a higher priority if the goal is to achieve maximum contrail reduction per flight.The characteristics of the group of transcontinental flights vary with weather so the priority of the group needs to be further evaluated based on the locations of the contrail areas during the day.The results provide a starting point for developing operational policies to reduce the impact of aviation on climate based on aircraft flight distances.Figure 1 .1Figure 1.Kansas City Center sector 28, 29, 30 and 31.
+Figure 2 .Figure 3 .23Figure 2. MAP values and sector counts before and after the contrail reduction strategies at 8AM EDT on April 23, 2010.
+Figure 4 .4Figure 4. Contrail reduction versus extra CO2 emissions with different α values for all flights on April 23, 2010.
+Figure 55shows the amount of contrail reductions versus extra CO 2 emissions using different α values for different flight ranges on April 3, 12 and 19, 2010.The strategies limit the aircraft cruise altitude changes to 4,000 feet.In the figures, different colors indicate groups of different flight ranges.Each curve has six dots showing different strategies (right to left: α=∞, 80, 40, 20, 10, 0, where α = ∞ is maximum reduction and α = 0 is no reduction).For example, in Fig 5a, the second from the right dot of the light blue curve indicates that the contrail reduction strategy applied to transcontinental flights with a trade-off factor α = 80.It reduced the CFI on April 12 by around 40,000 with extra CO 2 emissions of about 1,000 metric ton (1,000 kg)
+Figure 5 .5Figure 5. Contrail reduction versus extra CO2 emissions with different α values on April 12, April 19 and April 3, 2010.
+Table 1 .1Summary of contrail activities for different distances of flights.daterangenumber total distanceCFICFICFIof flightsof flights(1000 miles)(total) (per flight) (per 1000 miles)short1321236721279613.5April 12, 2010medium long8096 28645814 337852504 360216.5 12.69 10.7transcontinental195333786742034.520short133653707115280.93.1April 19, 2010medium long8250 27635932 325774249 546539 19.812.5 16.8transcontinental195333994072120.912short9423263760710.62.3April 3, 2010medium long6592 26874794 314362774 515049.5 19.213.1 16.4transcontinental170530423363319.711.1
+Table 2 .2Results of contrail reduction in CFIs with different α values.daterange of flights max reduction α=40 α=10short9.6%10.6% 12.6%April 12, 2010medium long33.3% 20.8%34.9% 38.1% 20.2% 20.9%transcontinental36.3%34.3% 28.3%short7.3%7.7%8.8%April 19,2010medium long39.9% 29.3%41.2% 43.8% 29.9% 29.9%transcontinental23.5%21.2% 17.5%short3.9%4.1%4.3%April 3,2010medium long39.5% 32.3%40.8% 42.6% 32.3% 32.2%transcontinental24.2%22.8% 20.9%
+Table 3 .3Contrail frequency index per 1000 miles after reduction with different α values.daterange of flights no reduction α=10 α=40 max reductionshort3.51.20.50.4April 12, 2010medium long9.0 10.75.7 8.13 4.62.4 3.5transcontinental20.717.710.28.1short3.110.40.3April 19, 2010medium long12.5 16.86.7 113 52.5 4.1transcontinental12.09.94.32.6short2.30.70.30.2April 3, 2010medium long13.1 16.46.4 92.9 4.42.5 3.5transcontinental11.17.42.91.6
+		
+		
+			
+
+				
+
+
+	
+		
+			IWaitz
+		
+		
+			JTownsend
+		
+		
+			JCutcher-Gershenfeld
+		
+		
+			EGreitzer
+		
+		
+			JKerrebrock
+		
+		Report to the United States Congress: Aviation and the Environment, A National Vision, Framework for Goals and Recommended Actions
+		London, UK
+		
+			December 2004
+		
+	
+	Tech. rep
+	Partnership for AiR Transportation Noise and Emissions Reduction
+	Waitz, I., Townsend, J., Cutcher-Gershenfeld, J., Greitzer, E., and Kerrebrock, J., "Report to the United States Congress: Aviation and the Environment, A National Vision, Framework for Goals and Recommended Actions," Tech. rep., Partnership for AiR Transportation Noise and Emissions Reduction, London, UK, December 2004.
+
+
+
+	
+		Radiative forcing by contrails
+		
+			RMeerkötter
+		
+		
+			USchumann
+		
+		
+			DRDoelling
+		
+		
+			PMinnis
+		
+		
+			TNakajima
+		
+		
+			YTsushima
+		
+		10.1007/s00585-999-1080-7
+	
+	
+		Annales Geophysicae
+		Ann. Geophys.
+		1432-0576
+		
+			17
+			8
+			
+			1999
+			Copernicus GmbH
+		
+	
+	Meerkotter, R., Schumann, U., Doelling, D. R., Minnis, P., Nakajima, T., and Tsushima, Y., "Radiative forcing by contrails," Annales Geophysicae, Vol. 17, 1999, pp. 1080-1094.
+
+
+
+	
+		Future Development of Contrail Cover, Optical Depth, and Radiative Forcing: Impacts of Increasing Air Traffic and Climate Change
+		
+			SMarquart
+		
+		
+			MPonater
+		
+		
+			FMager
+		
+		
+			RSausen
+		
+		10.1175/1520-0442(2003)016<2890:fdocco>2.0.co;2
+	
+	
+		Journal of Climate
+		J. Climate
+		0894-8755
+		1520-0442
+		
+			16
+			17
+			
+			September 2003
+			American Meteorological Society
+		
+	
+	Marquart, S., Ponater, M., Mager, F., and Sausen, R., "Future Development of Contrail Cover, Optical Depth, and Radiative Forcing: Impacts of Increasing Air Traffic and Climate Change," Journal of Climate, Vol. 16, September 2003, pp. 2890-2904.
+
+
+
+	
+		A Standing Royal Commission
+		
+			SusanOwens
+		
+		10.1093/acprof:oso/9780198294658.003.0003
+	
+	
+		Knowledge, Policy, and Expertise
+		London, UK
+		
+			Oxford University Press
+			2002
+			
+		
+	
+	"The Environmental Effects of Civil Aircraft in Flight," Tech. rep., Royal Commission on Environmental Pollution, London, UK, 2002.
+
+
+
+	
+		Aircraft induced contrail cirrus over Europe
+		
+			HermannMannstein
+		
+		
+			UlrichSchumann
+		
+		10.1127/0941-2948/2005/0058
+	
+	
+		Meteorologische Zeitschrift
+		metz
+		0941-2948
+		
+			14
+			4
+			
+			2005
+			Schweizerbart
+		
+	
+	Mannstein, H. and Schumann, U., "Aircraft induced contrail cirrus over Europe," Meteorologische Zeitschrift, Vol. 14, No. 4, 2005, pp. 549-554.
+
+
+
+	
+		A note on how to avoid contrail cirrus
+		
+			HermannMannstein
+		
+		
+			PeterSpichtinger
+		
+		
+			KlausGierens
+		
+		10.1016/j.trd.2005.04.012
+	
+	
+		Transportation Research Part D: Transport and Environment
+		Transportation Research Part D: Transport and Environment
+		1361-9209
+		
+			10
+			5
+			
+			September 2005
+			Elsevier BV
+		
+	
+	Mannstein, H., Spichtinger, P., and Gierens, K., "A note on how to avoid contrail cirrus," Transportation Research. Part D, Transport and environment, Vol. 10, No. 5, September 2005, pp. 421-426.
+
+
+
+	
+		The impact of cruise altitude on contrails and related radiative forcing
+		
+			ChristineFichter
+		
+		
+			SusanneMarquart
+		
+		
+			RobertSausen
+		
+		
+			DavidSLee
+		
+		10.1127/0941-2948/2005/0048
+	
+	
+		Meteorologische Zeitschrift
+		metz
+		0941-2948
+		
+			14
+			4
+			
+			August 2005
+			Schweizerbart
+		
+	
+	Fichter, C., Marquart, S., Sausen, R., and Lee, D. S., "The impact of cruise altitude on contrails and related radiative forcing," Meteorologische Zeitschrift, Vol. 14, No. 4, August 2005, pp. 563-572.
+
+
+
+	
+		Reducing the climate change impacts of aviation by restricting cruise altitudes
+		
+			VictoriaWilliams
+		
+		
+			RobertBNoland
+		
+		
+			RalfToumi
+		
+		10.1016/s1361-9209(02)00013-5
+	
+	
+		Transportation Research Part D: Transport and Environment
+		Transportation Research Part D: Transport and Environment
+		1361-9209
+		
+			7
+			6
+			
+			November 2002
+			Elsevier BV
+		
+	
+	Williams, V., Noland, R. B., and Toumi, R., "Reducing the climate change impacts of aviation by restricting cruise altitudes," Transportation Research. Part D, Transport and environment, Vol. 7, No. 6, November 2002, pp. 451-464.
+
+
+
+	
+		Variability of contrail formation conditions and the implications for policies to reduce the climate impacts of aviation
+		
+			VictoriaWilliams
+		
+		
+			RobertBNoland
+		
+		10.1016/j.trd.2005.04.003
+	
+	
+		Transportation Research Part D: Transport and Environment
+		Transportation Research Part D: Transport and Environment
+		1361-9209
+		
+			10
+			4
+			
+			July 2005
+			Elsevier BV
+		
+	
+	9 Williams, V. and Noland, R. B., "Variability of contrail formation conditions and the implications for policies to reduce the climate impacts of aviation," Transportation Research. Part D, Transport and environment, Vol. 10, No. 4, July 2005, pp. 269-280.
+
+
+
+	
+		Fuel efficient strategies for reducing contrail formations in United States airspace
+		
+			BanavarSridhar
+		
+		
+			NeilYChen
+		
+		10.1109/dasc.2010.5655533
+	
+	
+		29th Digital Avionics Systems Conference
+		Salt Lake City, UT
+		
+			IEEE
+			October 2010
+		
+	
+	Sridhar, B., Chen, N. Y., and Ng, H. K., "Fuel Efficient Strategies for Reducing Contrail Formations in United State National Air Space," 29th Digital Avionics Systems Conference, Salt Lake City, UT, October 2010.
+
+
+
+	
+		Tradeoff Between Contrail Reduction and Emissions in United States National Airspace
+		
+			NeilYChen
+		
+		
+			BanavarSridhar
+		
+		
+			HokKNg
+		
+		10.2514/1.c031680
+	
+	
+		Journal of Aircraft
+		Journal of Aircraft
+		0021-8669
+		1533-3868
+		
+			49
+			5
+			
+			2012
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	Chen, N. Y., Sridhar, B., and Ng, H. K., "Tradeoff between Contrail Reduction and Emissions in United States National Airspace," Journal of Aircraft, 2012, accepted.
+
+
+
+	
+		Über Bedingungen zur Bildung von Kondensstreifen aus Flugzeugabgasen
+		
+			UlrichSchumann
+		
+		10.1127/metz/5/1996/4
+	
+	
+		Meteorologische Zeitschrift
+		metz
+		0941-2948
+		
+			5
+			1
+			
+			1996
+			Schweizerbart
+		
+	
+	Schumann, U., "On Conditions for Contrail Formation from Aircraft Exhausts," Meteorologische Zeitschrift, Vol. 5, No. 1, 1996, pp. 4-23.
+
+
+
+	
+		Contrails in a comprehensive global climate model: Parameterization and radiative forcing results
+		
+			MichaelPonater
+		
+		
+			SMarquart
+		
+		
+			RSausen
+		
+		10.1029/2001jd000429
+	
+	
+		Journal of Geophysical Research
+		J. Geophys. Res.
+		0148-0227
+		
+			107
+			D13
+			
+			2002
+			American Geophysical Union (AGU)
+		
+	
+	Ponater, M., Marquart, S., and Sausen, R., "Contrails in a Comprehensive Global Climate Model: Parameterization and Radiative Forcing Results," Journal of Geophysical Research, Vol. 107, No. D13, 2002, pp. ACL 2-1.
+
+
+
+	
+		Determination of humidity and temperature fluctuations based on MOZAIC data and parametrisation of persistent contrail coverage for general circulation models
+		
+			KMGierens
+		
+		
+			USchumann
+		
+		
+			HG JSmit
+		
+		
+			MHelten
+		
+		
+			GZängl
+		
+		10.1007/s00585-997-1057-3
+	
+	
+		Annales Geophysicae
+		Ann. Geophys.
+		1432-0576
+		
+			15
+			8
+			
+			1997
+			Copernicus GmbH
+		
+	
+	Gierens, K. M., Schumann, U., Smit, H. G. J., Helten, M., and Zangl1, G., "Determination of humidity and temperature fluctuations based on MOZAIC data and parametrisation of persistent contrail coverage for general circulation models," Annales Geophysicae, Vol. 15, 1997, pp. 1057-1066.
+
+
+
+	
+		Estimated contrail frequency and coverage over the contiguous United States from numerical weather prediction analyses and flight track data
+		
+			DavidPDuda
+		
+		
+			PatrickMinnis
+		
+		
+			RabindraPalikonda
+		
+		10.1127/0941-2948/2005/0050
+	
+	
+		Meteorologische Zeitschrift
+		metz
+		0941-2948
+		
+			14
+			4
+			
+			June-July 2003
+			Schweizerbart
+			Friedrichshafen at Lake Constance, Germany
+		
+	
+	Duda, D. P., Minnis, P., Costulis, P. K., and Palikonda, R., "CONUS Contrail Frequency Estimated from RUC and Flight Track Data," European Conference on Aviation, Atmosphere, and Climate, Friedrichshafen at Lake Constance, Germany, June- July 2003.
+
+
+
+	
+		Impact of Uncertainty on the Prediction of Airspace Complexity of Congested Sectors
+		
+			BanavarSridhar
+		
+		
+			DeepakKulkarni
+		
+		
+			KapilSheth
+		
+		10.2514/atcq.19.1.1
+	
+	
+		Air Traffic Control Quarterly
+		Air Traffic Control Quarterly
+		1064-3818
+		2472-5757
+		
+			19
+			1
+			
+			2011. October 2010
+			American Institute of Aeronautics and Astronautics (AIAA)
+			Washington, DC
+		
+	
+	17 Federal Aviation Administration
+	Sridhar, B., Kulkarni, D., and Sheth, K., "Impact of Uncertainty on the Prediction of Airspace Complexity of Congested Sectors," Air Traffic Control Quarterly, Vol. 19, No. 1, 2011, pp. 1-23. 17 Federal Aviation Administration, Washington, DC, Aviation Environmental Design Tool (AEDT) User Guide: Beta1c, October 2010.
+
+
+
+	
+		Comparing the CO2 emissions and contrail formation from short and long haul air traffic routes from London Heathrow
+		
+			VictoriaWilliams
+		
+		
+			RobertBNoland
+		
+		10.1016/j.envsci.2005.10.004
+	
+	
+		Environmental Science & Policy
+		Environmental Science & Policy
+		1462-9011
+		
+			9
+			5
+			
+			June 2006
+			Elsevier BV
+		
+	
+	Williams, V. and Noland, R. B., "Comparing the CO2 emissions and contrail formation from short and long haul air traffic routes from London Heathrow," Environmental Science & Policy, Vol. 9, No. 5, June 2006, pp. 487-495.
+
+
+				
+			
+		
+	

file138.txt

@@ -0,0 +1,755 @@
+
+	
+	
+		
+I. IntroductionAircraft induced environmental impact has drawn attention in recent years. 1 The three largest emission impacts include direct emissions of greenhouse gases such as CO 2 , emissions of NOx, and persistent contrails.Contrails are clouds that are visible trails of water vapor made by the exhaust of aircraft engines.Contrails form when a mixture of warm engine exhaust gases and cold ambient air reaches saturation with respect to water, forming liquid drops which quickly freeze.They persist if the aircraft is flying in certain atmospheric conditions.Persistent contrails reduce incoming solar radiation and outgoing thermal radiation in a way that accumulates heat. 2 The global mean contrail cover in 1992 was estimated to double by 2015, and quadruple by 2050 due to an increase in air traffic. 3Studies suggest that the environmental impact from persistent contrail is estimated to be three to four times, 4 or even ten times 5 larger than the aviation induced emissions.Therefore, methods to reduce aircraft induced persistent contrails are needed to minimize the impact of aviation on climate.Efforts have been made in the past to reduce the persistent contrail formation.Gierens 6 and Noppel 7 reviewed various strategies for contrail avoidance.Mannstein 8 proposed a strategy to reduce the climate impact of contrails significantly by only small changes in individual flight altitude.Campbell 9 presented a methodology to optimally reroute aircraft trajectories to avoid the formation of persistent contrails with the use of mixed integer programming.Both methodologies require onboard contrail detection system and flight rerouting.Fichter 10 showed that the global annual mean contrail coverage could be reduced by downshifting the cruise altitude.Williams 11,12 proposed strategies for contrail reduction by identifying fixed and varying maximum altitude restriction policy.These restrictions generally imply more fuel burn, thus more emissions, and add congestion to the already crowded airspace at lower altitudes.The objective of this paper is to develop strategies to reduce persistent contrail formation with consideration to extra emissions and air space congestion.The concept of contrail frequency index is used to quantify the severity of contrail formation.The strategy for reducing persistent contrail formation is to reduce contrail frequency index by altering the aircraft's cruising altitude with minimal increase in emissions.A class of contrail reduction strategies that considers extra emissions is proposed.It provides a flexible way to trade off between contrail reduction and emissions.The results show that the contrail frequency index can be reduced with extra emissions and without adding congestion to airspace.The strategies provide a starting point for developing operational policies to reduce the impact of aviation on climate.The remainder of the paper is organized as follows.Section II provides descriptions of contrail model, definition of contrail frequency index, and the fuel burn and emission models.Next, contrail reduction strategies and the trade-offs between contrail reduction and emissions are described in Section III.Section IV shows the results.Finally, Section V presents a summary and conclusions.
+II. Data and Model
+II.A. Contrail ModelContrails are vapor trails caused by aircraft operating at high altitudes under certain atmospheric conditions.The contrail model in this paper uses atmospheric temperature and humidity data retrieved from the Rapid Updated Cycle (RUC) data, provided by the National Oceanic and Atmospheric Administration (NOAA).The horizontal resolution in RUC is 13-km.RUC data has 37 vertical isobaric pressure levels ranging between 100 and 1000 millibar (mb) in 25 mb increments.Since the vertical isobaric pressure levels do not correspond with 2,000 feet increments, linear interpolation was used to convert the RUC data to a vertical range from 26,000 feet to 44,000 feet with an increment of 2,000 feet.This range is chosen because it generally is too warm for contrails to form below 26,000 feet and most aircraft fly below 44,000 feet.Contrails form when a mixture of warm engine exhaust gases and cold ambient air reaches saturation with respect to water, forming liquid drops which quickly freeze.Contrails form in the regions of airspace that have ambient Relative Humidity with respect to Water (RHw) greater than a critical value r contr . 13Regions with RHw greater than or equal to 100% are excluded because clouds are already present. 14Contrails can persist when the environmental Relative Humidity with respect to Ice (RHi) is greater than 100%. 15In this paper, contrail favorable regions are defined as the regions of airspace that have r contr ≤ RHw < 100% and RHi ≥ 100%.The estimated critical relative humidity for contrail formation at a given temperature T (in Celsius) can be calculated asr contr = G(T -T contr ) + e liq sat (T contr ) e liq sat (T ) ,(1)where e liq sat (T ) is the saturation vapor pressure over water at a given temperature.The estimated threshold temperature for contrail formation at liquid saturation isT contr = -46.46 + 9.43ln(G -0.053) + 0.72ln 2 (G -0.053),(2)whereG = EI H2O C p P Q(1 -η) , (3)EI H2O is the emission index of water vapor (assumed to be 1.25); C p = 1004 (in JKg -1 K -1 ) is the isobaric heat capacity of air, P (in Pa) is the ambient air pressure, = 0.6222 is the ratio of molecular masses of water and dry air, Q = 43 • 10 6 (in JKg -1 ) is the specific combustion heat, and η = 0.3 is the average propulsion efficiency of the jet engine.The value of r contr is computed by Eq (1)-(3) using RUC measurements for RHw and temperatures.RHi is calculated by temperature and relative humidity using the following formula: 16 RHi = RHw • 6.0612e 18.102T /(249.52+T ) 6.1162e 22.577T /(237.78+T ) ,where T is the temperature in Celsius. Figure 1 shows the temperature, RHw, RHi, and contrail favorable regions at 8AM EDT on April 23, 2010 at an altitude of 34,000 feet.
+II.B. Contrail Frequency IndexContrail frequency index (CFI) is used to quantify the severity of contrail activities.This paper uses 13km RUC data instead of the 40km RUC data used in Ref. 17.The modified 13km RUC data divide the U.S. national airspace into a three dimensional grid with 337 elements along the latitude, 451 elements along the longitude, and 10 altitudes ranging from 26,000 feet to 44,000 feet.Air traffic in the U.S. can be mapped into the same volumetric grid.Contrail frequency index is the number of aircraft in a volumetric element which meets conditions for persistent contrail formation.Contrail frequency index is zero for volumetric elements which do not meet the conditions for persistent contrail formation.Precise definitions of contrail frequency index are provided by the following equations.The altitude level index l is defined as l = 1 . . . 10 corresponding to altitudes of 26, 000, 28, 000, . . ., 44, 000 feet.The persistent contrail formation matrix (contrail matrix) at time t at level l is defined asR l t =     r l 1,1,t r l 1,    ,(5)where r l i,j,t is 1 if r contr ≤ RHw < 100% and RHi ≥ 100% at grid (i, j), and 0 if the conditions are not met.The Center contrail frequency indices of twenty U.S. air traffic control centers at time t at level l are defined asC center,l,t = 337 i=1 451 j=1 r l i,j,t a l i,j,t c i,j ,(6)where a l i,j,t is the number of aircraft within RUC 13km grid (i, j) flying closest to altitude level l at time t, and c i,j is 1 when grid (i, j) is inside the center and 0 if not.The twenty U.S. air traffic control centers are listed in Table 1.The aircraft data were provided by the Federal Aviation Administration's Aircraft Situation Display to Industry (ASDI) data.For planning contrail reduction strategies, traffic flow managers need to know potentially high contrail regions in the next few hours.Therefore predicted contrail frequency indices are needed for contrail reduction strategies.Similar to the concept of Weather Impacted Traffic Index (WITI) introduced by Callaham et al. 18 and Sridhar, 19 and the three-dimensional index derived by Chen, 20 predicted contrail frequency index was defined as a convolution of predicted traffic data and forecast of atmospheric conditions.The index consists of the RUC forecast data and the predicted aircraft locations when t is a future time.The Center contrail frequency index can then be rewritten asC center,l,t =    337 i=1 451 j=1 r l i,j,t a l i,j,t c i,j if t <= t now , 337i=1451 j=1 rl i,j,t âl i,j,t c i,j if t > t now ,(7)where t now is the current time, rl i,j is defined in (5) with RUC forecast data, and âl i,j is the predicted number of aircraft within RUC 13km grid (i, j) flying closest to altitude level l at time t.Figure 2 illustrates how contrail frequency index is computed.The aircraft trajectories and contrail formations between 33,000 feet and 34,999 feet for the hour of 8AM EDT on April 23, 2010 are shown in Fig. 2a.An one-minute time interval is used.The blue polygons indicate the contrail favorable regions; grey dots are the aircraft between 33,000 feet and 34,999 feet.When the aircraft enter the blue polygons, contrails would form as indicated by blue dots.The number of blue dots is defined as the contrail frequency index.As shown in Fig. 2b, there are 148 blue dots for the hour in Kansas City Center.Therefore, the center contrail frequency index for Kansas City Center for the hour of 8AM EDT is 148.The total time, due to all aircraft that would form contrails during the hour, is 148 minutes.The Center contrail frequency indices for all 20 US air traffic control Centers at 34,000 feet at 8AM EDT on April 23, 2010 were computed and are shown in Fig. 3.As shown in the figure, Minneapolis Center (ZMP) and Chicago Center (ZAU) have high contrail frequency indices because there are large contrail favorable regions in the Centers and also high density of air traffic, as shown in Fig. 2a.Salt Lake City Center (ZLC) has large contrail favorable regions inside the Center but the contrail frequency index is low because not many aircraft fly through the Center.Contrail frequency index takes both atmospheric and air traffic data and quantifies the contrail activities.It will be used later in developing contrail reduction strategies.
+II.C. Fuel Burn and Emission ModelsThe computations of aircraft fuel burn and emissions are needed in order to study the trade-offs between contrail reductions and aircraft induced emissions.This paper uses the fuel consumption model in Eurocontrols Base of Aircraft Data Revision 3.7 (BADA). 21The air traffic data provide aircraft information including aircraft type, mass, altitude and speed to compute the fuel burn.There are five stages, climb, cruise, descent-idle, descent-approach, and descent-landing that are determined by the aircraft altitude and  speed.Only climb, cruise, and descent-idle models are used in this paper since the other two are used at the low altitudes.For climb stage, the fuel burn is computed using the following equation,F B = SF C • T • ∆t, (8)where F B is the fuel burn in kilograms, SF C (kg/min•kN) is the thrust specific fuel consumption, T is the trust in Newtons, and ∆t is the elapse time in minutes.For cruise, the fuel burn isF B = SF C • T • C f cr • ∆t,(9)where C f cr is the cruise fuel flow factor.For descent-idle, the fuel burn isF B = C f 3 (1 - h C f 4 ),(10)where C f 3 and C f 4 are descent fuel flow coefficients, and h is the altitude in meters.SFC in ( 8) and ( 9) are formulated asJet: SF C = C f 1 (1 + V T AS C f 2 ), Turboprop: SF C = C f 1 (1 - V T AS C f 2 ) • (V T AS /1000),(11)where V T AS is the true air speed in meters per second, and C f 1 and C f 2 are thrust specific fuel consumption coefficients.The thrust in (8) for climb stage is formulated asJet: T climb = C T c,1 (1 - h C T c,2 + C T c,3 • h 2 ), Turboprop: T climb = C T c,1 (1 - h C T c,2 )/V T AS + C T c,3 ,(12)where C T c,1 , C T c,2 and C T c,3 are climb thrust coefficients.For cruise, thrust is set equal to drag.Drag is computed byD = C D • ρ • V 2 T AS • S 2 , (13)where D is the drag in Newtons, C D is the drag coefficient, ρ (kg/m 3 ) is the air density, and S (m 2 ) is the wing reference area.The emission model is based on a prototype of the Aviation Environmental Design Tool (AEDT) developed by the Federal Aviation Administration (FAA). 22Five emissions are computed including CO 2 , SO x , CO, HC and NO x .Emissions of CO 2 and SO x (modeled as SO 2 ) are modeled based on fuel consumption. 23he emissions are computed byE CO2 = 3155 • F B, E SO2 = 0.8 • F B,(14)where E CO2 and E SO2 are emissions of CO 2 and SO 2 in grams, and FB is fuel burn in kilograms.Emissions of CO, HC and NO x are modeled through the use of the Boeing Fuel Flow Method 2 (BFFM2). 24The emissions are determined by aircraft engine type, altitude, speed, and fuel burn and the coefficients in International Civil Aviation Organization (ICAO) emission data bank.In the models, fuel burn is corrected to sea-level reference temperature (273.15K) and pressure (14.696 psi): 3.8 amb exp(0.2MF B c = (F B/δ amb )[θwhere F B c is the corrected fuel flow, P amb is the at-altitude ambient pressure, T amb is the at-altitude ambient temperature, and M is the Mach number.where EICO, EIHC and EIN O x are emission indices of CO, HC and N O x , H is the humidity correction factor, and ω is the specific humidity.The emissions are computed byE CO = EICO • F B, E HC = EIHC • F B, E N Ox = EIN O x • F B,(17)where E CO , E HC and E N Ox are emissions in grams.
+III. Contrail Reduction Strategies
+III.A. Use of contrail frequency indexContrail frequency index (CFI) quantifies the contrail activities.The strategy for reducing the persistent contrail formations is to minimize contrail frequency index by altering the aircraft's cruising altitude.Assume the aircraft at altitude level l in a Center are made to fly at a different level l .Both l and l range from 1 to 10, corresponding to altitudes of 26, 000, 28, 000, . . ., 44, 000 feet.The definition of the contrail frequency index is changed from (6) toC l center,l,t = 337 i=1 451 j=1 r l i,j,t a l i,j,t c i,j ,(18)A contrail frequency index matrix is formed asC center,t =       C 1 1,where the diagonal term C l l,t is the contrail frequency index at level l before changing cruising altitude, and C l l,t is the contrail frequency index when guiding aircraft at level l to level l .The contrail reduction from level l to l is ∆C l l,t = C l l,t -C l l,t .Note that when l > l, not all aircraft have the ability to fly from level l to level l .If altitude level l is higher than an aircraft's maximal flight altitude, it stays at level l and is not counted in C l l,t .In addition, if an aircraft crosses a sector boundary and causes congestion, it stays at level l and does not add to C l l,t .Additional conditions can be added to satisfy other operational procedures.The strategy is to find the altitude that would form least contrails.In other words, find the smallest element in each column of C center,t .If the aircraft are limited to alter ∆l levels, the solution is the smallest elementin [C l-∆l l,t . . . C l l,t . . . C l+∆l l,t] T in each column.The solution is denoted as [ l1 . . .l10 ].Each li means aircraft at flight level i is flying at level li .If li = i, the aircraft at level i do not alter.The total contrail reduction at the given center at time t can be expressed asΣ∆C t = 10 i=1 ∆C li i,t .(21)Consider the traffic situation at Kansas City Center.For ∆l = 2, the CFI matrix at 8AM EDT on April 23, 2010 was computed,C ZKC =                    0 0 0 × × × × × × × 0 0 0 0 × × × × × × 0 0 0 0 0 × × × × × × 0 0 0 0 0 × × × × × × 61 89 148 387 233 × × × × × × 35 102 230 154 83 × × × × × × 104 213 141 65 0 × × × × × × 164 67 22 0 0 × × × × × × 137 17 0 0 × × × × × × × 18 0 0                    , (22)where the elements not used are marked as ×.The center is divided into sectors horizontally and vertically.An air traffic controller monitors traffic in each sector and maintains separation between aircraft.The number of aircraft in a sector is kept below a maximum, referred to as Monitor Alert Parameter (MAP) in the current U.S. air traffic system, to keep the controllers workload within limits. 25The MAP is used to define the airspace capacity.The contrail reduction moves will not change the sector counts unless they cross the sector boundaries.The strategies only allow the moves such that the aircraft count in a sector does not exceed the sector capacity after the moves.In the previous example, Kansas City Center has 15 high sectors and 11 super-high sectors.Among them, sector 31 has the highest sector count during the hour.Sector 31 has a lower bound of 37,000 feet and is on top of sector 28, 29 and 30, shown in Fig. 4. The move from level 6 (35,000-36,999 feet) to level 7 (37,000-38,999 feet) would move some aircraft in sector 28, 29 and 30 to sector 31.Sector 28, 29, 30 and 31 have the MAP values of 18, 18, 19 and 21 respectively.Figure 5 shows the MAP values and the sector counts in sector 28, 29, 30 and 31 before and after the moves.The aircraft counts in sector 28, 29 and 30 decrease because some aircraft have been moved up to sector 31; the sector count in sector 31 increases but is still lower than the sector capacity of 21.Thus the contrail reduction moves are applied without exceeding the capacity of the airspace.The altitudes of the aircraft are changed as they enter a new Center.The number of altitude changes is not expected to result in frequent climb and descents to affect current operations.However, if needed, additional constraints can be imposed on the number of altitude changes.Data from a 24-hour period on April 23, 2010 was analyzed.The contrail reduction strategies were applied and the results are shown in Fig. 6.The center CFIs before reduction are shown in blue bars.When the aircraft altitudes are allowed to alter by 2,000 feet, the center CFIs after reduction are shown in light blue bars.The total reduction among all centers is 62%.When the aircraft altitudes are allowed to alter by 4,000 feet, the total reduction is 88% as indicated in green bars.Since allowing aircraft to alter 4,000 feet would eliminate most of the contrail formation, the strategies in this paper limit the altitude changes to 4,000 feet.
+III.B. Tradeoff between contrails and emissionsAltering cruising altitudes changes the aircraft fuel consumption and emissions.In order to analyze the environmental impact of contrail reduction strategies, fuel consumption and emissions are considered in the strategies.Fuel burn and emissions computations are based on the models described in Sec.II.C. Define E l l,t as the emissions for all aircraft at level l at a given center at time t before contrail reduction, and E l l,t as the total emissions when guiding aircraft from level l to level l .When aircraft change their flying altitude from level l to l , the difference in emissions is∆E l l,t = E l l,t -E l l,t .(23)∆E l l,t < 0 implies emission reduction.Define the emission matrix as∆E center,t =         0 ∆E        .0This matrix helps to study the emissions trade-offs when applying contrail reduction strategies.For the contrail reduction solution of [ l1 . . .l10 ], the change in emissions can be expressed asΣ∆E t = 10 i=1 ∆E li i,t .(25)Consider the same example in the previous subsection and study the trade-offs between contrail reduction and CO 2 emissions.The emission matrix for CO 2 was computed based on the models described in Sec.II.C and is the following:∆E ZKC =                    0 484 1130 × × × × × × × -27 0 531 3562 × × × × × × -41 -31 0 1674 3169 × × × × × × -28 11 0 1417 4542 × × × × × × 55 237 0 2143 3462 × × × × × × 285 1331 0 1683 1042 × × × × × × 961 1237 0 420 2 × × × × × × 434 1892 0 0 0 × × × × × × 70 106 0 0 × × × × × × × 128 0 0      ��              ,(26)where the elements not used are marked as × and the unit is kilograms.Assuming the environmental impact of the contrail frequency index of 1 is equivalent to CO 2 emissions of 10 kg, the move from level 5 to 4 makes sense because a reduction of 148 in CFI is greater than the impact of additional CO 2 of 1417 kg (148 • 10 -1417 > 0).However, the move from level 6 to 4 is not worth while because the net impact is negative (230 • 10 -4542 < 0).Instead, the move from 6 to 8 is preferred because it has a CFI reduction of 66 with additional CO 2 emissions of 434 kg and reduces the net impact (66 • 10 -434 > 0).Similarly, the move from level 7 to 8 and from 8 to 9 are not preferred because of the net negative impacts.Aircraft at level 7 and 8 are not altered.The new solution can be denoted as [1 2 3 4 4 8 7 8 9 10], resulting in a CFI reduction of 214, with additional CO 2 emissions of 1851 kg.Compared with the maximal reduction strategy, this strategy achieves less contrail reduction, 40% versus 84%, but emits much less CO 2 emissions, 1, 851 kg vs 7, 957 kg (77% less).This example shows that the proposed contrail reduction strategies have the capability to trade off contrail reduction with emissions.Considering the relative environment impact of emissions and contrails, the strategy would move aircraft only if the contrails reduction benefits exceed the environmental impact of additional emissions.The aircraft would be guided from level l to l only if∆C l l,t > 1 α ∆E l l,t ,(27)where ∆C i,t and ∆E l,t are defined in (20) and (23) and α is a user-defined trade-off factor.It can be interpreted as the equivalent emissions in kg that has the same environmental impact as the contrail frequency index of 1.For the maximal contrail reduction strategy, the effect of emissions is ignored.In other words, α = ∞.Also, α = 0 simply means no reduction strategy is applied because (27) will never be true.Higher values of α means more contrail reduction and more emissions (closer to maximal reduction strategy); lower α means less contrail reduction and less emissions (closer to no reduction).In the previous example, α is 10.The appropriate value of α can be determined in two different ways.It is possible to monetize the value of both contrails and emissions as suggested in Ref. 26.Another approach is to consider contrails and emissions as disturbances to the global climate equilibrium and measure their impact as changes to the global mean surface temperature. 27However, both these methods are beyond the scope of this paper and the value of α will be considered as a user-preference weighting factor in the rest of the paper.
+IV. ResultsThis section presents the results of contrail reduction strategies and the trade-offs between contrail reduction and extra emissions over a 24-hour period on April 23, 2010.The 24-hour period starts at 4AM EDT and ends at 4AM the next day.The strategies allow aircraft to move 4,000 feet up or down within a center and use various user-defined α values to trade off between contrail reduction and emissions.This paper focuses on the trade-offs between contrails and CO 2 emissions while other emissions like NOx, SO 2 , HC and CO have a similar trend.Figure 7 shows the hourly variations in contrail reduction and extra emissions with different trade-off factors during a 24-hour period over the entire U.S. In Fig. 7a, the blue line shows the hourly CFI during the day with no reduction strategy applied (α = 0).When reduction strategies are applied, it is consistent that higher α results in lower CFI, meaning more reduction.The maximal reduction strategy (α = ∞), shown in the magenta line, has the lowest CFI at every hour.On the other hand, Fig. 7b shows that higher α results in higher extra CO 2 emissions, and the maximal reduction strategy has the highest CO 2 emissions.The results show that contrails reduction results in extra CO 2 emissions.Looking at the Center level, Fig. 8 shows the daily contrail reduction and extra emissions in twenty U.S. air traffic control centers.The blue bars in Fig. 8a are the daily center contrail frequency index for each Center.It is consistent that for all Centers, higher α values results in more contrail reduction and the maximal reduction strategy achieves most contrail reduction in all twenty Centers.On the other hand, higher α also results in more CO 2 emissions, as shown in Fig. 8b, while the maximal reduction strategy has the most CO 2 emissions.Table 2 summarizes the trade-offs between contrail reduction and extra CO 2 emissions over the entire U.S. on April 23, 2010.On that day, the maximal reduction strategy has an 88% contrail reduction rate with extra CO 2 emissions of 3778 megagram (Mg).A smaller value of α lowers the contrail reduction ratio but has less emissions.For α = 40, the contrail reduction rate is 73% with 2,621 Mg extra CO 2 emissions, 31% less than the emissions in the maximal reduction strategy.If CO 2 has more environmental impact, using α = 10 results in a contrail reduction of 21% with 100 Mg extra CO 2 emissions, 97% less than the emissions in the maximal reduction strategy.As for fuel burn, considering all aircraft flying between 26, 000 feet and 44, 000 feet on a day with large contrail favorable regions, an 80% reduction in contrails can be achieved with around 1% extra fuel.The increase in fuel would be less on a day with smaller contrail favorable regions.The main focus of this paper is to study the trade-offs between contrail reduction and extra emissions.Therefore, the factor of the extra fuel burn is not taken into account in the strategies.Figure 9 shows the contrail reduction versus extra CO 2 emissions with various α values.In the figure, more contrail reduction takes place from left to right and more CO 2 emissions occurs from bottom to top.At the lower-left point, no reduction strategy is applied.The upper-right point is the maximal reduction strategy.As the values of α increases, the curve moves from lower-left to upper-right.The user-defined trade-off factor α provides a flexible way to trade off between contrail reduction and extra emissions.Better understanding of the trade-offs between contrails and emissions and impact on the climate need to beRelative Humidity with respect to ice (d) Contrail favorable regions
+Figure 1 .1Figure 1.Atmospheric data and contrail favorable regions at 34,000 feet at 8AM EDT on April 23, 2010.
+entire U.S. airspace (b) Kansas City Center
+Figure 2 .Figure 3 .23Figure 2. Aircraft trajectories and contrail favorable regions at 8AM EDT on April 23, 2010.
+F B c is used in ICAO emission data bank to determine the reference emission index REIHC, REICO and REIN O x for HC, CO and NO x .The emission indices are computed by EICO = REICO(θ 3.3 amb /δ 1.02 amb ), EIHC = REIHC(θ 3.3 amb /δ 1.02 amb ), EIN O x = REIN O x [exp(H)](δ 1.02 amb /θ 3.3 amb ) 0.5 , H = -19.0(ω-0.0063),
+Figure 4 .4Figure 4. Kansas City Center sector 28, 29, 30 and 31.
+Figure 5 .5Figure 5. MAP values and sector counts before and after the contrail reduction strategies at 8AM EDT on April 23, 2010.
+Figure 6 .6Figure 6.Results of contrail reduction strategies on April 23, 2010.
+Figure 7 .7Figure 7. Hourly contrail reduction and extra CO2 emissions using different trade-off factors on April 23, 2010.
+Figure 9 .9Figure 9. Contrail reduction versus extra CO2 emissions on April 23, 2010.
+Table 1 .1Center index of twenty continental U.S. air traffic control centers.IndexNameIndexName1Seattle Center (ZSE)11Chicago Center (ZAU)2Oakland Center (ZOA)12Indianapolis Center (ZID)3Los Angeles Center (ZLA)13Memphis Center (ZME)4Salt Lake City Center (ZLC)14Cleveland Center (ZOB)5Denver Center (ZDV)15Washington D. C. Center (ZDC)6Albuquerque Center (ZAB)16Atlanta Center (ZTL)7Minneapolis Center (ZMP)17Jacksonville Center (ZJX)8Kansas City Center (ZKC)18Miami Center (ZMA)9Dallas/Fort Worth Center (ZFW)19Boston Center (ZBW)10Houston Center (ZHU)20New York Center (ZNY)
+2 )], δ amb = P amb /14.696, θ amb = (T amb + 273.15)/273.15,
+The diagonal elements of the matrix show the current CFIs at various altitudes.First consider the case if the aircraft are allowed to move one level (2,000 feet) up or down to reduce contrail formation.All the aircraft between 33,000 feet and 34,999 feet (level 5) have a totalCFI of 148 (C ZKC (5, 5) = 148). Moving the aircraft to level 4 will result in zero CFI (C ZKC (4, 5) = 0), areduction of CFI by 148. Other contrail reduction moves include moving aircraft from level 6 to 7 (a CFIreduction of 17), 7 to 8 (a CFI reduction of 74) and 8 to 9 (a CFI reduction of 5). The solution is expressedas [1 2 3 4 4 7 8 9 9 10], resulting in a CFI reduction from 541 to 297, a 45% reduction. If the aircraft areallowed to move two levels up or down, even greater reductions can be achieved. The moves include movingaircraft from level 5 to 4, 6 to 4, 7 to 8 and 8 to 9. The solution is expressed as [1 2 3 4 4 4 8 9 9 10],resulting in a contrail reduction from 541 to 84, an 84% reduction.
+4s shown in the matrix and in(22), moving aircraft from level 5 to 4 results in a CFI reduction of 148 with additional CO 2 emissions of 1,417 kg (∆E45,t = 1417); moving from level 6 to 4 results in a CFI reduction of 230 with additional CO 2 of 4,542 kg; moving from level 7 to 8 results in a CFI reduction of 74 with additional CO 2 of 1892 kg; moving from level 8 to 9 results in a CFI reduction of 5 with additional CO 2 of 106 kg.This solution achieves the most contrail frequency index reduction of 457 with additional CO 2 emissions of 7,957 kg.
+		
+		
+			
+developed to fully utilize this class of contrail reduction strategies.
+V. ConclusionsA class of strategies for reducing persistent contrail formations with the capability to trade off between contrails and emissions has been developed.The concept of contrail frequency index is defined and used to quantify the contrail activities.The strategy of reducing the persistent contrail formations is to minimize the contrail frequency index by altering the aircraft's cruising altitude with consideration to extra emissions.The strategies use a user-defined factor to trade off between contrail reduction and extra emissions.The analysis results show that the contrails can be reduced with extra emissions and without adding congestion to airspace.For the day tested, the results show that the maximal contrail reduction strategy can achieve a contrail reduction of 88%.When a trade-off factor is used, the strategy can still achieve a 73% contrail reduction while emitting 31% less emissions compared to the maximal contrail reduction strategy, or achieve a 21% contrail reduction while only emitting 97% less emissions.The user-defined trade-off factor provides a flexible way to trade off between contrail reduction and extra emissions.Better understanding of the trade-offs between contrails and emissions and impact on the climate need to be developed to fully utilize this class of contrail reduction strategies.The strategies provide a starting point for developing operational policies to reduce the impact of aviation on climate.			
+			
+
+				
+
+
+	
+		
+			IWaitz
+		
+		
+			JTownsend
+		
+		
+			JCutcher-Gershenfeld
+		
+		
+			EGreitzer
+		
+		
+			JKerrebrock
+		
+		Report to the United States Congress: Aviation and the Environment, A National Vision, Framework for Goals and Recommended Actions
+		London, UK
+		
+			December 2004
+		
+	
+	Tech. rep
+	Partnership for AiR Transportation Noise and Emissions Reduction
+	Waitz, I., Townsend, J., Cutcher-Gershenfeld, J., Greitzer, E., and Kerrebrock, J., "Report to the United States Congress: Aviation and the Environment, A National Vision, Framework for Goals and Recommended Actions," Tech. rep., Partnership for AiR Transportation Noise and Emissions Reduction, London, UK, December 2004.
+
+
+
+	
+		Radiative forcing by contrails
+		
+			RMeerkötter
+		
+		
+			USchumann
+		
+		
+			DRDoelling
+		
+		
+			PMinnis
+		
+		
+			TNakajima
+		
+		
+			YTsushima
+		
+		10.1007/s00585-999-1080-7
+	
+	
+		Annales Geophysicae
+		Ann. Geophys.
+		1432-0576
+		
+			17
+			8
+			
+			1999
+			Copernicus GmbH
+		
+	
+	Meerkotter, R., Schumann, U., Doelling, D. R., Minnis, P., Nakajima, T., and Tsushima, Y., "Radiative forcing by contrails," Annales Geophysicae, Vol. 17, 1999, pp. 1080-1094.
+
+
+
+	
+		Future Development of Contrail Cover, Optical Depth, and Radiative Forcing: Impacts of Increasing Air Traffic and Climate Change
+		
+			SMarquart
+		
+		
+			MPonater
+		
+		
+			FMager
+		
+		
+			RSausen
+		
+		10.1175/1520-0442(2003)016<2890:fdocco>2.0.co;2
+	
+	
+		Journal of Climate
+		J. Climate
+		0894-8755
+		1520-0442
+		
+			16
+			17
+			
+			September 2003
+			American Meteorological Society
+		
+	
+	Marquart, S., Ponater, M., Mager, F., and Sausen, R., "Future Development of Contrail Cover, Optical Depth, and Radiative Forcing: Impacts of Increasing Air Traffic and Climate Change," Journal of Climate, Vol. 16, September 2003, pp. 2890-2904.
+
+
+
+	
+		A Standing Royal Commission
+		
+			SusanOwens
+		
+		10.1093/acprof:oso/9780198294658.003.0003
+	
+	
+		Knowledge, Policy, and Expertise
+		London, UK
+		
+			Oxford University Press
+			2002
+			
+		
+	
+	"The Environmental Effects of Civil Aircraft in Flight," Tech. rep., Royal Commission on Environmental Pollution, London, UK, 2002.
+
+
+
+	
+		Aircraft induced contrail cirrus over Europe
+		
+			HermannMannstein
+		
+		
+			UlrichSchumann
+		
+		10.1127/0941-2948/2005/0058
+	
+	
+		Meteorologische Zeitschrift
+		metz
+		0941-2948
+		
+			14
+			4
+			
+			2005
+			Schweizerbart
+		
+	
+	Mannstein, H. and Schumann, U., "Aircraft induced contrail cirrus over Europe," Meteorologische Zeitschrift, Vol. 14, No. 4, 2005, pp. 549-554.
+
+
+
+	
+		A Review of Various Strategies for Contrail Avoidance
+		
+			KlausGierens
+		
+		
+			LingLim
+		
+		
+			KostasEleftheratos
+		
+		10.2174/1874282300802010001
+	
+	
+		The Open Atmospheric Science Journal
+		TOASCJ
+		1874-2823
+		
+			2
+			1
+			
+			2008
+			Bentham Science Publishers Ltd.
+		
+	
+	The Open Atmospheric
+	Gierens, K., Limb, L., and Eleftheratos, K., "A Review of Various Strategies for Contrail Avoidance," The Open Atmo- spheric Science Journal, Vol. 2, 2008, pp. 1-7.
+
+
+
+	
+		Overview on Contrail and Cirrus Cloud Avoidance Technology
+		
+			FNoppel
+		
+		
+			RSingh
+		
+		10.2514/1.28655
+	
+	
+		Journal of Aircraft
+		Journal of Aircraft
+		0021-8669
+		1533-3868
+		
+			44
+			5
+			
+			2007
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	Noppel., F. and Singh, R., "Overview on Contrail and Cirrus Cloud Avoidance Technology," Journal of Aircraft, Vol. 44, No. 5, 2007, pp. 1721-1726.
+
+
+
+	
+		A note on how to avoid contrail cirrus
+		
+			HermannMannstein
+		
+		
+			PeterSpichtinger
+		
+		
+			KlausGierens
+		
+		10.1016/j.trd.2005.04.012
+	
+	
+		Transportation Research Part D: Transport and Environment
+		Transportation Research Part D: Transport and Environment
+		1361-9209
+		
+			10
+			5
+			
+			September 2005
+			Elsevier BV
+		
+	
+	Mannstein, H., Spichtinger, P., and Gierens, K., "A note on how to avoid contrail cirrus," Transportation Research. Part D, Transport and environment, Vol. 10, No. 5, September 2005, pp. 421-426.
+
+
+
+	
+		An Optimal Strategy for Persistent Contrail Avoidance
+		
+			ScotCampbell
+		
+		
+			NatashaNeogi
+		
+		
+			MichaelBragg
+		
+		10.2514/6.2008-6515
+		AIAA-2008-6515
+	
+	
+		AIAA Guidance, Navigation and Control Conference and Exhibit
+		Honolulu, HI
+		
+			American Institute of Aeronautics and Astronautics
+			August 2008
+		
+	
+	Campbell1, S. E., Neogi, N. A., and Bragg, M. B., "An Optimal Strategy for Persistent Contrail Avoidance," AIAA Guidance, Navigation and Control Conference, AIAA-2008-6515, AIAA, Honolulu, HI, August 2008.
+
+
+
+	
+		The impact of cruise altitude on contrails and related radiative forcing
+		
+			ChristineFichter
+		
+		
+			SusanneMarquart
+		
+		
+			RobertSausen
+		
+		
+			DavidSLee
+		
+		10.1127/0941-2948/2005/0048
+	
+	
+		Meteorologische Zeitschrift
+		metz
+		0941-2948
+		
+			14
+			4
+			
+			August 2005
+			Schweizerbart
+		
+	
+	Fichter, C., Marquart, S., Sausen, R., and Lee, D. S., "The impact of cruise altitude on contrails and related radiative forcing," Meteorologische Zeitschrift, Vol. 14, No. 4, August 2005, pp. 563-572.
+
+
+
+	
+		Reducing the climate change impacts of aviation by restricting cruise altitudes
+		
+			VictoriaWilliams
+		
+		
+			RobertBNoland
+		
+		
+			RalfToumi
+		
+		10.1016/s1361-9209(02)00013-5
+	
+	
+		Transportation Research Part D: Transport and Environment
+		Transportation Research Part D: Transport and Environment
+		1361-9209
+		
+			7
+			6
+			
+			November 2002
+			Elsevier BV
+		
+	
+	Williams, V., Noland, R. B., and Toumi, R., "Reducing the climate change impacts of aviation by restricting cruise altitudes," Transportation Research. Part D, Transport and environment, Vol. 7, No. 5, November 2002, pp. 451-464.
+
+
+
+	
+		Variability of contrail formation conditions and the implications for policies to reduce the climate impacts of aviation
+		
+			VictoriaWilliams
+		
+		
+			RobertBNoland
+		
+		10.1016/j.trd.2005.04.003
+	
+	
+		Transportation Research Part D: Transport and Environment
+		Transportation Research Part D: Transport and Environment
+		1361-9209
+		
+			10
+			4
+			
+			July 2005
+			Elsevier BV
+		
+	
+	Williams, V. and Noland, R. B., "Variability of contrail formation conditions and the implications for policies to reduce the climate impacts of aviation," Transportation Research. Part D, Transport and environment, Vol. 10, No. 4, July 2005, pp. 269-280.
+
+
+
+	
+		Contrails in a comprehensive global climate model: Parameterization and radiative forcing results
+		
+			MichaelPonater
+		
+		
+			SMarquart
+		
+		
+			RSausen
+		
+		10.1029/2001jd000429
+	
+	
+		Journal of Geophysical Research
+		J. Geophys. Res.
+		0148-0227
+		
+			107
+			D13
+			
+			2002
+			American Geophysical Union (AGU)
+		
+	
+	Ponater, M., Marquart, S., and Sausen, R., "Contrails in a Comprehensive Global Climate Model: Parameterization and Radiative Forcing Results," Journal of Geophysical Research, Vol. 107, No. D13, 2002, pp. ACL 2-1.
+
+
+
+	
+		Determination of humidity and temperature fluctuations based on MOZAIC data and parametrisation of persistent contrail coverage for general circulation models
+		
+			KMGierens
+		
+		
+			USchumann
+		
+		
+			HG JSmit
+		
+		
+			MHelten
+		
+		
+			GZängl
+		
+		10.1007/s00585-997-1057-3
+	
+	
+		Annales Geophysicae
+		Ann. Geophys.
+		1432-0576
+		
+			15
+			8
+			
+			1997
+			Copernicus GmbH
+		
+	
+	Gierens, K. M., Schumann, U., Smit, H. G. J., Helten, M., and Zangl1, G., "Determination of humidity and temperature fluctuations based on MOZAIC data and parametrisation of persistent contrail coverage for general circulation models," Annales Geophysicae, Vol. 15, 1997, pp. 1057-1066.
+
+
+
+	
+		Estimated contrail frequency and coverage over the contiguous United States from numerical weather prediction analyses and flight track data
+		
+			DavidPDuda
+		
+		
+			PatrickMinnis
+		
+		
+			RabindraPalikonda
+		
+		10.1127/0941-2948/2005/0050
+	
+	
+		Meteorologische Zeitschrift
+		metz
+		0941-2948
+		
+			14
+			4
+			
+			June-July 2003
+			Schweizerbart
+			Friedrichshafen at Lake Constance, Germany
+		
+	
+	Duda, D. P., Minnis, P., Costulis, P. K., and Palikonda, R., "CONUS Contrail Frequency Estimated from RUC and Flight Track Data," European Conference on Aviation, Atmosphere, and Climate, Friedrichshafen at Lake Constance, Germany, June- July 2003.
+
+
+
+	
+		Improved Magnus Form Approximation of Saturation Vapor Pressure
+		
+			OlegAAlduchov
+		
+		
+			RobertEEskridge
+		
+		10.1175/1520-0450(1996)035<0601:imfaos>2.0.co;2
+	
+	
+		Journal of Applied Meteorology
+		J. Appl. Meteor.
+		0894-8763
+		1520-0450
+		
+			35
+			4
+			
+			April 1996
+			American Meteorological Society
+		
+	
+	Alduchov, O. A. and Eskridge, R. E., "Improved Magnus Form Approximation of Saturation Vapor Pressure," Journal of Applied Meteorology, Vol. 35, No. 4, April 1996, pp. 601-609.
+
+
+
+	
+		Prediction and Use of Contrail Frequency Index for Contrail Reduction Strategies
+		
+			NeilChen
+		
+		
+			BanavarSridhar
+		
+		
+			HokNg
+		
+		10.2514/6.2010-7849
+	
+	
+		AIAA Guidance, Navigation, and Control Conference
+		Toronto, Ontario
+		
+			American Institute of Aeronautics and Astronautics
+			August 2010
+		
+	
+	Chen, N. Y., Sridhar, B., and Ng, H. K., "Prediction and Use of Contrail Frequency Index for Contrail Reduction Strategies," AIAA Guidance, Navigation, and Control Conference, Toronto, Ontario, August 2010.
+
+
+
+	
+		Assessing NAS Performance: Normalizing for the Effects of Weather
+		
+			MBCallaham
+		
+		
+			JSDearmon
+		
+		
+			ACooper
+		
+		
+			JHGoodfriend
+		
+		
+			DMoch-Mooney
+		
+		
+			GSolomos
+		
+	
+	
+		4th USA/Europe Air Traffic Management R&D Symposium
+		Santa Fe, NM
+		
+			December 2001
+		
+	
+	Callaham, M. B., DeArmon, J. S., Cooper, A., Goodfriend, J. H., Moch-Mooney, D., and Solomos, G., "Assessing NAS Performance: Normalizing for the Effects of Weather," 4th USA/Europe Air Traffic Management R&D Symposium, Santa Fe, NM, December 2001.
+
+
+
+	
+		Relationship Between Weather, Traffic and Delay Based on Empirical Methods
+		
+			BanavarSridhar
+		
+		
+			SeanSwei
+		
+		10.2514/6.2006-7760
+	
+	
+		6th AIAA Aviation Technology, Integration and Operations Conference (ATIO)
+		Wichita, KS
+		
+			American Institute of Aeronautics and Astronautics
+			September 2006
+		
+	
+	Sridhar, B. and Swei, S., "Relationship between Weather, Traffic and Delay Based on Empirical Methods," 6th AIAA Aviation Technology, Integration and Operations Conference (ATIO), Wichita, KS, September 2006.
+
+
+
+	
+		Estimation of Air Traffic Delay Using Three Dimensional Weather Information
+		
+			NeilChen
+		
+		
+			BanavarSridhar
+		
+		10.2514/6.2008-8916
+	
+	
+		The 26th Congress of ICAS and 8th AIAA ATIO
+		Anchrorage, AK; Washington, DC
+		
+			American Institute of Aeronautics and Astronautics
+			September 2008. March 2009. October 2010
+			3
+		
+	
+	Aviation Environmental Design Tool (AEDT) User Guide: Beta1c
+	Chen, N. Y. and Sridhar, B., "Estimation of Air Traffic Delay Using Three Dimensional Weather Information," 8th AIAA Aviation Technology, Integration and Operations Conference (ATIO), Anchrorage, AK, September 2008. 21 EUROCONTROL Validation Infrastructure Centre of Expertise, France, User manual for the base of aircraft data (BADA), 3rd ed., March 2009. 22 Federal Aviation Administration, Washington, DC, Aviation Environmental Design Tool (AEDT) User Guide: Beta1c, October 2010.
+
+
+
+	
+		The Characteristics of Future Fuels
+		
+			OJHadaller
+		
+		
+			AMMomenthy
+		
+		
+			1989
+		
+	
+	Project Report D6-54940, Boeing publication
+	Hadaller, O. J. and Momenthy, A. M., "The Characteristics of Future Fuels," Project Report D6-54940, Boeing publica- tion, 1989.
+
+
+
+	
+		Development of an EMF Measurements Database, EMF Rapid Program, Project #5, Interim Report: April 1995-December 1996
+		
+			SBaughcuma
+		
+		
+			TTritz
+		
+		
+			SHenderson
+		
+		
+			DPickett
+		
+		10.2172/2440
+		NASA CR 4700
+		
+			April 1996
+			Office of Scientific and Technical Information (OSTI)
+		
+	
+	Project Report
+	Baughcuma, S., Tritz, T., Henderson, S., and Pickett, D., "Scheduled Civil Aircraft Emission Inventories for 1992: Database Development and Analysis," Project Report NASA CR 4700, April 1996.
+
+
+
+	
+		Impact of Uncertainty on the Prediction of Airspace Complexity of Congested Sectors
+		
+			BanavarSridhar
+		
+		
+			DeepakKulkarni
+		
+		
+			KapilSheth
+		
+		10.2514/atcq.19.1.1
+	
+	
+		Air Traffic Control Quarterly
+		Air Traffic Control Quarterly
+		1064-3818
+		2472-5757
+		
+			19
+			1
+			
+			2011
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	Sridhar, B., Kulkarni, D., and Sheth, K., "Impact of Uncertainty on the Prediction of Airspace Complexity of Congested Sectors," Air Traffic Control Quarterly, Vol. 19, No. 1, 2011, pp. 1-23.
+
+
+
+	
+		Methods for Evaluating Environmental and Performance Tradeoffs for Air Transportation Systems
+		
+			MGregoryO'neill
+		
+		
+			Jean-MarieDumont
+		
+		
+			TomReynolds
+		
+		
+			JohnHansman
+		
+		10.2514/6.2011-6816
+	
+	
+		11th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference
+		Virginia Beach, VA
+		
+			American Institute of Aeronautics and Astronautics
+			September 2011
+		
+	
+	ONeill, M., Dumont, J., Reynolds, T., and Hansman, J., "Methods for Evaluating Environmental and Performance Tradeoffs for Air Transportation Systems," 11th AIAA Aviation Technology, Integration and Operations Conference (ATIO), Virginia Beach, VA, September 2011.
+
+
+
+	
+		Transport impacts on atmosphere and climate: Metrics
+		
+			JSFuglestvedt
+		
+		
+			KPShine
+		
+		
+			TBerntsen
+		
+		
+			JCook
+		
+		
+			DSLee
+		
+		
+			AStenke
+		
+		
+			RBSkeie
+		
+		
+			GJ MVelders
+		
+		
+			IAWaitz
+		
+		10.1016/j.atmosenv.2009.04.044
+	
+	
+		Atmospheric Environment
+		Atmospheric Environment
+		1352-2310
+		
+			44
+			37
+			
+			December 2010
+			Elsevier BV
+		
+	
+	Fuglestvedta, J., Shineb, K., Berntsen, T., Cook, J., Lee, D., Stenke, A., Skeie, R., Velders, G., and Waitz, I., "Transport impacts on atmosphere and climate: Metrics," Atmospheric Environment, Vol. 44, No. 37, December 2010, pp. 4648-4677.
+
+
+				
+			
+		
+	

file139.txt

@@ -0,0 +1,660 @@
+
+	
+	
+		
+I. IntroductionAircraft-induced environmental impact has drawn attention in recent years. 1 A recent study estimates that aviation is responsible for 13% of transportation-related fossil fuel consumption and 2% of all anthropogenic carbon dioxide emissions. 2Domestic air traffic is expected to grow at an annual rate of 3.5% over the next 20 years, and the global air traffic is expected to grow more rapidly at an annual rate of 4.8% from 2011 to 2030. 3 To address the aviation environment impacts with the forecast in air traffic growth, various methods have been proposed.The three largest environmental impacts for enroute air traffic include direct emissions of greenhouse gases such as carbon dioxide (CO 2 ), emissions of nitrogen oxides (NO x ), and persistent contrails.CO 2 and NO x emissions are related to fuel burn therefore minimizing fuel consumption results in minimal emission solutions.Various procedures have been proposed in the past to reduce the persistent contrail formation, including promising approaches based on changing aircraft flight altitudes.Mannstein 4 proposed a strategy to reduce the climate impact of contrails significantly by only small changes in individual flight altitude.Williams 5,6 proposed strategies for contrail reduction by identifying fixed and varying maximum altitude restriction policies.However, these restrictions generally imply more fuel burn, thus more emissions, and add congestion to the already crowded airspace at lower altitudes.Sridhar, 7 Chen, 8 and Wei 9 proposed contrail reduction strategies by altering an aircraft's cruising altitude in a fuel-efficient way, but these strategies did not address the environmental impact from aircraft emissions.Recently, the Absolute Global Temperature Potential was introduced in Ref. 10 and 11 to study the combined effect of CO 2 emissions and contrail formation on the reduction strategies, and effect of NO x was added in Ref. 12.However, none of the above evaluates both the reduction in environmental cost and the increase in operational costs for the reduction strategies.The idea of placing a financial cost to the impact aircraft operations have on the environment has been used by Virgin America airlines.Virgin America offers passengers the option to pay for carbon-offset based on the length of their flight. 13A methodology can be developed to evaluate a policy that seeks to minimize the environmental impact due to aircraft operations while considering the cost to the airline for invoking such a policy.The objective of this paper is to evaluate the tradeoff between environmental impact reduction and the corresponding operational costs for enroute air traffic.First, a linear climate model was used to convert climate effects of CO 2 emissions and aircraft contrails to changes in Absolute Global Temperature Potential, 14 a metric that measures the mean surface temperature change due to aircraft emissions and persistent contrail formations.NO x is not considered since its effect on the reduction strategy is minor. 12Next, the concept of social cost of carbon 15 and the carbon auction price from California's 2013 cap-and-trade system 16 were used to provide an estimate of the environmental cost of CO 2 , which was used to estimate the cost of contrails.Even though the estimate of the cost is highly uncertain, 17 a suggested value was used and sensitivity analysis was conducted.The environmental impact reduction strategy uses a previously developed fuelefficient contrail reduction strategy 8 to minimize the combined impacts of emissions and contrails.The strategy minimizes the environmental impact by altering the aircraft's cruising altitude while computing the additional fuel burn and emissions.Some policies may consider this strategy to be favorable when the reduction in the combined environmental cost exceeds the increase in operational cost with a certain tradeoff factor.This paper evaluates how the net environmental benefit varies with different decision-making time-horizons, carbon and fuel costs, and atmospheric conditions.Introducing the cost models provides a method to tradeoff environmental cost and operational cost that will result in maximal net environmental benefit.The remainder of the paper is organized as follows.Section II provides descriptions of the linear climate models, the environmental impact reduction strategy, and the environmental and cost models.Next, Section III shows the results and analysis of environmental reduction strategies with various parameters.Finally, Section IV presents a summary and conclusions.
+II. Models and Methods
+II.A. Linear Climate ModelsThe climate response to aviation emission and contrails can be modeled as outputs from a series of linear dynamic systems.The carbon cycle models describe the changes to the CO 2 concentration due to the transport and absorption of CO 2 by the land mass and various ocean layers.The Radiative Forcing (RF) for CO 2 emissions is comprised of a steady-state component and three exponentially decaying components. 18oncentration dynamics of other non-CO 2 greenhouse gases can be described by first order linear systems.Radiative Forcing due to different emissions affects the climate by changing the Earth's global average nearsurface air temperature and the temperature response and energy balance to RF can be modeled using either a first order linear model 19 or a second order linear model. 20ontrails form when a mixture of warm engine exhaust gases and cold ambient air reaches saturation with respect to water, forming liquid drops which quickly freeze.Contrails occur at different regions of the earth and add non-uniform sources of RF to the atmosphere.The latest estimates indicate that contrails caused by aircraft may be causing more climate warming today than all the residual CO 2 emitted by aircraft. 21The net RF for contrails includes the effect of trapping outgoing longwave radiation from the Earth and that of reflecting incoming shortwave radiation from the sun.Energy Forcing (EF) is the net energy flux induced to the atmosphere by a unit length of contrail over its lifetime.Estimates of EF given the RF forcing due to contrails are described in Ref. 22.The lifetime associated with different emissions and contrails varies from a few hours to several hundred years.The impact of certain gases depends on the amount and location of the emission, and the decision-making time horizon, H in years, when the impact is estimated.These variations make it necessary to develop a common yardstick to measure the impact of various gases.Several climate metrics have been developed to assess the impact of the aviation emissions.Using linear climate response models, the Absolute Global Temperature Potential (AGTP) measures the mean surface temperature change because of different aircraft emissions and persistent contrail formations. 14AGTP provides a way to express the combined environmental cost of emissions and contrails as a function of the fuel cost.Only CO 2 emissions are considered in this paper, as the effect of NO x emissions are relatively small compared with CO 2 . 12Assume that the RF due to contrails is independent of the location of the contrails, the near surface temperature change ∆T , in Kelvin (K), for the decision-making time horizon of H years, can be approximated as∆T (H) = ∆T CO2 (H) + ∆T Con (H),(1)where ∆T CO2 (H) is the contribution to AGTP from CO 2 emissions for the time horizon of H years and is a linear function of additional CO 2 emissions, and ∆T Con (H) is the contribution to AGTP from contrails for the time horizon of H years and is a linear function of contrail length.The units of ∆T CO2 and ∆T Con are also in Kelvin.The coefficients of the linear functions depend on the linear models for RF, the specific forcing because of CO 2 , energy forcing because of contrails, energy balance model and the duration of the climate effect horizon. 10 Using the coefficients described in Ref. 12, Eq.( 1) can be rewritten as∆T (H) = α(H)E CO2 + β(H)L Con ,(2)where α(H) is the coefficient of AGTP due to CO 2 for the time horizon of H in K/kg, β(H) is the coefficient of AGTP due to contrails for the time horizon of H in K/km, E CO2 is the amount of CO 2 emissions in kg, and L Con is the contrail length in km.A list of α(H) and β(H), derived from Ref. 12, is shown in Table 1.Notice that the AGTP coefficient for contrails is much larger at shorter time horizons and smaller at longer time horizons, as contrails have more short-term environmental impact; the AGTP coefficient for CO 2 does not change much with different time horizons.Table 1.AGTP coefficients for CO2 and contrails for three different time horizonsTime Horizon H = 25 years H = 50 years H = 100 years α(H), K/kg 6.71×10 -16 5.78×10 -16 5.07×10 -16 β(H), K/km 2.99×10 -14 6.98×10 -15 5.10×10 -15
+II.B. Environmental Impact ReductionPrevious research 3 shows that the aviation environmental effect can be reduced efficiently by only changing the flight cruise altitude.This paper modifies the contrail reduction strategy described in Ref. 8 and uses the approach to reduce AGTP rather than contrails.The strategy divides the U.S. National Airspace System into twenty regions horizontally based on the twenty continental U.S. Air Traffic Control Centers (Centers), and ten levels vertically, from 26,000 feet to 44,000 feet at increments of 2,000 feet.At each hour, the strategy looks at all aircraft cruising in a Center at the same flight level, alters their cruise altitude by -4,000, -2000, +2000, or +4,000 feet, and selects the optimal cruise altitude that provides the minimal ∆T .The strategy also computes the additional fuel burn needed for such a move, and uses a fuel-efficient index, the ratio of the ∆T reduction and the additional fuel burn, to determine the temperature to fuel changes ratio.For example, if (a) moving all the aircraft at a Center up 2,000 feet will burn 1,000 kg more fuel for the climb and the remainder of the flight in the Center, and reduce ∆T by 2 × 10 -10 K, or if (b) moving the aircraft down 2,000 will reduce ∆T by 3 × 10 -10 K but will burn 10,000 kg additional fuel for the descent and the remainder of the flight in the Center, the strategy to minimize the climate impact will choose (b) to move aircraft 2,000 feet lower to achieve a greater reduction in ∆T .However, if the strategy looks at the fuel-efficiency index and only moves aircraft when the fuel-efficient index is greater than 10 -10 K/ 1000 kg, the strategy will choose (a) to move aircraft 2,000 feet higher, even though the ∆T reduction is 10 -10 K less, and the additional fuel burn is 10 times less.Using the different thresholds on the fuel-efficient index allows the strategy to tradeoff fuel burn with ∆T .Note that the strategy is applied to each Center at each  hour independently.Also these altitude changes are subject to the cruise altitude limits of each aircraft.An additional constraint is added such that where an aircraft crosses a sector boundary and causes congestion, it will stay at the original cruise altitude.Figure 1 presents the results from a 24-hour simulation based on historical data on April 19, 2010.The environmental impact reduction strategy, which allows the aircraft cruise altitudes to change in the range of -4,000 to +4,000 feet, was applied to the historical data, and the trade-off between AGTP due to CO 2 emissions, AGTP due to contrails, and total AGTP and additional fuel consumption for the decision-making time horizon of 100 years were summarized in Fig. 1a.The corresponding reduction in contrail length and additional CO 2 emissions are shown in Fig, 1b.In Fig. 1a, the contribution to AGTP from CO 2 emissions, the black line, increases linearly with additional fuel burn.The AGTP due to contrails, the green line, decreases faster at the beginning, and slower with more additional fuel burn.This is because the strategy selected the altitude changes with higher fuel-efficiency index first, resulting in more AGTP reduction with less additional fuel burn at the beginning (left end of the curve); the changes with lower fuel-efficiency index were then selected that slowed down the AGTP reduction rate (right end of the curve).The cumulative AGTP, the blue line, decreases initially with reduction in contribution from contrails and is eventually offset by the increase in contribution from CO2 emissions.The curves show that even if the cost of fuel is not taken into consideration, under certain conditions, reducing contrails beyond a certain level may neither be economical nor good environmental policy.
+II.C. Cost ModelThe United States Government recently concluded a process to develop a range of values representing the monetized damages associated with an incremental increase in CO 2 emissions, commonly referred to as the social cost of carbon. 15These values were used in benefit-cost analyses to assess potential federal regulations.In California, the state has a carbon cap-and-trade system which is the largest of its kind in the U.S. and the second-biggest carbon market in the world behind the European Unions. 16California cites its program as an example for the rest of the world to follow, and plans to use it and other emissions-reduction measures to cut greenhouse-gas pollution to 1990 levels by 2020.The cap-and-trade system recently sold carbon allowances for $13.62 per metric ton.This paper attempts to relate AGTP due to CO 2 emissions and aircraft contrails to the environmental cost in dollar amounts in order to perform a quantitative analysis of the environmental benefit resulting from the environmental impact reduction strategy.Using the social cost of carbon dioxide as an estimate of environmental cost of CO 2 due to warming, the additional contribution to environmental cost from CO 2 emissions, ∆Cost CO2 , can be formulated as∆Cost CO2 = SCC • ∆E CO2 1000 ,(3)where SCC is the social cost of carbon in dollar per metric ton, and ∆E CO2 is the changes in CO 2 emissions in kg.In order to quantify the environmental cost of contrails, the environmental cost of temperature changes, specifically one Kelvin of AGTP, was defined using the SCC and the AGTP coefficient of CO 2 for time horizon H years,ECK = SCC 1000 • α(H) , (4)where ECK is the equivalent environmental cost of temperature change in dollars per Kelvin and α(H) is the AGTP coefficient of CO 2 for the time horizon of H years listed in Table 1.Using the ECK to relate the environmental cost from contrails, ∆Cost H Con , to ∆Cost CO2 assuming that the same ∆T CO2 and ∆T Con have the same environmental cost for the time horizon of H years, ∆Cost H Con can be formulated as∆Cost H Con = ECK • ∆T Con (H) = SCC 1000 • β(H) α(H) • ∆L Con ,(5)where ∆L Con is the change in contrail length, and β(H) is the AGTP coefficient of contrails for the time horizon of H years listed in Table 1.In general, ∆L Con is negative as the strategy is reducing the contrail length and ∆Cost CO2 is positive due to the additional fuel burn.The superscript H in ∆Cost H Con indicates the environment cost due to contrails depends on the decision-making time horizon.The combined environmental cost changes, ∆Cost H Env , from both CO 2 and contrails for time horizon of H years can be written as ∆CostH Env = ∆Cost CO2 + ∆Cost H Con ,(6)All ∆Cost H Env , ∆Cost CO2 , and ∆Cost H Con are in US dollars.Note that ∆Cost H Env is always negative after the environmental impact reduction strategy.The net environmental benefit index, N BI H Env , is defined asN BI H Env = -∆Cost H Env -∆Cost Opr ,(7)where ∆Cost Opr is the additional operational cost of applying the environmental impact reduction strategy.Only the cost of additional fuel burn is considered as additional operational cost in this paper.Note that since ∆Cost H Env is always negative after the environmental impact reduction strategy, the first term in Eq.( 7), -∆Cost H Env , indicates the environmental cost savings.For the same example in the previous subsections, using a social cost of CO 2 of $21 per metric ton suggested by the United States Government 15 as an estimate of the environmental cost of CO 2 , the fuel cost of $4 per gallon, and the fuel density of 0.82 kilogram per liter, the AGTP and additional fuel burn in Fig. 1a were converted into the environmental cost reduction, -∆Cost H Env , and additional operational cost, ∆Cost H Opr , are shown in Fig. 2a.The blue curve shows the environmental cost reduction versus the additional operational cost after the environmental reduction strategy.The black dash line is a straight line with a slope of one.When the blue curve is above the black line, it suggests that the reduction strategy provided a positive net benefit.The net benefit versus the additional operational cost is shown in Fig. 2b.At the apex of the curve, marked as 'x,' that the strategy could provide a positive N BI H=100 Env of around $57, 000, or equivalent to around 2, 700 tons of CO 2 , after applying the reduction strategy at the point that the strategy will burn an additional 1.05 × 10 5 kg fuel for all aircraft.When the blue curve falls below the black line in Fig. 2a and Fig. 2b, it suggests that the additional cost for the strategy exceeded the environmental benefit thus the strategy is not recommended.Introducing the cost model provides a solution to select the fuel-efficiency index described in Section II.B that will result in the most net environmental benefit.
+III. AnalysisThe cost models introduced in the previous section can be used to evaluate the environmental impact reduction strategy with different parameters, including the decision-making time-horizon of environmental impact and the cost estimate of CO 2 , and the fuel cost.The variation due to different days are also shown in this section.The social cost of carbon was used as an estimate of the environmental cost of CO 2 .The social cost of temperature changes, defined in Eq.( 4), was used to relate the environmental cost of contrails to CO 2 .
+III.A. Varying Decision-Making Time-HorizonSince CO 2 emissions and aircraft contrails have different life times, a parameter of decision-making timehorizon H needs to be defined to compute the Absolute Global Temperature Potential and evaluate the  environmental impact.Three different time horizons, 25, 50, and 100 years were considered.Figure 3a shows the environmental cost saving versus the additional operational cost with different time horizons.The social cost of CO 2 at $21 per metric ton was used as an estimate of the environmental cost of CO 2 , the social cost of temperature changes at time horizon 100 years was used to estimate the environmental cost of contrails, and the fuel cost of $4 per gallon was used in this analysis.The blue line in the figure is the same as in Fig. 2a for H = 100, and the green and magenta lines are for H = 50 and H = 25 respectively.As shown in the figure, the magenta line is much higher than the blue and green lines, and also above the black dashed-line all the time.This indicates that shorter time horizon would result in more short-term environmental cost savings for the same operational cost.This is because aircraft contrails have shorter life time than CO 2 so the benefit from contrail reductions is more obvious in a shorter time-horizon.For longer time-horizons, the impact of contrails decays and the relative impact from CO 2 becomes larger.The net environmental benefit for different time-horizons after applying the environmental impact reduction strategy described in Sec.II.B can be seen in Fig. 3b.Same as in Fig. 2a, at H = 100 (blue line), the strategy could result in an net environmental benefit of around $57, 000, or around 2, 700 tons of CO 2 equivalent for all aircraft in the U.S. on April 19, 2010, indicated at the blue 'x' in Fig. 3b.For a shorter time-horizon such as H = 50 (green line), the strategy could result in net environmental benefit of around $129, 000, or around 6, 100 tons of CO 2 equivalent, indicated at the green 'x'.For H = 25 (magenta line), the strategy could result in net environmental benefit of around $1, 421, 000, or around 67, 700 tons of CO 2 equivalent, indicated at the magenta 'x.'It is worth mentioning that the environmental cost saving and net benefit are time-horizon-dependent, meaning a net gain in benefit in a 25-year time horizon might turn into net loss in benefit at 50-or 100-year time horizons because the benefit from reducing contrails decays with the length of the time-horizon.Figure 4 shows how the maximum net benefit decays with time.The upper right magenta 'x' in the figure is the same as the magenta 'x' in Fig. 3b, showing an net environmental benefit of $1,421,000 at H = 25.The benefit decays to -$267, 000 at H = 50 and -$400,000 at H = 100, as the magenta line suggested.If the decision-making time horizon for the reduction strategy is H = 50, the net benefit decays from $129,000 at H = 50 to $6,100 at H = 100 (green line), which happens to be the net benefit for the strategy with decision time horizon of H = 100.This is because the strategy for decision time horizon H = 50 and H = 100 are the same in this case.The strategy may behave differently with different time horizons and the net environmental benefit may also vary.
+III.B. Varying Estimate of the Cost of Carbon DixocideEven though an approximate social cost of CO 2 is suggested, 15 the estimate of the cost is highly uncertain. 17n addition to the suggested price at $21 per ton of CO 2 , a sensitivity analysis was conducted using prices of $5 and $64 suggested in Ref. 15.Another good reference of the carbon cost is the auction price under California's cap-and-trade system in 2013, at $13.62 per metric ton of CO 2 . 16igure 5a is the same as Fig. 3b and is placed here for easier comparison.Figures 5b,5c, and 5d show  the net environmental benefit curves after reduction strategy for three different time horizons with different estimates of CO 2 cost with the fuel cost of $4 per gallon.Note that the scales on y-axis in these figures are different in order to shows the variations of the three curves in each individual plot.The maximum net benefit from the strategy is marked as 'x.'If the 'x' is located at the origin, it means there is no feasible solution to reduce environmental impact given the time horizon and the estimate of CO 2 cost.With higher estimate of CO 2 cost of $65, shown in Fig. 5b, the strategy results in more net benefit compared to that in Fig. 5a.On the other hand, when the estimated cost of CO 2 is small, the environmental benefit was offset by the relatively high operational cost.When the cost is $5, the strategy can only achieve net benefit at the 25-year time horizon, shown in Fig. 5c.Even with the estimate cost of CO 2 at $13.62, the current California auction price, the strategy cannot find a feasible solution for the net environmental benefit for time horizons of 50-and 100-years; the strategy can only achieve net benefit in a 25-year time horizon.In order to achieve more net benefit with a given set of time horizons and estimates of CO 2 costs, the efficiency of the environmental impact reduction strategy needs to be improved.Note that the strategy used in this paper is very conservative.It alters the cruise altitudes for all the aircraft within a Center to certain specified altitudes.The strategy can be improved by using a finer spatial resolution 9 and a resulting increase in net environmental benefit.Increasing the carbon cost or reducing the fuel cost will help the strategy to achieve more net environmental benefit.The net benefit with different estimated costs of CO 2 and fuel costs are shown in Table 2.For the environmental impact reduction strategy used in this paper, the net benefit will turn positive at a CO 2 price of $20 per ton with the fuel cost of $4 per gallon for H = 100, about 47% more than the current California auction price.
+III.C. Variation on Different DaysThe same simulation and analysis were applied to the entire month of April, 2010 based on the historical air traffic and atmospheric data with the estimated environmental cost of CO 2 at $21 and the fuel cost at $4.The daily net environmental benefit for the month with time horizon 25, 50, and 100 years are shown in Fig. 6.The daily net environmental benefits vary on different days mainly because of different atmospheric  conditions.The net benefit with decision-making time horizon of 25 years (magenta bar) are much higher than the net benefit with time horizon of 50 years (green bars) and 100 years (blue bars).The average daily net benefit for the month is $773,000 for H = 25, $102,000 for H = 50, and $63,000 for H = 100.The results show that the environmental impact reduction can achieve net benefit (environmental cost reduction is greater than the operational cost) for all time horizons on all 30 days in April, 2010.The daily total aircraft contrail length is also shown in the figure (green line).The daily contrail length is normalized so that it has the same magnitude as the environmental net cost at H = 25 (magenta bars).It is clear that the daily net benefit for H = 25 is highly correlated with the daily total contrail length; the correlation coefficient is 0.92.It is not surprising as the net benefit of the reduction strategy mainly comes from the reduction in contrail length, and in general more aircraft contrails can be reduced on days with more contrail formations.The correlations are not as high for H = 50 and H = 100.The results show that the environmental impact reduction strategy can reduce environmental cost effectively so that it outweighs the additional operational cost on days with different atmospheric conditions.
+IV. ConclusionsThis paper provides a method to evaluate the tradeoffs between environmental impact and the corresponding operational costs for enroute air traffic.A linear climate model and the concept of social carbon cost and Absolute Global Temperature Change Potential were used to provide an estimate of the aviation environmental costs.An environmental impact reduction strategy was introduced to reduce environmental costs by changing aircraft's cruise altitude while computing additional operational costs.Depending on the specific environmental policy, the strategy is considered favorable when the reduction in environmental costs exceeds the increase in operational costs.It is shown that the reduction strategy can achieve more environmental benefit with shorter decision-making time horizons.The results show at the current suggested social cost of CO 2 at $21 per metric ton and higher, the reduction strategy can achieve net benefits in 25-, 50-, and 100-year time horizons.However, at the recent California carbon auction price of $13.62 per metric ton, the strategy can only achieve net benefit at the 25-and 50-year time horizons.The auction price needs to be about 47% more than the current price in order to see net benefit in 100-year time-horizon.Increasing the efficiency of the strategy or reducing the operational cost would also gain more net benefit.The results also show that the reduction strategy can achieve net environmental benefit on days with different atmospheric conditions, and the daily net benefit for the 25-year time horizon is highly correlated with the daily aircraft contrail formations.This tradeoff study provides guidance to environmental policy that will result in the most net environmental benefit.CO 2 emissions and contrails
+Figure 1 .1Figure 1.AGTP (H=100), CO2 emissions, and contrail length versus additional fuel burn after the environmental reduction strategy for all flights on April 19, 2010.
+Figure 2 .2Figure 2. Environmental cost saving and net benefit for all flights on April 19, 2010.
+Figure 3 .Figure 4 .34Figure 3. Environmental cost saving index and factor with different time horizons for all flights on April 19, 2010.
+Estimate cost of CO 2 =$21 per ton, fuel cost $4 per gallon , US$ Environmental Net Benefit, US$ (b) Estimate cost of CO 2 =$65 per ton, fuel cost $4 per gallon , US$ Environmental Net Benefit, US$ (c) Estimate cost of CO 2 =$5 per ton, fuel cost $4 per gallon , US$ Environmental Net Benefit, US$ (d) Estimate cost of CO 2 =$13.62 per ton, fuel cost $4 per gallon
+Figure 5 .5Figure 5. Net environmental benefit index with different social cost of CO2 for all flights on April 19, 2010.
+Figure 6 .6Figure 6.Daily maximum net benefit with different decision-making time horizon and contrails for all flights in April, 2010.
+Table 2 .2Net environmental benefit after impact reduction strategy for all flights on April 19,2010Estimate of CO 2 costFuel CostH = 25 years H = 50 years H = 100 years$5 per ton$4 per gallon$140,000$0$0$13.62 per ton$4 per gallon$750,000$36,000$0$21 per ton$4 per gallon$1,421,000$129,000$57,000$65 per ton$4 per gallon$6,103,000$826,000$483,000$21 per ton$3 per gallon$1,606,000$162,000$91,000$21 per ton$4 per gallon$1,421,000$129,000$57,000$21 per ton$5 per gallon$1,276,000$95,000$23,000$21 per ton$6 per gallon$1,173,000$61,000$015x 10 5Environmental Net Benefit, US$5 10net benefit (H=25) net benefit (H=50) net benefit (H=100) normalized contrail length0 051015202530day of April, 2010
+		
+		
+			
+
+				
+
+
+	
+		
+			IWaitz
+		
+		
+			JTownsend
+		
+		
+			JCutcher-Gershenfeld
+		
+		
+			EGreitzer
+		
+		
+			JKerrebrock
+		
+		Report to the United States Congress: Aviation and the Environment, A National Vision, Framework for Goals and Recommended Actions
+		London, UK
+		
+			December 2004
+		
+	
+	Tech. rep
+	Partnership for AiR Transportation Noise and Emissions Reduction
+	Waitz, I., Townsend, J., Cutcher-Gershenfeld, J., Greitzer, E., and Kerrebrock, J., "Report to the United States Congress: Aviation and the Environment, A National Vision, Framework for Goals and Recommended Actions," Tech. rep., Partnership for AiR Transportation Noise and Emissions Reduction, London, UK, December 2004.
+
+
+
+	
+		Impact of Aviation on Climate
+		
+			GuyPBrasseur
+		
+		
+			MohanGupta
+		
+		10.1175/2009bams2850.1
+	
+	
+		Bulletin of the American Meteorological Society
+		Bull. Amer. Meteor. Soc.
+		0003-0007
+		1520-0477
+		
+			91
+			4
+			
+			2010
+			American Meteorological Society
+		
+	
+	Brasseur, G. P. and Gupta, M., "Impact of Aviation on Climate: Research Priorities," Bulletin of the American Meteo- rological Society, Vol. 91, No. 4, 2010, pp. 461-463.
+
+
+
+	
+		Call for Papers
+		
+			BSridhar
+		
+		
+			NYChen
+		
+		
+			HKNg
+		
+		10.1027/2192-0923/a000067
+	
+	
+		Aviation Psychology and Applied Human Factors
+		Aviation Psychology and Applied Human Factors
+		2192-0923
+		2192-0931
+		
+			4
+			2
+			
+			June 2013
+			Hogrefe Publishing Group
+			Chicago, IL
+		
+	
+	Tenth USA
+	Sridhar, B., Chen, N. Y., and Ng, H. K., "Energy Efficient Contrail Mitigation Strategies for Reducing the Environmental Impact of Aviation," Tenth USA/Europe Air Traffic Management Research and Development Seminar , Chicago, IL, June 2013.
+
+
+
+	
+		A note on how to avoid contrail cirrus
+		
+			HermannMannstein
+		
+		
+			PeterSpichtinger
+		
+		
+			KlausGierens
+		
+		10.1016/j.trd.2005.04.012
+	
+	
+		Transportation Research Part D: Transport and Environment
+		Transportation Research Part D: Transport and Environment
+		1361-9209
+		
+			10
+			5
+			
+			September 2005
+			Elsevier BV
+		
+	
+	Mannstein, H., Spichtinger, P., and Gierens, K., "A note on how to avoid contrail cirrus," Transportation Research. Part D, Transport and environment, Vol. 10, No. 5, September 2005, pp. 421-426.
+
+
+
+	
+		Reducing the climate change impacts of aviation by restricting cruise altitudes
+		
+			VictoriaWilliams
+		
+		
+			RobertBNoland
+		
+		
+			RalfToumi
+		
+		10.1016/s1361-9209(02)00013-5
+	
+	
+		Transportation Research Part D: Transport and Environment
+		Transportation Research Part D: Transport and Environment
+		1361-9209
+		
+			7
+			6
+			
+			November 2002
+			Elsevier BV
+		
+	
+	Williams, V., Noland, R. B., and Toumi, R., "Reducing the climate change impacts of aviation by restricting cruise altitudes," Transportation Research. Part D, Transport and environment, Vol. 7, No. 6, November 2002, pp. 451-464.
+
+
+
+	
+		Variability of contrail formation conditions and the implications for policies to reduce the climate impacts of aviation
+		
+			VictoriaWilliams
+		
+		
+			RobertBNoland
+		
+		10.1016/j.trd.2005.04.003
+	
+	
+		Transportation Research Part D: Transport and Environment
+		Transportation Research Part D: Transport and Environment
+		1361-9209
+		
+			10
+			4
+			
+			July 2005
+			Elsevier BV
+		
+	
+	Williams, V. and Noland, R. B., "Variability of contrail formation conditions and the implications for policies to reduce the climate impacts of aviation," Transportation Research. Part D, Transport and environment, Vol. 10, No. 4, July 2005, pp. 269-280.
+
+
+
+	
+		Fuel efficient strategies for reducing contrail formations in United States airspace
+		
+			BanavarSridhar
+		
+		
+			NeilYChen
+		
+		10.1109/dasc.2010.5655533
+	
+	
+		29th Digital Avionics Systems Conference
+		Salt Lake City, UT
+		
+			IEEE
+			October 2010
+		
+	
+	Sridhar, B., Chen, N. Y., and Ng, H. K., "Fuel Efficient Strategies for Reducing Contrail Formations in United State National Air Space," 29th Digital Avionics Systems Conference, Salt Lake City, UT, October 2010.
+
+
+
+	
+		Tradeoff Between Contrail Reduction and Emissions in United States National Airspace
+		
+			NeilYChen
+		
+		
+			BanavarSridhar
+		
+		
+			HokKNg
+		
+		10.2514/1.c031680
+	
+	
+		Journal of Aircraft
+		Journal of Aircraft
+		0021-8669
+		1533-3868
+		
+			49
+			5
+			
+			2012
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	Chen, N. Y., Sridhar, B., and Ng, H. K., "Tradeoff between Contrail Reduction and Emissions in United States National Airspace," Journal of Aircraft, Vol. 49, No. 5, 2012, pp. 1367-1375.
+
+
+
+	
+		A Linear Programming Approach to the Development of Contrail Reduction Strategies Satisfying Operationally Feasible Constraints
+		
+			PengWei
+		
+		
+			BanavarSridhar
+		
+		
+			NeilChen
+		
+		
+			DengfengSun
+		
+		10.2514/6.2012-4754
+	
+	
+		AIAA Guidance, Navigation, and Control Conference
+		Minneapolis, MN
+		
+			American Institute of Aeronautics and Astronautics
+			August 2012
+		
+	
+	Wei, P., Sridhar, B., Chen, N., and Sun, D., "A Linear Programming Approach to the Development of Contrail Reduc- tion Strategies Satisfying Operationally Feasible Constraints," AIAA Guidance, Navigation and Control Conference, AIAA, Minneapolis, MN, August 2012.
+
+
+
+	
+		Integration of Linear Dynamic Emission and Climate Models with Air Traffic Simulations
+		
+			BanavarSridhar
+		
+		
+			HokNg
+		
+		
+			NeilChen
+		
+		10.2514/6.2012-4756
+	
+	
+		AIAA Guidance, Navigation, and Control Conference
+		Minneapolis, MN
+		
+			American Institute of Aeronautics and Astronautics
+			August 2012
+		
+	
+	Sridhar, B., Ng, H., and Chen, N., "Integration of Linear Dynamic Emission and Climate Models with Air Traffic Simulations," AIAA Guidance, Navigation and Control Conference, AIAA, Minneapolis, MN, August 2012.
+
+
+
+	
+		Aircraft Trajectory Optimization and Contrails Avoidance in the Presence of Winds
+		
+			BanavarSridhar
+		
+		
+			HokNg
+		
+		
+			NeilChen
+		
+		10.2514/6.2010-9139
+	
+	
+		10th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference
+		Brisbane, Australia
+		
+			American Institute of Aeronautics and Astronautics
+			September 2012
+		
+	
+	Sridhar, B., Ng, H., and Chen, N., "Uncertainty Quantification in the Development of Aviation Operations to Reduce Aviation Emissions and Contrails," 28th International Congress of the Aeronautical Sciences, AIAA, Brisbane, Australia, September 2012.
+
+
+
+	
+		Aircraft Trajectory Design Based on Reducing the Combined Effects of Carbon-Di-Oxide, Oxides of Nitrogen and Contrails
+		
+			BanavarSridhar
+		
+		
+			NeilYChen
+		
+		
+			HokNg
+		
+		10.2514/6.2014-0807
+		
+	
+	
+		AIAA Modeling and Simulation Technologies Conference
+		National Harbor, MD
+		
+			American Institute of Aeronautics and Astronautics
+			Jan 2014
+			13
+			
+		
+	
+	Sridhar, B., Chen, N. Y., and Ng, H. K., "Aircraft Trajectory Design Based on Reducing the Combined Effects of Carbon- Dioxide, Oxides of Nitrogen and Contrails," AIAA Modeling and Simulation Technologies (MST) Conference, National Harbor, MD, Jan 2014. 13 https://carbonfund.org/index.php?option=com_zoo&task=item&item_id=3&Itemid=216, Accessed: 2013-12-05.
+
+
+
+	
+		Transport impacts on atmosphere and climate: Shipping
+		
+			VeronikaEyring
+		
+		
+			IvarS AIsaksen
+		
+		
+			TerjeBerntsen
+		
+		
+			WilliamJCollins
+		
+		
+			JamesJCorbett
+		
+		
+			OyvindEndresen
+		
+		
+			RoyGGrainger
+		
+		
+			JanaMoldanova
+		
+		
+			HansSchlager
+		
+		
+			DavidSStevenson
+		
+		10.1016/j.atmosenv.2009.04.059
+	
+	
+		Atmospheric Environment
+		Atmospheric Environment
+		1352-2310
+		
+			44
+			37
+			
+			2010
+			Elsevier BV
+		
+	
+	Eyring, V., Isaksen, I. S., Berntsen, T., Collins, W. J., Corbett, J. J., Endresen, O., Grainger, R. G., Moldanova, J., Schlager, H., and Stevenson, D. S., "Transport impacts on Atmosphere and Climate: Metrics," Atmosphere Environment, Vol. 44, No. 37, 2010, pp. 4648-4677.
+
+
+
+	
+		Estimating the Social Cost of Carbon for Use in U.S. Federal Rulemakings: A Summary and Interpretation
+		
+			MichaelGreenstone
+		
+		
+			ElizabethKopits
+		
+		
+			AnnWolverton
+		
+		10.3386/w16913
+		
+			2011. 2013
+			National Bureau of Economic Research
+		
+	
+	16 California Environmental Protection Agency, A. R. B., "California Air Resources Board Quarterly Auction 2
+	Greenstone, M., Kopits, E., Wolverton, A., Greenstone, M., Kopits, E., and Wolverton, A., "Estimating the Social Cost of Carbon for Use in U.S. Federal Rulemakings: A Summary and Interpretation," 2011. 16 California Environmental Protection Agency, A. R. B., "California Air Resources Board Quarterly Auction 2," 2013.
+
+
+
+	
+		Consideration of costs and damages avoided and/or benefits gained
+		
+			RJKlein
+		
+		
+			SHuq
+		
+		
+			FDenton
+		
+		
+			TEDowning
+		
+		
+			RGRichels
+		
+		
+			JBRobinson
+		
+		
+			FLToth
+		
+		
+			OCanziani
+		
+		
+			JPalutikof
+		
+		
+			PVan Der Linden
+		
+		
+			CHanson
+		
+	
+	
+		Climate Change 2007: Impacts, Adaptation and Vulnerability Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change
+		
+			MParry
+		
+		Cambridge, United Kingdom and New York, NY
+		
+			Cambridge University Press
+			2007
+		
+	
+	Klein, R. J., Huq, S., Denton, F., Downing, T. E., Richels, R. G., Robinson, J. B., and Toth, F. L., "Consideration of costs and damages avoided and/or benefits gained," In: Climate Change 2007: Impacts, Adaptation and Vulnerability Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, edited by M. Parry, O. Canziani, J. Palutikof, P. van der Linden, and C. Hanson, Cambridge University Press, Cambridge, United Kingdom and New York, NY, 2007.
+
+
+
+	
+		Anthropogenic and Natural Radiative Forcing pages 705 to 740
+		
+			PForster
+		
+		
+			VRamaswamy
+		
+		
+			PArtaxo
+		
+		
+			TBerntsen
+		
+		
+			RBetts
+		
+		
+			DFahey
+		
+		
+			JHaywood
+		
+		
+			JLean
+		
+		
+			DLowe
+		
+		
+			GMyhre
+		
+		
+			JNganga
+		
+		
+			RPrinn
+		
+		
+			GRaga
+		
+		
+			MSchulz
+		
+		
+			RVDorland
+		
+		
+			MManning
+		
+		
+			ZChen
+		
+		
+			MMarquis
+		
+		
+			KAveryt
+		
+		
+			MTignor
+		
+		
+			HMiller
+		
+		10.1017/cbo9781107415324.019
+	
+	
+		Climate Change 2013 - The Physical Science Basis
+		
+			SSolomon
+		
+		
+			DQin
+		
+		Cambridge, United Kingdom and New York, NY
+		
+			Cambridge University Press
+			2007. 2007. 19
+			
+		
+	
+	Forster, P., Ramaswamy, V., Artaxo, P., Berntsen, T., Betts, R., Fahey, D., Haywood, J., Lean, J., Lowe, D., Myhre, G., Nganga, J., Prinn, R., Raga, G., Schulz, M., and Dorland, R. V., "Changes in Atmospheric Constituents and in Radiative Forcing," In: Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, edited by S. Solomon, D. Qin, M. Manning, Z. Chen, M. Marquis, K. Averyt, M.Tignor, and H. Miller, Cambridge University Press, Cambridge, United Kingdom and New York, NY, 2007. 19
+
+
+
+	
+		Global Physical Climatology
+		
+			DLHartmann
+		
+		
+			1994
+			Academic Press
+			San Diego, CA
+		
+	
+	Hartmann, D. L., Global Physical Climatology, Academic Press, San Diego, CA, 1994.
+
+
+
+	
+		Climate trade-off between black carbon and carbon dioxide emissions
+		
+			OBoucher
+		
+		
+			MSReddy
+		
+		10.1016/j.enpol.2007.08.039
+	
+	
+		Energy Policy
+		Energy Policy
+		0301-4215
+		
+			36
+			1
+			
+			2008
+			Elsevier BV
+		
+	
+	Boucher, O. and Reddy, M., "Climate trade-off between black carbon and carbon dioxide emissions," Energy Policy, Vol. 36, 2008, pp. 193-200.
+
+
+
+	
+		Seeing through contrails
+		
+			OlivierBoucher
+		
+		10.1038/nclimate1078
+	
+	
+		Nature Climate Change
+		Nature Clim Change
+		1758-678X
+		1758-6798
+		
+			1
+			1
+			
+			2011
+			Springer Science and Business Media LLC
+		
+	
+	Boucher, O., "Atmospheric science: Seeing through contrails," Nature Climate Change, Vol. 1, 2011, pp. 24-25.
+
+
+
+	
+		Potential to reduce the climate impact of aviation by flight level changes
+		
+			UlrichSchumann
+		
+		
+			KasparGraf
+		
+		
+			HermannMannstein
+		
+		10.2514/6.2011-3376
+		AIAA-2011-3376
+	
+	
+		3rd AIAA Atmospheric Space Environments Conference
+		Honolulu, HI
+		
+			American Institute of Aeronautics and Astronautics
+			June 2011
+		
+	
+	Schumann, U., Graf, K., and Mannstein, H., "Potential to Reduce the Climate Impact of Aviation by Flight Level Changes," AIAA Modeling and Simulation Technologies Conference, AIAA-2011-3376, AIAA, Honolulu, HI, June 2011.
+
+
+				
+			
+		
+	

file140.txt

@@ -0,0 +1,672 @@
+
+	
+	
+		
+I. IntroductionI nterest in the effect of aircraft condensation trails or contrails on climate change has increased in recent years. 1 Contrails form in the wake of aircraft for various reasons but the most important is the emission of water vapor. 2They appear and persist if the aircraft location has certain atmospheric conditions.The global mean contrail cover in 1992 is estimated to double by 2015, and quadruple by 2050 due to the increase in air traffic. 3Studies suggest that the environmental impact from persistent contrails may be three to four times, 4 or even ten times 5 larger than that from aviation induced emissions.Therefore, methods of persistent contrail reduction need to be explored to reduce aircraft induced environmental impact.Efforts have been made in the past years to design strategies for reducing persistent contrail formations.Gierens 6 and Noppel 7 reviewed various strategies for contrail avoidance including changing engine architecture, enhancing airframe and engine integration, using alternate fuels, and modifying traffic flow management procedures.Among the traffic flow management solutions, Mannstein 8 proposed a strategy to reduce the climate impact of contrails significantly by small changes to each aircraft's flight altitude.Campbell 9 presented a mixed integer programming methodology to optimally reroute aircraft to avoid the formation of persistent contrails.Fichter 10 showed that the global annual mean contrail coverage could be reduced by reducing the cruise altitude.Williams 11,12 proposed strategies for contrail reduction by identifying fixed and varying maximum altitude restrictions.Sridhar proposed strategies that minimize the persistent contrail formation by altering aircraft cruising altitudes in a fuel efficient way 13 and proposed methods of contrails avoidance in the presence of winds. 14These methods did not take into account the effect of existing cloud coverage and the presence of severe weather.The objective of this paper is to develop contrail reduction strategies using diverse weather resources.Two weather products were used in this paper, the wind, temperature and humidity forecast provided in the National Oceanic and Atmospheric Administration's Rapid Update Cycle and the weather forecast provided by the FAA's Corridor Integrated Weather System.4][15][16] This paper uses the two weather products in a complementary way to develop contrail reduction strategies in the presence of different weather conditions.The concept of contrail frequency index was used to quantify the severity of contrail formation. 17The contrail frequency index was defined as the number of aircraft flying through regions that would form contrails during a period of time.When the weather resource indicated that there were already clouds formed in the regions, the index was discounted.The indices are used to identify the air traffic control centers and altitudes with high potential contrail formation in the next few hours such that different strategies can be used to alter the aircraft cruising altitudes to reduce the contrail formation in the entire airspace.When severe weather presents, the feasibility of the strategies will be evaluated using the weather information.The proposed contrail reduction strategies are suitable for reducing the contrail formation in the entire airspace with the presence of clouds and severe weather conditions.The remainder of the paper is organized as follows.Section II provides the descriptions of weather and contrail weather models.Section III demonstrates the correlations between contrail and cloud coverage regions.Section IV derives the contrail frequency index using different weather resources and its usage for contrail reduction strategies.Finally, a summary and conclusions are presented in Section V.
+II. Models
+A. Weather ModelThis paper uses the FAA's Corridor Integrated Weather System (CIWS) weather model. 18CIWS provides accurate, low-latency, high-resolution three-dimensional (3D) weather information and forecasts for up to two hours.It has a temporal resolution of two and a half minutes and a spatial resolution of one kilometer by one kilometer.The CIWS 3D weather depiction is composed of two main product types: precipitation, or vertically integrated liquid (VIL), and echo tops.Figure 1a shows the CIWS precipitation information at 8AM eastern daylight time (EDT) on April 23, 2010.The colors indicate the VIL level, where blue is VIL level 1 and 2, light blue is 3, yellow is 4, and red is 5 and 6.VIL level 1 and above indicate the presence of precipitations and were used to identify the clouds coverage regions.In addition, pilots would generally penetrate VIL level less than 3 and avoid regions with VIL level 3 and above. 19IWS also provides echo tops information, which indicate the storm heights.Echo tops were used to provide three dimensional weather information.It is assumed that the storm is having impacts from ground level up to the height of the storm tops.
+B. Contrail ModelContrails are vapor trails caused by aircraft operating at high altitudes under certain atmospheric conditions.The contrail model in this paper uses atmospheric temperature and humidity data retrieved from the Rapid Updated Cycle (RUC) data, provided by the National Oceanic and Atmospheric Administration (NOAA).The horizontal resolution in RUC is 13-km.RUC data has 37 vertical isobaric pressure levels ranging between 100 and 1000 millibar (mb) in 25 mb increments.Since the vertical isobaric pressure levels do not correspond with 2,000 feet increments, linear interpolation was used to convert the RUC data to a vertical range from 26,000 feet to 44,000 feet with an increment of 2,000 feet.The range is chosen because it generally is too warm to form contrails below 26,000 feet and most aircraft fly below 44,000 feet.Contrails form when a mixture of warm engine exhaust gases and cold ambient air reaches saturation with respect to water, forming liquid drops which quickly freeze.Contrails form in the regions of airspace that have ambient Relative Humidity with respect to Water (RHw) greater than a critical value r contr . 20Regions with RHw greater than or equal to 100% are excluded because clouds are already formed. 21Contrails can persist when the environmental Relative Humidity with respect to Ice (RHi) is greater than 100%. 22In this paper, contrail favorable regions are defined as the regions of airspace that have r contr ≤ RHw < 100% and RHi ≥ 100%.The estimated critical relative humidity for contrails formation at a given temperature T (in Celsius) can be calculated asr contr = G(T -T contr ) + e liq sat (T contr ) e liq sat (T ) ,(1)where e liq sat (T ) is the saturation vapor pressure over water at a given temperature.The estimated threshold temperature for contrails formation at liquid saturation isT contr = -46.46 + 9.43ln(G -0.053) + 0.72ln 2 (G -0.053),(2)whereG = EI H2O C p P Q(1 -η) ,(3)EI H2O is the emission index of water vapor (assumed to be 1.25); C p = 1004 (in JKg -1 K -1 ) is the isobaric heat capacity of air, P (in Pa) is the ambient air pressure, = 0.6222 is the ratio of molecular masses of water and dry air, Q = 43 × 10 6 (in JKg -1 ) is the specific combustion heat, and η = 0.3 is the average propulsion efficiency of the jet engine.The value of r contr is computed by Eq (1)-(3) using RUC measurements for RHw and temperatures.RHi is calculated by temperature and relative humidity using the following formula:RHi = RHw × 6.0612e 18.102T /(249.52+T ) 6.1162e 22.577T /(237.78+T ) , (4)where T is the temperature in Celsius. Figure 1b shows the contrail favorable regions at 8AM EDT on April 23, 2010 at an altitude of 34,000 feet.
+III. AnalysisContrail favorable regions can be computed and predicted using the weather forecasts.One might ask what if there are already clouds in regions favorable for contrail formation.In that case, it does not matter if contrails are formed or not since the incoming and outgoing radiations are reduced anyway.That is, aircraft should be able to fly through these regions without causing negative environmental impact.On the other hand, if the weather condition is too severe to safely fly through, aircraft should avoid the regions even though doing so would reduce the contrail formations.In order to take into account these effects, precipitation prediction is used to modify the contrail models.To observe the correlations between contrail favorable and precipitation (VIL level 1 and above) regions, the two regions in Fig. 1 were overlapped and shown in Fig. 2. In the figure, the green polygons indicates the regions of CIWS precipitation and the grey contours are the contrail favorable regions over twenty U.S. air traffic control centers.As shown in the figure, the locations of contrail favorable and CIWS precipitation are correlated.In fact, there are 48% of CIWS precipitation regions overlapped with contrail favorable regions.Therefore, the effect should not be ignored.The average ratios of CIWS precipitation overlapped with contrail favorable regions at different altitudes over the year of 2010 are shown in Fig. 4. The coverage ratios vary with altitudes.The maximum coverage ratio at each month ranges from 19.0% (November, 34,000 feet) to 45% (January, 34,000 feet).It is noted that in summer the coverage ratios are larger at higher altitude (July and August at 40,000 feet), while the ratios are higher at lower altitude in winter (January and December at 34,000 feet).The observation can be interpreted as contrail favorable regions are larger at higher altitude in summer because of the warm temperature at lower altitude.In July and August, there are no contrail favorable regions formed at 26,000 and 28,000 feet.Since there are overlapped regions between the two that can not be ignored, the development of contrail formation analysis and reduction strategies should take into account the effect of precipitation.Another observation is that the contrail reduction strategies should be more efficient in winter than in summer since the contrail favorable regions in summer happen more frequently at high altitude (40,000 feet and above) where fewer aircraft are flying.The contrail formation analysis and reduction strategies in the presence of different weather conditions are described in the next section.
+IV. Methodologies
+A. Contrail Frequency IndexThis paper modifies the contrail frequency index (CFI) defined in Ref. 17 by adding the component of cloud coverage information and using 13km RUC data instead of 40km RUC data.The altitude level index, l, is defined as l = 1 . . . 10 corresponding to altitudes of 26, 000, 30, 000, . . ., 44, 000 feet.The potential persistent contrail formation matrix (contrail matrix) at time t at level l is defined asR l t =     r l 1,1,t r l 1,where r l i,j,t is 1 if r contr ≤ RHw < 100% and RHi ≥ 100% at grid (i, j), and 0 if not.The Center contrail frequency indices of twenty U.S. air traffic control centers at time t at level l are defined asCF I center,l,t = 337 i=1 451 j=1 r l i,j,t a l i,j,t c i,j ,(6)where a l i,j,t is the number of aircraft within RUC 13km grid (i, j) flying closest to altitude level l at time t, and c i,j is 1 when grid (i, j) is inside the center and 0 if not.The twenty U.S. air traffic control centers are listed in Table 1.Contrail frequency index is used to quantify the severity of contrail activities.For planning for contrail reduction stretegies, traffic flow managers need to know the potential high contrail regions in the next few hours.Therefore predicted contrail frequency index is needed for contrail reduction strategies.Similar to the concept of Weather Impacted Traffic Index (WITI) introduced by Callaham et al. 23 and Sridhar, 24 and the three-dimensional index derived by Chen, 15 predicted contrail frequency index was defined as a convolution of predicted traffic data and forecast of atmospheric conditions.The index consists of the RUC forecast data and the predicted aircraft locations when t is a future time.The Center contrail frequency index can then be rewritten asCF I center,l,t =    337 i=1 451 j=1 r l i,j,t a l i,j,t c i,j if t <= t now , 337i=1451 j=1 rl i,j,t âl i,j,t c i,j if t > t now ,(7)where t now is the current time, rl i,j is defined in Eq. ( 5) with RUC forecast data, and âl i,j is the predicted number of aircraft within RUC 13km grid (i, j) flying closest to altitude level l at time t.Based on the assumption that there are already clouds in regions with VIL level 1 and above, CFI should be discounted when the contrail favorable and CIWS VIL level 1 and above regions overlapped.However, regions with VIL level 3 and above are consider severe storms and pilots would try to avoid those regions.Contrail favorable regions are considered as soft constraint areas, meaning it is fine to fly through, but better to avoid for reducing the environmental impact.Severe weather regions are considered as hard constraint areas that pilots would try to avoid.In order to incorporate these constraints in the contrail reduction strategies, the VIL level matrix that maps the CIWS data to 13km RUC grid is defined asW l t =     w l 1,    ,(8)where w l i,j,t is the CIWS VIL level if the echo top at grid (i, j) is higher than altitude level l at time t.w l i,j,t is zero if the echo top at grid (i, j) is lower than altitude lvl l.The r i,j,t , l in Eq (5) will be set to zero if w l i,j,t is 1 or 2. This matrix will later be used to evaluate the feasibility of the contrail reduction strategies.As an example, the Center contrail frequency indices with and without weather information at 34,000 feet at 8AM EDT on April 23, 2010 were computed and are shown in Fig. 5.The blue bars are the CFI without weather information.When the contrail favorable regions are covered by CIWS VIL level 1 or 2, the CFI was discounted, as represented by the green bars.For ZMP, ZKC, and ZAU centers, the CFIs without weather information are 212, 148 and 237 respectively, while the CFIs without weather information are 188, 98 and 174 respectively.The CFI differences in these centers are more significant as there are more regions with VIL 1 and 2 overlapped with contrail favorable regions (see Fig. 2).Figure 6 shows the aircraft trajectories and contrail formations in Kansas City Center.In the figures, the grey polygon indicates the contrail favorable regions.Blue dots are the flying aircraft between 33,001 feet and 35000 feet.When the aircraft enter the grey polygon, contrails would form, as indicated by grey dot.The number of grey dots in Fig. 6a are the CFI without weather information, which is 148.The green polygon in Fig. 6b are the regions with CIWS VIL level 1 and 2. When an aircraft flies through a grey polygon (contrail favorable regions) overlapped with a green polygon, the aircraft is already covered by clouds therefore it should not be counted toward CFI.There are total of 50 grey dots that are inside green polygons, bringing down the CFI with weather information to 98.In order to determine if it is feasible to perform the contrail reduction strategies, the Weather Severity Index (WSI) is defined as the number of aircraft that would be impacted by the severe weather in a future time, formulated asW SI center,l,t = 337 i=1 451 j=1 ŵl i,j,t âl i,j,t ,(9)where ŵl i,j is defined in Eq. ( 8) with CIWS forecast data, and âl i,j predicted number of aircraft within RUC 13km grid (i, j) flying closest to altitude level l at time t.Note that the WSI is the same concept with WITI in Ref. 24 and is used to determine the feasibility of applying the proposed contrail reduction strategies.
+B. Contrail Reduction StrategiesThis paper extends the contrail reduction strategies developed in Ref. 13  and 2, as described in the previous subsection.Second is to use the weather severity index as an indicator to decide whether the strategy is feasible or not.Contrail frequency indices are used to quantify the severity of contrail formation.The strategy for reducing the persistent contrail formations is to minimize contrail frequency index by altering the aircraft's cruising altitude.Assume the aircraft at altitude level l at a center are made to fly a different level l .The contrail frequency index changes toCF I l center,l,t = 337 i=1 451 j=1 r l i,j,t a l i,j,t c i,j ,(10)A contrail frequency index matrix is formed asCFI center,t =       CF I 1 1,t CF Iwhere the diagonal term CF I l l,t is the contrail frequency index at level l before contrail reduction, and CF I l l,t is the contrail frequency index when guiding aircraft at level l to aircraft at level l .The contrail reduction is ∆CF I l l,t = CF I l l,t -CF I l l,t .Note that when l > l, not all aircraft have the ability to fly from level l to level l .If altitude level l is higher than an aircraft's maximal flight altitude, it stays at level l and is not counted in CF I l l,t .In addition, if an aircraft crosses a sector boundary and causes congestion, it stays at level l and does not add to CF I l l,t .The strategy is to find the smallest element in each column of CFI center,t .If the aircraft are limited to alter ∆l levels, the solution is the smallest element in [CF I l-∆l l,t . . .CF I l l,t . . .CF I l+∆l l,t] T in each column.The solution is denoted as [l 1 . . .l 11 ].Each l i means aircraft at flight level i is flying at level l i .If l i = i, the aircraft at level i did not alter.In order to determine whether the contrail reduction strategy is feasible in the presence of severe weather, a weather severity matrix is used, defined asWSI center,t =       W SI 1 1,t W SIThe increase of W SI if aircraft flying at level l are altered to level l is ∆W SI l l,t = W SI l l,t -W SI l l,t .The value helps to determine if it is feasible to alter the aircraft cruising altitude from level l to level l .If ∆W SI l l,t ≤ 0,that means it is even safer to fly at level l than at l, therefore the contrail reduction move can be applied.If ∆W SI l l,t > ε, where ε is a threshold value, the contrail reduction move might not be feasible because of the increase of weather impact.Further analysis is needed to determine the value of ε.As an example, the CFI matrix for Kansas City Center at 8AM EDT on April 23, 2010 was computed,CFI ZKC =                    0 0 0 × × × × × × × 0 0 0 0 × × × × × × 0 0 0 0 0 × × × × × × 0 0 0 0 0 × × × × × × 29 47 98 230 100 × × × × × × 17 51 124 39 32 × × × × × × 39 101 23 16 0 × × × × × × 91 18 15 0 0 × × × × × × 14 6 0 0 × × × × × × × 7 0 0                    . (16)Assume that the aircraft are allowed to move one level (2,000 feet) up or down to reduce contrails formation.All the aircraft between 33,001 feet and 35000 feet (level 5) have the total CFI of 98 (CFI ZKC (5, 5) = 98).Moving the aircraft to level 4 will result in 0 in CFI (CFI ZKC (4, 5) = 0), a 98 reduction.Other contrail reduction moves include moving aircraft from level 6 to 7, 7 to 8 and 8 to 9. The solution is expressed as [1 1 1 1 4 7 8 9 9 10], resulting in a contrail reduction from 260 to 125, a 51.9% reduction.If the aircraft are allowed to move two levels up or down, it can achieve more reduction.The moves include moving aircraft from level 5 to 4, 6 to 4, 7 to 9 and 8 to 9. The solution is expressed as [1 1 1 1 4 4 9 9 9 10], resulting in a contrail reduction from 260 to 20, a 92.3% reduction.To evaluate the impact of severe weather, WSI matrix was computed,WSI ZKC =                    1 2 3 × × × × × × × 1 2 3 3 × × × × × × 1 2 3 3 0 × × × × × × 2 3 3 0 12 × × × × × × 3 2 0 4 1 × × × × × × 2 0 4 1 0 × × × × × × 0 0 1 0 0 × × × × × × 0 1 0 0 0 × × × × × × 1 0 0 0 × × × × × × × 0 0 0                    . (17)For the case that allows aircraft to move two level up or down, the move from level 6 to level 4 results in a CFI reduction of 124 with WSI increased by 8 (WSI ZKC (4, 6) -WSI ZKC (6, 6) = 8).If the threshold ε is set to 10, the move is still feasible even with an increase of weather impact.If the threshold ε is set to 0, the move is not feasible.Instead, the aircraft in level 6 will be moved to level 8, resulting in a CFI reduction from 124 to 91 and a WSI reduction of 4. The contrail reduction strategies described in this section incorporates the weather data so that they are feasible in the presence of severe weather.Data from 24-hour period on April 23, 2010 were analyzed.The Center contrail frequency indices at all altitudes, with and without CIWS percipication information, are shown in Fig. 7.For the day, the difference of CFI with and without CIWS precipitation information among all centers are 14%.As shown in the figure, it has more impact in the centers including Denver (ZDV), Minneapolis (ZMP), Kansas City (ZKC), Chicago (ZAU) and Memphis Center (ZME).The effect of cloud coverage on computing contrail frequency index is significant and should not be ignored.Next, the contrail reduction strategies were applied using the CFI with weather information and WSI threshold ε = 0 .The results are shown in Fig. 8.The center CFIs with weather information before reduction are shown in green bars.When the aircraft are allowed to alter 2,000 feet, the center CFIs after reduction are shown in light green bars.The total reduction among all centers is 63.7%.When the aircraft are allowed to alter 4,000 feet, the total reduction is 92.6%.
+V. ConclusionsThis paper describes contrail reduction strategies using different weather resources.Two weather products were used in this paper, Rapid Update Cycle for the wind, temperature and humidity forecast and Corridor Integrated Weather System for the weather forecast.The two weather products are used in a complementary way to develop contrail reduction strategies.Analysis shows that the contrail favorable regions are correlated with regions with precipitation.During the month of April 2010, there are 53% of CIWS VIL level 1 and above overlapped with contrail favorable regions.For the day tested, adding precipitation weather information reduces the contrail frequency index by 14%.The strategies use the concept of contrail frequency index with weather information to achieve maxima contrail reduction while avoiding severe weather regions.The contrail reductions are achieved by changing the aircraft pre-departure cruising altitudes that would form less contrails.For the day tested, the strategies achieve a 63.7% contrail reduction taking into account the weather impact if the aircraft are allowed to change their cruising altitude 2,000 feet up or down.When the aircraft are allowed to move 4,000 feet up or down, it has a 92.6% reduction.The proposed contrail reduction strategies are suitable for reducing contrail formation in the entire airspace with the presence of different weather conditions.Figure 1 .1Figure 1.CIWS weather and contrail regions at 8AM EDT on April 23, 2010.
+Figure 2 .2Figure 2. Contrail favorable and CIWS precipitation regions overlapped at 8AM EDT on April 23, 2010.
+Figure 33Figure3shows the correlations between CIWS precipitation and contrail favorable regions in April 2010.The blue line indicates the size of the CIWS precipitation, defined as the number of grid points in 13km RUC 451 × 337 grid with CIWS VIL level 1 and above.The green line is the number of grid points that are overlapped with contrail favorable regions.During the month, there are 53% of CIWS VIL level 1 and above overlapped with contrail favorable regions.The yearly data of 2010 have the similar trend.
+Figure 3 .3Figure 3. CIWS precipitation regions overlapped with contrail favorable regions in April 2010.
+Figure 4 .4Figure 4. Overlap ratio between CIWS precipitation and contrail favorable regions in 2010.
+Figure 6 .6Figure 6.CIWS precipitation and contrail regions at 8AM EDT on April 23, 2010.
+Figure 7 .7Figure 7.Total contrail frequency index with and without weather information on April 23, 2010.
+Figure 8 .8Figure 8. Results of contrail reduction strategies on April 23, 2010.
+Table 1 .1Center index of twenty continental U.S. air traffic control centers.IndexNameIndexName1Seattle Center (ZSE)11Chicago Center (ZAU)2Oakland Center (ZOA)12Indianapolis Center (ZID)3Los Angeles Center (ZLA)13Memphis Center (ZME)4Salt Lake City Center (ZLC)14Cleveland Center (ZOB)5Denver Center (ZDV)15Washington D. C. Center (ZDC)6Albuquerque Center (ZAB)16Atlanta Center (ZTL)7Minneapolis Center (ZMP)17Jacksonville Center (ZJX)8Kansas City Center (ZKC)18Miami Center (ZMA)9Dallas/Fort Worth Center (ZFW)19Boston Center (ZBW)10Houston Center (ZHU)20New York Center (ZNY)
+			of 11 American Institute of Aeronautics and Astronautics
+		
+		
+			
+
+				
+
+
+	
+		
+			IWaitz
+		
+		
+			JTownsend
+		
+		
+			JCutcher-Gershenfeld
+		
+		
+			EGreitzer
+		
+		
+			JKerrebrock
+		
+		Report to the United States Congress: Aviation and the Environment, A National Vision, Framework for Goals and Recommended Actions
+		London, UK
+		
+			December 2004
+		
+	
+	Tech. rep
+	Partnership for AiR Transportation Noise and Emissions Reduction
+	Waitz, I., Townsend, J., Cutcher-Gershenfeld, J., Greitzer, E., and Kerrebrock, J., "Report to the United States Congress: Aviation and the Environment, A National Vision, Framework for Goals and Recommended Actions," Tech. rep., Partnership for AiR Transportation Noise and Emissions Reduction, London, UK, December 2004.
+
+
+
+	
+		Über Bedingungen zur Bildung von Kondensstreifen aus Flugzeugabgasen
+		
+			UlrichSchumann
+		
+		10.1127/metz/5/1996/4
+	
+	
+		Meteorologische Zeitschrift
+		metz
+		0941-2948
+		
+			5
+			1
+			
+			1996
+			Schweizerbart
+		
+	
+	Schumann, U., "On Conditions for Contrail Formation from Aircraft Exhausts," Meteorologische Zeitschrift, Vol. 5, No. 1, 1996, pp. 4-23.
+
+
+
+	
+		Future Development of Contrail Cover, Optical Depth, and Radiative Forcing: Impacts of Increasing Air Traffic and Climate Change
+		
+			SMarquart
+		
+		
+			MPonater
+		
+		
+			FMager
+		
+		
+			RSausen
+		
+		10.1175/1520-0442(2003)016<2890:fdocco>2.0.co;2
+	
+	
+		Journal of Climate
+		J. Climate
+		0894-8755
+		1520-0442
+		
+			16
+			17
+			
+			September 2003
+			American Meteorological Society
+		
+	
+	Marquart, S., Ponater, M., Mager, F., and Sausen, R., "Future Development of Contrail Cover, Optical Depth, and Radiative Forcing: Impacts of Increasing Air Traffic and Climate Change," Journal of Climate, Vol. 16, September 2003, pp. 2890-2904.
+
+
+
+	
+		A Standing Royal Commission
+		
+			SusanOwens
+		
+		10.1093/acprof:oso/9780198294658.003.0003
+	
+	
+		Knowledge, Policy, and Expertise
+		London, UK
+		
+			Oxford University Press
+			2002
+			
+		
+	
+	"The Environmental Effects of Civil Aircraft in Flight," Tech. rep., Royal Commission on Environmental Pollution, London, UK, 2002.
+
+
+
+	
+		Aircraft induced contrail cirrus over Europe
+		
+			HermannMannstein
+		
+		
+			UlrichSchumann
+		
+		10.1127/0941-2948/2005/0058
+	
+	
+		Meteorologische Zeitschrift
+		metz
+		0941-2948
+		
+			14
+			4
+			
+			2005
+			Schweizerbart
+		
+	
+	Mannstein, H. and Schumann, U., "Aircraft induced contrail cirrus over Europe," Meteorologische Zeitschrift, Vol. 14, No. 4, 2005, pp. 549-554.
+
+
+
+	
+		A Review of Various Strategies for Contrail Avoidance
+		
+			KlausGierens
+		
+		
+			LingLim
+		
+		
+			KostasEleftheratos
+		
+		10.2174/1874282300802010001
+	
+	
+		The Open Atmospheric Science Journal
+		TOASCJ
+		1874-2823
+		
+			2
+			1
+			
+			2008
+			Bentham Science Publishers Ltd.
+		
+	
+	The Open Atmospheric
+	Gierens, K., Limb, L., and Eleftheratos, K., "A Review of Various Strategies for Contrail Avoidance," The Open Atmo- spheric Science Journal, Vol. 2, 2008, pp. 1-7.
+
+
+
+	
+		Overview on Contrail and Cirrus Cloud Avoidance Technology
+		
+			FNoppel
+		
+		
+			RSingh
+		
+		10.2514/1.28655
+	
+	
+		Journal of Aircraft
+		Journal of Aircraft
+		0021-8669
+		1533-3868
+		
+			44
+			5
+			
+			2007
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	Noppel., F. and Singh, R., "Overview on Contrail and Cirrus Cloud Avoidance Technology," Journal of Aircraft, Vol. 44, No. 5, 2007, pp. 1721-1726.
+
+
+
+	
+		A note on how to avoid contrail cirrus
+		
+			HermannMannstein
+		
+		
+			PeterSpichtinger
+		
+		
+			KlausGierens
+		
+		10.1016/j.trd.2005.04.012
+	
+	
+		Transportation Research Part D: Transport and Environment
+		Transportation Research Part D: Transport and Environment
+		1361-9209
+		
+			10
+			5
+			
+			September 2005
+			Elsevier BV
+		
+	
+	Mannstein, H., Spichtinger, P., and Gierens, K., "A note on how to avoid contrail cirrus," Transportation Research. Part D, Transport and environment, Vol. 10, No. 5, September 2005, pp. 421-426.
+
+
+
+	
+		An Optimal Strategy for Persistent Contrail Avoidance
+		
+			ScotCampbell
+		
+		
+			NatashaNeogi
+		
+		
+			MichaelBragg
+		
+		10.2514/6.2008-6515
+		AIAA-2008-6515
+	
+	
+		AIAA Guidance, Navigation and Control Conference and Exhibit
+		Honolulu, HI
+		
+			American Institute of Aeronautics and Astronautics
+			August 2008
+		
+	
+	Campbell1, S. E., Neogi, N. A., and Bragg, M. B., "An Optimal Strategy for Persistent Contrail Avoidance," AIAA Guidance, Navigation and Control Conference, AIAA-2008-6515, AIAA, Honolulu, HI, August 2008.
+
+
+
+	
+		The impact of cruise altitude on contrails and related radiative forcing
+		
+			ChristineFichter
+		
+		
+			SusanneMarquart
+		
+		
+			RobertSausen
+		
+		
+			DavidSLee
+		
+		10.1127/0941-2948/2005/0048
+	
+	
+		Meteorologische Zeitschrift
+		metz
+		0941-2948
+		
+			14
+			4
+			
+			August 2005
+			Schweizerbart
+		
+	
+	Fichter, C., Marquart, S., Sausen, R., and Lee, D. S., "The impact of cruise altitude on contrails and related radiative forcing," Meteorologische Zeitschrift, Vol. 14, No. 4, August 2005, pp. 563-572.
+
+
+
+	
+		Reducing the climate change impacts of aviation by restricting cruise altitudes
+		
+			VictoriaWilliams
+		
+		
+			RobertBNoland
+		
+		
+			RalfToumi
+		
+		10.1016/s1361-9209(02)00013-5
+	
+	
+		Transportation Research Part D: Transport and Environment
+		Transportation Research Part D: Transport and Environment
+		1361-9209
+		
+			7
+			6
+			
+			November 2002
+			Elsevier BV
+		
+	
+	Williams, V., Noland, R. B., and Toumi, R., "Reducing the climate change impacts of aviation by restricting cruise altitudes," Transportation Research. Part D, Transport and environment, Vol. 7, No. 5, November 2002, pp. 451-464.
+
+
+
+	
+		Variability of contrail formation conditions and the implications for policies to reduce the climate impacts of aviation
+		
+			VictoriaWilliams
+		
+		
+			RobertBNoland
+		
+		10.1016/j.trd.2005.04.003
+	
+	
+		Transportation Research Part D: Transport and Environment
+		Transportation Research Part D: Transport and Environment
+		1361-9209
+		
+			10
+			4
+			
+			July 2005
+			Elsevier BV
+		
+	
+	Williams, V. and Noland, R. B., "Variability of contrail formation conditions and the implications for policies to reduce the climate impacts of aviation," Transportation Research. Part D, Transport and environment, Vol. 10, No. 4, July 2005, pp. 269-280.
+
+
+
+	
+		Fuel efficient strategies for reducing contrail formations in United States airspace
+		
+			BanavarSridhar
+		
+		
+			NeilYChen
+		
+		10.1109/dasc.2010.5655533
+	
+	
+		29th Digital Avionics Systems Conference
+		Salt Lake City, UT
+		
+			IEEE
+			October 2010
+		
+	
+	Sridhar, B., Chen, N. Y., and Ng, H. K., "Fuel Efficient Strategies for Reducing Contrail Formations in United State National Air Space," 29th Digital Avionics Systems Conference, Salt Lake City, UT, October 2010.
+
+
+
+	
+		Aircraft Trajectory Optimization and Contrails Avoidance in the Presence of Winds
+		
+			BanavarSridhar
+		
+		
+			HokNg
+		
+		
+			NeilChen
+		
+		10.2514/6.2010-9139
+		AIAA-2010-9139
+	
+	
+		10th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference
+		Fort Worth, TX
+		
+			American Institute of Aeronautics and Astronautics
+			September 2010
+		
+	
+	Sridhar, B., Ng, H. K., and Chen, N. Y., "Aircraft Trajectory Optimization and Contrails Avoidance in the Presence of Winds," The 10th AIAA Aviation Technology, Integration, and Operations Conference, AIAA-2010-9139, AIAA, Fort Worth, TX, September 2010.
+
+
+
+	
+		Estimation of Air Traffic Delay Using Three Dimensional Weather Information
+		
+			NeilChen
+		
+		
+			BanavarSridhar
+		
+		10.2514/6.2008-8916
+	
+	
+		The 26th Congress of ICAS and 8th AIAA ATIO
+		Anchrorage, AK
+		
+			American Institute of Aeronautics and Astronautics
+			September 2008
+		
+	
+	Chen, N. Y. and Sridhar, B., "Estimation of Air Traffic Delay Using Three Dimensional Weather Information," 8th AIAA Aviation Technology, Integration and Operations Conference (ATIO), Anchrorage, AK, September 2008.
+
+
+
+	
+		Modeling Convective Weather Avoidance in Enroute Airspace
+		
+			RADelaura
+		
+		
+			MRobinson
+		
+		
+			MLPawlak
+		
+		
+			JEEvans
+		
+	
+	
+		13th Conference on Aviation, Range, and Aerospace Meteorology (ARAM)
+		New Orleans, LA
+		
+			2008
+			17
+		
+	
+	DeLaura, R. A., Robinson, M., Pawlak, M. L., and Evans, J. E., "Modeling Convective Weather Avoidance in Enroute Airspace," 13th Conference on Aviation, Range, and Aerospace Meteorology (ARAM), New Orleans, LA, 2008. 17
+
+
+
+	
+		Prediction and Use of Contrail Frequency Index for Contrail Reduction Strategies
+		
+			NeilChen
+		
+		
+			BanavarSridhar
+		
+		
+			HokNg
+		
+		10.2514/6.2010-7849
+	
+	
+		AIAA Guidance, Navigation, and Control Conference
+		Toronto, Ontario
+		
+			American Institute of Aeronautics and Astronautics
+			August 2010
+		
+	
+	Chen, N. Y., Sridhar, B., and Ng, H. K., "Prediction and Use of Contrail Frequency Index for Contrail Reduction Strategies," AIAA Guidance, Navigation, and Control Conference, Toronto, Ontario, August 2010.
+
+
+
+	
+		Description of the Corridor Integrated Weather System (CIWS) Weather Products
+		
+			JEvans
+		
+		
+			DKlingle-Wilson
+		
+		ATC-317
+		
+			Aug 2005
+		
+		
+			MIT Lincoln Laboratory
+		
+	
+	Project Report
+	Evans, J. and Klingle-Wilson, D., "Description of the Corridor Integrated Weather System (CIWS) Weather Products," Project Report ATC-317, MIT Lincoln Laboratory, Aug 2005.
+
+
+
+	
+		An Exploratory Study of Modeling Enroute Pilot Convective Storm Flight Deviation Behavior
+		
+			RDelaura
+		
+		
+			JEvans
+		
+	
+	
+		12th Conference on Aviation, Range, and Aerospace Meteorology (ARAM)
+		Atlanta, GA
+		
+			2006
+		
+	
+	DeLaura, R. and Evans, J., "An Exploratory Study of Modeling Enroute Pilot Convective Storm Flight Deviation Behavior,," 12th Conference on Aviation, Range, and Aerospace Meteorology (ARAM), Atlanta, GA, 2006.
+
+
+
+	
+		Contrails in a comprehensive global climate model: Parameterization and radiative forcing results
+		
+			MichaelPonater
+		
+		
+			SMarquart
+		
+		
+			RSausen
+		
+		10.1029/2001jd000429
+	
+	
+		Journal of Geophysical Research
+		J. Geophys. Res.
+		0148-0227
+		
+			107
+			D13
+			
+			2002
+			American Geophysical Union (AGU)
+		
+	
+	Ponater, M., Marquart, S., and Sausen, R., "Contrails in a Comprehensive Global Climate Model: Parameterization and Radiative Forcing Results," Journal of Geophysical Research, Vol. 107, No. D13, 2002, pp. ACL 2-1.
+
+
+
+	
+		Determination of humidity and temperature fluctuations based on MOZAIC data and parametrisation of persistent contrail coverage for general circulation models
+		
+			KMGierens
+		
+		
+			USchumann
+		
+		
+			HG JSmit
+		
+		
+			MHelten
+		
+		
+			GZängl
+		
+		10.1007/s00585-997-1057-3
+	
+	
+		Annales Geophysicae
+		Ann. Geophys.
+		1432-0576
+		
+			15
+			8
+			
+			1997
+			Copernicus GmbH
+		
+	
+	Gierens, K. M., Schumann, U., Smit, H. G. J., Helten, M., and Zangl1, G., "Determination of humidity and temperature fluctuations based on MOZAIC data and parametrisation of persistent contrail coverage for general circulation models," Annales Geophysicae, Vol. 15, 1997, pp. 1057-1066.
+
+
+
+	
+		Estimated contrail frequency and coverage over the contiguous United States from numerical weather prediction analyses and flight track data
+		
+			DavidPDuda
+		
+		
+			PatrickMinnis
+		
+		
+			RabindraPalikonda
+		
+		10.1127/0941-2948/2005/0050
+	
+	
+		Meteorologische Zeitschrift
+		metz
+		0941-2948
+		
+			14
+			4
+			
+			June-July 2003
+			Schweizerbart
+			Friedrichshafen at Lake Constance, Germany
+		
+	
+	Duda, D. P., Minnis, P., Costulis, P. K., and Palikonda, R., "CONUS Contrail Frequency Estimated from RUC and Flight Track Data," European Conference on Aviation, Atmosphere, and Climate, Friedrichshafen at Lake Constance, Germany, June- July 2003.
+
+
+
+	
+		Assessing NAS Performance: Normalizing for the Effects of Weather
+		
+			MBCallaham
+		
+		
+			JSDearmon
+		
+		
+			ACooper
+		
+		
+			JHGoodfriend
+		
+		
+			DMoch-Mooney
+		
+		
+			GSolomos
+		
+	
+	
+		4th USA/Europe Air Traffic Management R&D Symposium
+		Santa Fe, NM
+		
+			December 2001
+		
+	
+	Callaham, M. B., DeArmon, J. S., Cooper, A., Goodfriend, J. H., Moch-Mooney, D., and Solomos, G., "Assessing NAS Performance: Normalizing for the Effects of Weather," 4th USA/Europe Air Traffic Management R&D Symposium, Santa Fe, NM, December 2001.
+
+
+
+	
+		Relationship Between Weather, Traffic and Delay Based on Empirical Methods
+		
+			BanavarSridhar
+		
+		
+			SeanSwei
+		
+		10.2514/6.2006-7760
+	
+	
+		6th AIAA Aviation Technology, Integration and Operations Conference (ATIO)
+		Wichita, KS
+		
+			American Institute of Aeronautics and Astronautics
+			September 2006
+		
+	
+	Sridhar, B. and Swei, S., "Relationship between Weather, Traffic and Delay Based on Empirical Methods," 6th AIAA Aviation Technology, Integration and Operations Conference (ATIO), Wichita, KS, September 2006.
+
+
+				
+			
+		
+	

file141.txt

@@ -0,0 +1,600 @@
+
+	
+	
+		
+I. INTRODUCTIONThis paper describes NASA's ATD-2 Phase 3 prototype capability, its initial demonstration in the North Texas, and its expected benefits.The Phase 3 concept expands on Phases 1 and 2 of the Integrated Arrival, Departure, and Surface (IADS) traffic management system and incorporates the use of TOS to identify alternative departure routes when flights are predicted to incur surface delay due to demand/capacity imbalances.
+A. BackgroundThe National Aeronautics and Space Administration (NASA) has collaborated with the Federal Aviation Administration (FAA) and industry partners to investigate and test IADS technologies over the last decade [1,2,3,4,5].For Phases 1 and 2 of the ATD-2 field demonstration, NASA developed a prototype Surface Metering Program (SMP) for the FAA's Terminal Flight Data Manager (TFDM) [6,7].This system has been used since 2018 by the American Airlines ramp and Air Traffic Control (ATC) Tower personnel at Charlotte Douglas International Airport (KCLT) with positive results [8].From November, 29 2017 to April, 30 2020, 25,748 departures (3.8% of departures) were held at the gate for an average of 5.9 minutes.Gate holds were estimated to have saved about 2,883,410 pounds of fuel and reduced CO2 emissions by 8,880,901 pounds, the equivalent of 66,038 urban trees [9].The maturation of the Phase 2 system which enables a strategic SMP [10], was based on a series of human-in-the-loop simulations [11], users' feedback [12,13], as well as refinements of the scheduler's takeoff, pushback and taxi times [14].The technologies and lessons learned have been transferred to the FAA.The FAA plans to replace the ATD-2 IADS system in CLT with TFDM by late 2021 and has started the process of developing new requirements for its future implementations.The goal for Phase 3 was to develop and operationally test a capability to support the management of departure demand/capacity imbalances in a multi-airport airspace.The Dallas-Fort Worth TRACON (D10) metroplex was chosen due to its high traffic demand and frequent delays due to inclement weather.This airspace includes two major hub airports, Dallas Love Field (KDAL) and Dallas-Fort Worth (KDFW), where both Southwest Airlines and American Airlines, NASA's ATD-2 field demonstration partners, have a large presence.The desire to incorporate TOS operations in the ATD-2 Phase 3 capability and evaluation stemmed from both industry partners' support and early analyses.In recent years, the concept of TOS has gained recognition by the FAA and the airline industry as a potentially useful resource to help ATC manage flowconstrained areas.Use cases of TOS outside of the Collaborative Trajectory Option Program (CTOP) were investigated by The MITRE Corporation, with the FAA and Collaborative Decision Making (CDM) groups [15].In addition, NASA's early benefit analysis indicated that rerouting departure flights on alternative routes, when deemed beneficial, may result in delay reductions for both rerouted flights and subsequent flights.The initial concept received strong support from field demonstration partners, the airline industry, the CDM Flow Evaluation Team (FET), the Surface CDM Team (SCT) group, and the FAA NextGen organization (FAA/ANG).
+B. Trajectory Option Set (TOS)In current-day operations, Flight Operators (FOs) can file only one flight plan per flight.While airline dispatchers may consider various routes before filing a flight plan, ATC assume that pilots will fly the filed route.This assumption is reasonable when flights are not subjected to surface or airborne delay.When demand/capacity imbalances occur, however, both the FOs and the ATCs may be looking for opportunities to offload demand to reduce congestion and flight delay.A TOS consists of a set of preferred routes, each of which has an associated Relative Trajectory Cost (RTC).The RTC is a weighted time that accounts for the cost of flying additional miles.Adding alternative TOS routes to the filed plans may provide opportunities to compare routes and determine when flights could depart or arrive earlier and benefit from a reroute.This is analogous to when drivers look up alternative routes on a map application to assess if other routes would enable them to arrive at their destination sooner.This determination requires the ability to predict when benefits (delay savings) outweigh the cost (RTC) of flying an alternative versus the filed route (net savings) in real-time.This is, in essence, what the Phase 3 capability attempts to provide for departure flights to both the FOs and ATC in a tactical manner.The concept of TOS has been integrated in FAA's Traffic Flow Management System (TFMS).It is a key component of the CTOP to help ATC to strategically manage flows in constrained areas.CTOP analyzes demand at a constrained area and determines when flights could be routed outside of constrained areas based on TOS submitted no later than 45 min prior to departure.Flights that cannot be rerouted would receive an Expect Departure Clearance Time (EDCT).TOS has also been integrated in Pre-Departure ReRoute (PDRR) and Airborne ReRoute (ABRR) capabilities.Both PDRR and ABRR would enable Traffic Management Coordinators (TMCs) to amend flight plans, based on submitted TOS, via the Route Amendment Dialog (RAD).However, to this date, CTOP has been mainly used for analysis purposes, and TOS has not been used operationally.There are three types of potential TOS scenario that we considered for the development of Phase 3:1) Departure Demand/Capacity Imbalances: Excess demand and/or reduced capacity may result in surface departure delay for the flight's filed route, but not necessarily for alternative routes.Under normal circumstances, the D10 airspace has enough route capacity to absorb the departure demand from both major airports and the surrounding airports.D10 has four terminal gates (North, East, South, and West), with four departure routes each (for a total of 16 departure routes).When convective weather occurs in the region, routes may be closed and/or increased space between departures may be required, i.e., Miles-in-Trail (MIT) between departures.The most common restriction compresses the capacity from four routes down to one with 10 MIT.In this situation, both KDFW and KDAL airports are requested to space departures, and as a result, flights frequently incur surface delays.While departures to the East gate may be delayed, departures to the North or South gates may not be.Therefore, opportunities for flights to depart earlier on a north or south alternative route may provide advantageous delay savings despite the cost of flying a longer route.In this scenario, rerouting flights would effectively off load demand and reduce delay to the East gate as well, as shown in Fig. 1.In addition to terminal restrictions, different taxi times and demand loads across runways may also provide options for flights to take off earlier and thus may contribute to potential delays savings.2) Arrival Demand/Capacity Imbalances: Arrival flights may also experience airborne delay due to excess demand or reduced capacity at the arrival metering fixes or runways.Recent HITL simulations, using a research version of CTOP, assessed the potential benefits of using TOS to strategically offload arrivals across routes into Time-Based Flow Management (TBFM)'s arrival metering.Results indicated that delays could be reduced while maintaining throughput at the destination [16].This scenario was not addressed in Phase 3.3) En-route Demand/Capacity Imbalances: Special Use Airspace (SUA), Aircraft Hazard Areas (AHA), and weather events may constrain airspace and create demand/capacity imbalances, as well.The Air Traffic Control System Command Center (ATCSCC, or DCC) may issue Traffic Management Initiatives (TMIs), such as Airspace Flow Programs (AFPs), or route advisories to manage flows.AFPs may be used to reduce the demand near constrained airspaces.Flights that are subject to AFPs receive an EDCT and are usually delayed.The flights that are delayed are those that cross metering arcs set near the constrained airspaces.Under this type of TMI, TOS may be used to look for alternative routes that would not cross metering arcs and would therefore enable flights to depart earlier.Mandatory route advisories may also be issued to manage the flow of traffic away from constrained areas onto specific protected route segments.In this case, comparing TOS routes to the protected segments may help to identify which TOS routes may be available.There may be other constraints, such as predicted arrival time, flight duty time limitations, or airport curfews, that may be considered by the FOs.These may act as compounding factors to the demand/capacity imbalance scenario described above.
+II. CONCEPT OVERVIEWAt a minimum, the tactical identification of beneficial alternative TOS routes depends on the ability to consider constraints on available resources, to predict real-time delay savings benefits, and for the FOs and ATC to coordinate the submission and the approval of flight reroutes.The Phase 3 concept can be summarized as consisting of four essential processes: 1) creation of a TOS, 2) estimation of delay, 3) determination of candidate TOS routes, and 4) submission and approval of TOS reroutes.These four processes are described at a high-level below.1) Creation of a TOS: A TOS service generates a TOS, that is a set of alternative routes, for every departure flight.It computes an RTC for each TOS route.In this early prototype and evaluation, Coded Departure Routes (CDRs) were used as pre-defined TOS routes.CDRs are full-route procedures that are published, and therefore accessible to both ATC and the FO.Using CDRs also enables the system to continuously assess the impact of restrictions and delays on these routes.2) Delay estimation: A scheduling service provides offtimes and delay estimates for both the filed route, and each TOS route, every 10 seconds.This is based on demand and capacity predictions at the runways and at the terminal boundaries, and delays that are imposed on the surface.The scheduler then predicts delay savings for each TOS route, as well as aggregate delay saving for subsequent flights that would not need to be rerouted themselves.Lastly, real-time metrics indicate the probability that a delay savings will occur based on the accuracy of the previous system predictions and the expected size of the delay savings.3) Determination of Candidate TOS Route: A TOS service identifies when and which TOS routes may be beneficial to fly.It compares the TOS route RTC (cost) with predicted delay savings (benefits).When the delay savings exceeds the RTC (net savings), the system identifies the TOS route as a candidate for reroute.
+4) Submission and Approval of TOS reroute:A Graphical User Interface (GUI) enables the FO to view and submit one or more TOS routes for a flight, and enables ATC to view and approve a reroute of a flight on a TOS route.Audio/visual alerts are available to ATC users when a TOS route is submitted and to FO users when a TOS route is approved by ATC.The scheduler updates the estimated schedule and predicted delays for all flights after each TOS reroute approval.Dispatch and pilots concur on the new flight route.The ATC tower updates the flight plan in the FAA's legacy system and clears the pilots on the new flight route.In the following sections, the technology, the GUI and how the system is used are described in more depth.
+III. PHASE-3 CAPABILITY
+A. ATD-2 Phase 3 Built Upon Phase 1 & 2 CapabilitiesThe Phase 3 TOS capability was built upon the Phase 1 and 2 single airport IADS traffic management system.The purpose of the single airport system is to provide accurate prediction of future surface demand/capacity imbalances, to propose SMPs, and to provide gate hold advisories to reduce excess taxi time for departures [6,7,10].The scheduling capability used in Phases 1 and 2 is fully compatible with the Phase 3 TOS capability.However, it is worth noting that the Phase 3 prototype is being evaluated on its own in the North Texas metroplex.Another precursor of the Phase 3 TOS prototype was the Tactical Departure Scheduling -Terminal software that was developed to provide a terminal-wide scheduling system that would support the management of demand/capacity at the terminal boundary [5].Some of the terminal capability has been integrated in the Phase 3 system described below.
+B. Traffic Management Inititiatives (TMI) ServiceThis service parses TMI restrictions from across the NAS and provides the constraints needed by the scheduler service to schedule takeoff times and predict delays at the runways and crossing times at the terminal boundaries.It also identifies constraints on the CDRs used by the TOS service to determine when CDRs are available in the flight's TOS.The TMI service parses restrictions that are available in the TFM Flow data via the System Wide Information Management (SWIM).The system accounts for Ground Stops, EDCTs from Ground Delay Programs (GDPs) and AFPs, Approval Request / Call-for-Releases (APREQ/CFRs), as well as terminal fix closures and MITs.TMCs at the Fort Worth Air Route Traffic Control Center (ZFW) agreed to enter terminal restrictions imposed on the D10 TRACON in the National Traffic Management Log (NTML).These entries were standardized so that the ATD-2 system could parse them reliably.Both fix closures and MITs are entered in the NTML's MIT restriction tab.Alternative fixes to the closed fixes are entered in a qualifier field.The TMI service assesses whether CDR routes are impacted by the DCC's route advisories that mandate traffic to fly specific protected route segments.The system compares the CDRs with the protected segments and filters out excluded CDR routes from the flight's TOS accordingly.
+C. Trajectory Option Set (TOS) ServiceFor this first TOS prototype and evaluation, field demonstration partners agreed to use CDRs in the flights' TOS.These pre-defined routes provide full procedures from KDFW and KDAL to over 95% of their destinations.Examples of CDRs to La Guardia Airport (KLGA) are shown in Fig. 2. Some CDRs were removed from the database, including, CDRs to international destinations and those that could only be used when arrival traffic demand is light at a specific D10 arrival gate.There are several advantages to using CDRs as static TOS routes.First, they are known by ATCs and FOs, and thus do not require the FO to submit a TOS to ATC via SWIM.This enables the TOS service to compare the CDRs to the filed routes on a constant basis.Second, ATC can use the CDR alphanumerical codes to expeditiously amend the flight plan in the ATC legacy system.Third, the CDR codes may also be used by the ATC Tower to relay the new route to the fight crew without having to go through a full route readout, provided a Letter of Agreement to that effect is in place.The TOS service determines TOS states.The eligibility state indicates whether a route is available and if it is beneficial for a reroute.The eligibility states are: Potential, Candidate, Excluded, and Expired.The coordination state enables communication between the FO and ATC about the TOS routes.The coordination states are: FO Submitted, ATC Approved, Reroute Filed, ATC Excluded, and FO Excluded.Before the TOS Service determines if any TOS route is a candidate for a reroute, the service first determines if any CDRs are available.This information is provided by the TMI service.CDRs may be impacted by DCC mandatory reroutes, terminal fix closures, or be excluded due to APREQs, EDCTs or Ground Stops.In these instances, the routes will be labeled "ATC Excluded."Note that the FOs can also exclude their own flights.The service then compares the TOS route delay savings to its RTC.Predicted delay savings are provided by the scheduler service.When delay savings rise above the RTC (net savings), the route state is changed from the "Potential" state to "Candidate" (for reroute).Flights' TOS routes are initially shown as in a "Potential" state until they become a "Candidate."The TOS states are updated every 10 seconds.The RTC of a TOS route is a weighted travel time (in minutes).It is computed by multiplying the difference between mileage of the TOS route and the filed route by a cost factor, divided by the filed speed.The cost factor accounts for the FO's additional cost to operate in the air versus on the surface, or other business priorities.As an example, the FOs' costs could be based on types of aircraft, destinations, and time of day.The FOs' cost factors are static and accessed by the TOS service.A configurable minimum RTC is also included.This sets a threshold for the minimum delays savings that is deemed beneficial.A future system may consider cost factor parameters to be dynamic and managed by the FOs.Fig. 2 shows an example of additional nautical miles and ranges of RTC values for two CDRs via the North terminal gates (in blue and orange) compared to the commonly filed route to KLGA via the East gate (in green).In this example, the routes to the north add about 57 to 60nm, and depending on cost factors, the RTC values could range from a weighted time of about 8 min to 19 min, depending on the variability of FOs' cost factors.The concept includes a buffering feature that prevents the eligibility state to change from "Potential" to "Candidate" too frequently when the delay savings value is close to the RTC value.Thresholds can be set under and above the RTC value to allow for some fluctuations.When the thresholds are set, the state doesn't change unless the delay savings crosses the threshold.So far, the scheduler has proven stable enough to not use these buffers.The TOS service identifies a "top" route amongst the flight's TOS.The top route is the route with the highest predicted net savings, that is, the difference between predicted delay savings and the RTC.The top route is displayed in a TOS Table .Users can also view each TOS route for a particular flight in that flight's TOS Menu.When the FO determines that a flight would benefit from a reroute, the FO submits the route in the ATD-2 GUI.The route coordination state then changes to "FO Submitted."When the ATC Tower determines that they can approve the reroute, they approve the route in the GUI.The route coordination state then changes to "ATC Approved."With this prototype, no data is sent to the TFMS.The ATC Tower enters the CDR code in the FAA legacy system to amend the flight plan.Once the flight plan is updated in SWIM, the ATD-2 system will detect the change and update the coordination state to "Reroute Filed" on the GUI.Note that the system allows the FO to submit any route at any time, whether the route is in a "Potential" or "Candidate" state.Indeed, there may be circumstances when the FO may want a flight to take off earlier on an alternative route despite the RTC value, for example, to accommodate other time constraints.Lastly, the prototype includes the option that TOS routes can be set to expire.A Required Minimum Notification Time (RMNT) can be tailored to certain flight events.For example, this type of threshold could be used by ATC or FO personnel to prevent TOS submissions to occur after certain events, such as pushback.This feature has not been used by the field demonstration partners, yet.
+D. Scheduling ServiceThe scheduling service computes predicted delay savings and provides opportunities to use TOS in a tactical manner.It uses multiple algorithms to provide: Below is a high-level description of the Phase 3 scheduler's capability.More details are provided in [14,17].The Phase 3 scheduler computes Estimated Takeoff Times (ETOTs) based on various inputs and constraints: 1) the surface scheduler that provides surface transit times, runway predictions, aircraft movement detections and position of traffic on the surface; 2) the tower TMC's runway utilization entered on the ATD-2 GUI; 3) FOs' estimated departure times, such as the Earliest Off-Block Times (EOBTs) that are used to predict the Undelayed Takeoff Times (UTOTs)-the earliest times flights could depart independent of other flights and separation applied at the runway; 4) Terminal restrictions from the TMI service which may require combining departures over certain fixes when other fixes are closed, as well as increased spacing for MIT compliance; and lastly 5) the competing demand between major airports and engine types over certain departure fixes that may result in terminal delay that is passed back to the departure runways.The best estimate of a flight's takeoff time on the filed route is the ETOT.Predicted total delay on the filed route is computed as the difference between the ETOT and the UTOT.In addition, the scheduler computes what-if ETOTs for all TOS routes for all flights.It assumes the same UTOT on the filed route but then computes a new ETOT for each specific TOS route.Because the restrictions and the demand may vary across the different routes and runways, it is possible that the ETOT for an alternative route may differ from the ETOT for the filed route.The difference between the two ETOTs is then used to compute delay savings.The following factors may influence delay savings: 1) when a terminal restriction may be in effect on the filed route, but not on the alternative route, 2) when departures on the alternative route depart from a different runway that has a lower demand than the filed route, and 3) when departures on an alternative route have a shorter taxi time.Fig. 3 provides an example of AAL1560's UTOT, filed route ETOT and a TOS route ETOT.In this example, the ETOT of filed route indicates a delay of 13 minutes (middle timeline).However, the ETOT for the southern TOS route (right timeline) is as early at the UTOT (left timeline) thus providing a delay saving of 13 minutes.Note that the delay savings accounts for different UTOTs at different runways, as well as when flights push from the gate (out events).This helps to account for accrued delays, in case route submissions and reroute decisions are made after pushback.Typically, an increase in accuracy is observed after an aircraft pushes back and gets closer to the runway.However, the closer the flight gets to the runway, the fewer the opportunities there will be for departing earlier on an alternative route, and the higher the workload for the FO and ATC to reroute the flight.Lastly, the scheduling service provides a probability (%) that the flight's delay savings for the "top" route will occur.Essentially, it compares the likelihood of a flight's ETOT with previous Actual Takeoff Times (ATOTs) from historical data.The difference between the probability of an ETOT and the distribution of ATOTs then yields a probability for various delay values.The method to calculate this probability is described in detail in [17].The probability can indicate how likely a delay savings is going to be higher than zero, or higher than the RTC value.When compared to the RTC value, the probability typically reaches 50% or higher when a TOS route becomes candidate which is when the delay savings is estimated to be higher than the RTC.The probability is influenced by 1) the accuracy of estimates (the higher the accuracy, the higher the probability), and 2) the size of the delay savings (the larger the size, the higher the probability).In addition to individual delay savings estimates, the scheduling service also provides what-if estimates of aggregated delay savings for other flights, up to 60 minutes, behind a given flight if the given flight were rerouted on its top TOS route.The aggregated delay savings is a sum of the delay savings for other flights plus the delay savings for the given rerouted flight.The aggregate delay savings is computed at the following levels: for the air carrier (e. g.SWA), for the major's fleet (e. g., the American Airlines Group operates with several sub air carriers), for each airport, and for the metroplex (KDFW and KDAL).Consider the following example: Flight ABC123's filed route is under a 10 MIT restriction to the terminal East gate.It shows a delay savings of 10 minutes on TOS route #1.There are 10 flights under the same MIT restriction that are predicted to take off behind ABC123 at the same airport.Each of the 10 have an additional one minute of delay, due to the increased spacing to comply with the MIT.Four of those belong to the same air carrier.An additional three of those belong to the same fleet.There are five other flights from another airport that are also impacted by the MIT.In this example, rerouting ABC123 on the TOS route #1 would actually delay two flights from the same fleet by a minute each, because the rerouted flight would be predicted to move in front of these 2 flights.The breakdown of the aggregate delays is shown in Table 1.
+IV. GRAPHICAL USER INTERFACEThe Phase 3 GUI, called Metroplex Planner, provides key display elements such as, Timelines, Maps, and Tables.Many of these elements are part of the Surface Trajectory Based Operations (STBO) GUI that was developed for Phases 1 and 2 -See [6,7].Modifications to the GUI for Phase 3 are highlighted in the sections below.
+A. TOS TablesTwo new tables were created for TOS information as shown in Fig 4 .The TOS Departure Table (TDT) lists flights' filed route data, the "top" TOS route data, and the TOS states.Users can set various relevant information to be displayed in the TDT.Fig. 4 shows examples of flight's filed route, equipage, TMI information, ETOT, delay, plus the top route's individual and aggregate delay savings, the probability of delay savings, and the TOS eligibility and coordination states.An editable scratchpad field was added to allow users to make notes for themselves and to communicate with NASA researchers.Several TDTs can be stacked on top of each other.Using filters, tables can be set up by TOS states, flight events, and time or value thresholds, thus enabling an ad-hoc organization of flights.For example, one table may list flights with candidate routes, and another may list the flights with submitted and approved routes.The other table is the TOS Flight Menu (FM).It lists all the routes for a specific flight as shown in the bottom of Fig. 4. The filed route is in the first row, and in the following rows are the CDRs that are available.Each of these rows indicates the route procedure, the mileage, additional mileage compared to the filed route, the RTC, as well as the ETOT and estimated delay.In the TOS Flight Menu shown in Fig. 4, the CDRs via the North gate show earlier ETOTs than the filed route and delay savings above the RTC.In Fig. 4 the FM for AAL2334 lists two candidate routes via the North gate with 45min of delay savings on each route.On the TDT display above, the top route is the CDR DFWPHLJ3.This is due to its lowest additional mileage of 54nm (as seen under "Add nm" in the FM).The ETOT of this route has a probability of 86% of delay savings above the RTC value.The aggregate delay savings are 54min for 27 flights (2min per flight) at the air carrier level, 59min at the fleet level, 68min at the airport level, and 91min for both airports.Both tables can be used to submit or approve TOS routes.To do so, an FO user right-clicks on a flight route and selects which TOS route to submit.The FO can submit more than one route per flight.Once the FO submits one, or more, routes, the coordination state of the route changes to "FO Submitted."To approve a reroute, the ATC user right-clicks on the flight route and approves the route.This will update the Coordination State to "ATC Approved."The FO has the ability to undo a submission until ATC approves a submitted route.ATC can also undo an approval, as needed.Optional pop-up windows can be set to alert ATC users when a route has been submitted or to alert the FO users when a route has been approved.Measures were taken to protect any sensitive information of the FOs.Each operator had their own criteria to compute the RTC's cost factor and minimum RTC values.Each FO can see only their own flight and TOS information in the TOS tables, but can see all the traffic on the timeline and map on their respective clients.However, ZFW, KDFW and KDAL ATC facilities can see both FOs' flights and TOS information across airports.
+B. TimelineThe purpose of the Timeline is to present the likely time of departure or arrival to a reference point, such as a runway or a fix.Datablocks provide flight-related information such as delay, restrictions, aircraft type, destinations, origin, runway, and parking gates.In Phase 3, timelines were upgraded to show the ETOT of the flights' filed route, the predicted delay, as well as two new letter indicators.The letter D indicates that the aircraft is equipped with Controller Pilot Data Link Communications-Departure Clearance (CPDLC-DCL).The letter T, highlighted in white, indicates that a TOS route has been submitted for a flight, and the letter T, highlighted in green, indicates that the TOS route has been approved by ATC.Both the FO and ATC can submit and approve routes by right-clicking on a flight's datatag.The users can also bring up the Flight's TOS Menu from there.
+C. MapThe Map provides a planview of D10 arrival and departure traffic, as well as airspace elements, such as airports, fix names The color indicates the level of restriction on the CDRs.When the color is green, the CDRs are available to all possible destinations.The CDR is shown as either "Potential" or "Candidate" state in the flight's TOS.When the color is red, the CDRs are not available to any destination.In this case, the CDR is shown as "ATC Excluded" state in the flight's TOS.When the color is yellow, some CDRs are not available to specific destinations.The CDR is shown as "ATC Excluded" state in the flight's TOS to these destinations only.The user can hover over the indicator to see which destinations are effectively excluded.
+D. Management of TOS OperationsA TOS Operation tab and a DCC Route Advisory tab were added to the Traffic Management Menu in the GUI.The TOS Operation tab provides ways for the center's TM Unit to turn TOS operation ON or OFF, as well as to manage the exclusions of destinations or CDR routes in the flights' TOS as shown in Fig. 6.The upper left corner of Fig. 6 (highlighted in red) shows where the TM Unit can activate or deactivate TOS operations.When TOS operations are active, the FO can submit TOS routes, and ATC Tower can approve them.When TOS operations are  active, a TOS icon (not shown in the document) changes from gray to orange to provide situation awareness.The top section of Fig. 6 (highlighted in blue) shows where the TM Unit can enter destinations that are restricted from TOS rerouting.That is when none of the CDRs are available to these destinations.The TMI service automatically provides destinations under TMIs such as those with EDCTs under GDPs, and APREQ/CFRs.However, the parsing of the TMIs may not include all destinations that need to be excluded.The main section in Fig. 6 displays CDR indicators by cardinal direction.The CDRs listed under each direction are the most frequently used for reroutes via adjacent terminal gates.For instance, flights that are filed over the East gate may use the 1N or 1S CDR via the North and South gates respectively as alternate routes.The TMI Service will determine when certain CDRs are excluded to certain destinations in the DCC mandatory routes.These destinations are then listed next to the given CDR indicators.For example, in Fig. 6, the CDRs ending in 1N are excluded in the flights' TOS to EWR, JFK, and LGA (highlighted in green).The user can manually add or subtract destinations, as needed.
+V. ATD-2 PHASE 3 PROCESS FLOW AND AGREEMENTS
+A. Phase 3 -Process FlowThe Metroplex Planner was designed to be adaptable to various decision processes.It can help either the FOs or ATC to identify flights that would benefit from being rerouted.Below is the approximate process flow developed by the FOs and ATC after a period of shadow evaluation, early tests, and discussion amongst field demonstration partners in 2019.
+1)The en route ZFW TM Unit determines terminal restrictions and makes entries in NTML.This unit also decides if TOS operations could be initiated based on the traffic situation and workload involved in the terminal airspace and airports.2) The FO's ATC coordinator coordinates with dispatch about potential flights that could be rerouted.They discuss whether flights could be rerouted based on fuel and equipment.They may decide to include CDR information in the flight plan release for pilots.3) The FO's ATC coordinator monitors the TOS Table .As the flight approaches 30min before pushback, a TOS route may be inspected for weather and wind information.If dispatch and the ATC coordinator concur, the TOS route is submitted on the ATD-2 GUI.
+4)The ATC Tower verifies that there is an opportunity to reroute the flight on the alternative route.The ATC may approve the reroute and coordinate with Clearance Delivery or Ground Control, depending on which position is controlling the flight at the moment.The flight plan is amended in the Flight Data Input/Output system.5) The FO's ATC coordinator may alert dispatch when ATC approves the reroute on the ATD-2 interface.Dispatch may see that the flight plan amendment has been updated on the FO's system.Dispatch is then legally required to concur with the pilots on the new route and the amount of fuel on board.
+6)The ATC Tower clears the pilots on the new route.The clearance may be carried via CPDLC-DCL or VHF.CPDLC-DCL allows the Tower to send the amendment via datalink to flights that indicated this capability in the flight plan.VHF requires a full route readout.Some airlines may have a Letters Of Agreement with ATCT to use the CDR code to relay the new route, in the absence of CPDLC-DCL equipment.
+7)The pilots eventually enter, or select, the new route in the Flight Management System, verify fuel and recompute weight and balances on the new departure procedure.
+B. Phase 3 -Procedural AgreementsAgreements were made to facilitate the above process.These may not be applicable to other airspaces.The ZFW TM Unit agreed to make standardized NMTL entries.Restriction durations are up to two hours, but these may be updated as needed, based on evolving conditions and weather predictions.The ZFW TM Unit determines if TOS submissions and approval are acceptable, based on workload and constraints in the airspace.However, the ATC Towers approve the reroutes, since they were in control of the flights 45 min prior to departure time.FO submissions are treated as reroute requests and are approved by the ATC Towers, as able.It was agreed that ATC Towers don't need to approve or reject each of the TOS reroute requests and instead, don't act upon a TOS submission if they are not able to evaluate the request due to higher priority duties.This, however, has not happened yet.The system allows the FO to submit one or more TOS routes, regardless of its "Potential" or "Candidate" state.If several are submitted, ATC picks the route they deem beneficial and can approve.However, in practice, the FO has submitted only one alternative route per flight, so far.Both the FO and ATC have preferred to submit and approve a TOS reroute before pushback to reduce workload on the pilots and the controllers.There is a mutual desire to avoid amending the flight plan once the flight is past the spot.Note that so far in North Texas, the onus has been on the FO to submit TOS routes when reroutes are deemed beneficial for a flight.The Phase 3 system supports that approach which seems to work particularly well when multiple routes are available, and the demand on alternative routes is not saturated.This may not be the case in other circumstances when ATC is constrained to impose reroutes, such as during Severe Weather Avoidance Plan (SWAP) events.
+VI. BENEFITS
+A. Individual and Aggregate Delay SavingsThe Phase 3 system calculates predicted delay savings for both individual benefits (for the rerouted flights) and aggregate (system-wide) benefits (for multiple subsequent flights).These predicted benefits may weigh differently in the decisions made by the FOs or by ATC.The combination of individual and system-wide benefits can result in the following:1) High individual delay savings, and high aggregate delay savings: This is the "no-brainer" case when the cost of rerouting the flight provides both individual and system-wide benefits, which may or may not span across air carriers and airports.2) Low individual delay savings, and low aggregate delay savings: This is the nominal scenario where demand is not exceeding capacity, and no delay savings can be produced.3) High individual delay savings, and low aggregate delay savings: In this case, rerouting the flight may result in delaying other subsequent flights.An aggregate savings value lower than the individual value indicates that delay may increase for some flights, since the aggregate delay savings includes the individual delay savings as well.
+4) Low individual delay savings, and high aggregate delay savings:In this case, rerouting the flight may not result in individual delay savings, but in aggregate delay savings for the air carrier, the fleet, for the airport, or for the metroplex.When the aggregate delay savings benefits the air carrier or the fleet, the FO may consider rerouting 'low-impact' flights.ATC may have different criteria for selecting flights to reroute.The aggregate delay savings computation may be used to justify rerouting a particular flight.The operational use of the system should provide additional insights in how the individual and aggregate delay savings weigh in the decisions to submit and approve reroutes.A future system could leverage both individual and aggregate delay savings, as well as other inherent costs to determine possible outcomes when several flights are considered for a reroute, or to provide recommendations on which flights would provide targeted benefits.Such a capability could be designed to support both FOs or ATC priorities.
+B. Stormy 2019 and 2020 EvaluationsAn initial TOS prototype was deployed at the end of Spring 2019 to AAL and SWA ATC Coordinators at their respective Operation Centers, to the TMC positions at ZFW, D10, and to KDFW and KDAL ATC Towers (there is no TMC position at KDAL).After a period of training and shadow observations, the system was tested with users for four-hour periods over several days in the summer of 2019.On a few occasions, the weather resulted in restrictions and delay savings and candidate routes were observed.Several TOS routes were submitted, and flights were rerouted on the TOS routes.This provided opportunities to evaluate the system, and identify additional features to be added, such as aggregate delay savings and the probability of delay savings.A field test was initially planned during the stormy season of 2020, which usually takes place from late May to early September.However, the health-related measures implemented to contain the COVID-19 pandemic resulted in a significant reduction of air traffic.By May 2020, the combined average daily departures at KDFW and KDAL had dropped to 47% of January 2020's average (1226/day average in January to 572/day in May).Departure demand was reduced from eight to three banks at KDFW.In addition to traffic, both FO and ATC staffing was reduced.By the end of June, neither the FO nor the ATC had effectively used the system.However, inclement weather developed as expected.Severe weather took place on several days which resulted in fix closures, MITs, and some departure delays, despite reduced demand.The TOS system indicated candidate TOS routes on several occasions.Below is one example to illustrate possible benefits in the absence of actual reroutes.On May 15 th , fix closures restricted the East gates.Traffic from two departure routes were combined over one, with one other departure route remaining unimpacted.That level of restriction lasted for 90min (21:30 to 01:00 UTC) and coincided with the evening departure bank at KDFW (the first OUT occurred at 23:25 and the last OFF occurred at 01:30).During that period of time, the TOS system indicated that seven flights had candidate TOS routes.No flights were actually rerouted during this period.Therefore, realized aggregate delay savings are not reported here.Because rerouting flights impact the ETOTs and delay savings of subsequent flights, and without knowing how many flights could be effectively rerouted, we opted to show the potential benefits for only one flight in that period of time.In the example below, the flight was the first in a bank that met the criterion of having a candidate TOS route at pushback (OUT).The example flight was AAL1446 from KDFW to KTUL.Normally, flights to KTUL are filed over the North gate.However, due to weather, this flight was filed on the FORCK route via the East gate and was expected to depart from the East runway.When it pushed from the gate at 23:25, its ETOT was 23:44 with a predicted delay of three minutes at Runway 17R.At the same time, an alternate route to the West gate from Runway 18L had an earlier ETOT of 23:35 with no predicted delay at the runway.Due to differences in taxi times and delay at the runway, the system indicated a delay savings of about nine minutes and a positive net time savings on the TOS route.The flight's probability of delay savings above its RTC was about 60%.The aggregate delay savings was projected to be about 32min at the air carrier level (AAL), 60min at the fleet level (AAL plus ENY, ASH, and SKW), 63min at the airport level (all airlines at KDFW), and 63min at the metroplex level (both KDFW and KDAL).It is worth noting that in this example, the predicted delay savings on the TOS route was based upon an advantageous taxi time and the absence of delay at a different departure runway.Also, our decision to pick the first flight in the bank with delay savings greater than the RTC may result in larger aggregate delay savings than if the flight was later in the bank.The delay savings in this example may, or may not, reflect the FO's own criteria to make a TOS submission.Lastly, the FOs may have additional considerations and metrics that are not yet being accounted for in the ATD-2 system.
+C. Operational BenefitsThe integration of TOS in FO and ATC operations has been shown to provide many advantages.Below are identified operational benefits from the use of the Phase 3 capability.These may be worth considering for the development of future systems:1) Real-time predictions: Estimates of demand, off-time, delay, delay savings, and likelihood of delay savings could support traffic flow management decisions such as load balancing.These estimates could also help identify other sources of delay.2) Integration of TMIs: It is critical to account for the impact of TMI restrictions when calculating predicted delay, alternative routes, and destinations.Considering TMIs helps to determine when flights are eligible to submit a TOS.3) Information on which flights are eligible for viable alternative TOS routes: This information helps ATC to identify whether flights are able to fly alternative routes, based on criteria such as equipment and fuel.4) Information on when alternative TOS routes are beneficial for reroute: This helps to inform FOs and ATC rerouting decisions, based on agreed-upon benefit criteria.5) More efficient workflow: TOS information helps to increase ATC and FO's awareness of reroute opportunities and decisions.It also reduces the coordination effort between them.6) Reduction of risks/costs: Avoiding departure delay could help to minimize costs related to flights returning to the parking gate to refuel, to offload passengers to avoid tarmac rules, or to avoid crew Flight Duty Time limitations.Reducing delay may help FOs with passenger, crew, and aircraft connectivity at destinations, as well as help ATC with maintaining throughput.
+VII. CONCLUSIONTOS has been identified as a resource to meet NextGen objectives by the Industry and FAA.Among those objectives are better communications via data exchange across the airspace system and its users, fuel savings, a reduction in emissions and reduced congestion.The ATD-2 Phase 3 system enables both the FOs and ATC to tactically identify reroute opportunities for departure flights impacted for departure fights impacted by demand/capacity imbalances.The Phase 3 field evaluation has helped to identify uses cases and operational benefits.The remainder of the field evaluation will provide opportunities to analyze benefits, identify new areas of development as well document lessons learned.In the future, evaluating TOS opportunities in other airspace configurations and traffic demands, as well as efforts to integrate TOS information and predictive engines into the FO's own flight planning system and into the FAA's TFMS and TFDM systems may be considered.Fig. 1 .1Fig. 1.Alternative TOS departure routes to offload demand and reduce delay.
+Fig. 2 .2Fig. 2. RTC values and CDRs from KDAL, KDFW to KLGA.The green line is the CDR route via the East gate that is commonly filed, and the blue and orange lines are two CDRs via the North gate.
+Fig. 3 .3Fig. 3. Example of filed route UTOT and ETOT, and TOS route ETOT and delay savings for AAL1560.
+Fig. 4 .4Fig. 4. Top: Example of a TOS Departure Table (image shown in two parts: left and right halves) which shows two flights with candidate TOS routes.Bottom: Example of a TOS Flight Menu which shows the filed route and each alternative TOS route for AAL2334.
+and radar video maps.New information about restrictions at the terminal boundary and on the CDRs were added to the Metroplex Planner Map.Terminal restrictions impacting departure fixes, such as fix closures and MITs are now indicated next to the fix name, and restrictions on CDRs are located on the outskirts of the D10 airspace.An example is shown in Fig. 5.In Fig. 5, Fix ZERLU (ZER) is displayed in gray.This indicates that the fix is closed."ZER" is however displayed next to the adjacent fix, TRRCH (TRR).This indicates that flights bound to ZERLU are being routed over the alternate fix TRRCH.The number 10 in front of the fix indicates that there is a 10 MIT in place over TRRCH.In Fig. 5, color-coded indicators of CDR availability in the D10 airspace are shown, grouped by Departure Fix.These indicators are based on the last two alphanumeric characters of each CDR.Many CDRs have the same route name since many use the same Standard Instrument Departure (SID) routes exiting in the D10 airspace into the ZFW airspace.For instance, numerous CDRs ending in 1N fly over the AKUNA (AKU) fix.
+Fig. 5 .5Fig. 5. Map with Departure Fixes, TRACON radar video map, and TOS Route availability labels.
+Fig. 6 .6Fig. 6.Metroplex Planner TOS Operation Tab.
+TABLE I .IAGGR.DELAY SAVINGS (IN MIN) IF FLIGHT ABC123 DEPARTED ON TOS ROUTE #1
+Aggregate Levels Delay Values Added Together (min) Total Flight Delay Savings Flights With Reduced Delay Flights With Increased Delay Aggregate Delay SavingsAir carrier-10-40-14Fleet-10-72-15Airport-10-102-18Metroplex-10-152-23
+		
+		
+
+			
+ACKNOWLEDGMENTNone of the concept, prototype development, and field evaluation would have been possible without the commitment and support of our field demonstration partners in North Texas, as well as the CDM Stakeholders Group (CSG), FET and SCT groups, and the FAA/ANG.
+			
+
+			
+
+				
+
+
+	
+		An Integrated Collaborative Decision Making and Tactical Advisory Concept for Airport Surface Operations Management
+		
+			GautamGupta
+		
+		
+			WaqarMalik
+		
+		
+			YoonJung
+		
+		10.2514/6.2012-5651
+	
+	
+		12th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference and 14th AIAA/ISSMO Multidisciplinary Analysis and Optimization Conference
+		
+			American Institute of Aeronautics and Astronautics
+			June 2014
+		
+	
+	FAA Air Traffic Organization Surface Operations Office
+	FAA Air Traffic Organization Surface Operations Office, "U.S. airport surface collaborative decision making (CDM) concept of operations (ConOps) in the near-term: Application of the Surface Concept at United States Airports," June 2014.
+
+
+
+	
+		NASA's ATM technology demonstration 1 (ATD-1): Integrated concept of arrival operations
+		
+			BBaxley
+		
+		
+			HSwenson
+		
+		
+			TPrevot
+		
+		
+			TCallantine
+		
+		10.1109/dasc.2012.6382981
+		NASA/TM 2013-218040, Version 2.0
+	
+	
+		2012 IEEE/AIAA 31st Digital Avionics Systems Conference (DASC)
+		
+			IEEE
+			September 2013
+		
+	
+	B. Baxley, W. Johnson, H. Swenson, J. Robinson, T. Prevot, T. Callantine, J. Scardina, and M. Greene, "Air Traffic Management Technology Demonstration 1 Concept of Operations (ATD-1 ConOps)," NASA/TM 2013-218040, Version 2.0, September 2013.
+
+
+
+	
+		Performance Evaluation of SARDA: An Individual Aircraft-based Advisory Concept for Surface Management
+		
+			YoonJung
+		
+		
+			TyHoang
+		
+		
+			MiwaHayashi
+		
+		
+			WaqarMalik
+		
+		
+			LeonardTobias
+		
+		
+			GautamGupta
+		
+		10.2514/atcq.22.3.195
+	
+	
+		Air Traffic Control Quarterly
+		Air Traffic Control Quarterly
+		1064-3818
+		2472-5757
+		
+			22
+			3
+			
+			2015
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	Y. Jung, W. Malik, L. Tobias, G. Gupta, T. Hoang, and M. Hayashi, "Performance Evaluation of SARDA: An individual aircraft-based advisory concept for surface management," Air Traffic Control Quarterly, Vol. 22, Number 3, 2015, p. 195-221.
+
+
+
+	
+		
+			SEngelland
+		
+		
+			ACapps
+		
+		
+			KDay
+		
+		
+			MKistler
+		
+		
+			FGaither
+		
+		
+			GJuro
+		
+		NASA/TM-2013-216533
+		Precision Departure Release Capability (PDRC) Final Report
+		
+			June 2013
+		
+	
+	S. Engelland, A. Capps, K. Day, M. Kistler, F. Gaither, and G. Juro, "Precision Departure Release Capability (PDRC) Final Report," NASA/TM-2013-216533, June 2013.
+
+
+
+	
+		Design Characteristics of a Terminal Departure Scheduler
+		
+			AlanCapps
+		
+		
+			MathewSKistler
+		
+		
+			ShawnAEngelland
+		
+		10.2514/6.2014-2020
+	
+	
+		14th AIAA Aviation Technology, Integration, and Operations Conference
+		Atlanta, Georgia
+		
+			American Institute of Aeronautics and Astronautics
+			June 2014
+		
+	
+	A. Capps, M. Kistler, and S. Engelland, "Design characteristics of a terminal departure scheduler," 14th AIAA Aviation Conference, Atlanta, Georgia, June 2014.
+
+
+
+	
+		ATD-2) Phase 1 Concept of Use (ConUse)
+		
+			YJung
+		
+		
+			SEngelland
+		
+		
+			ACapps
+		
+		
+			RCoppenbarger
+		
+		
+			BHooey
+		
+		
+			SSharma
+		
+		
+			LStevens
+		
+		
+			SVerma
+		
+		
+			GLohr
+		
+		
+			EChevalley
+		
+		
+			VDulchinos
+		
+		
+			WMalik
+		
+		
+			LMorganRuszkowski
+		
+		NASA/TM-2018-219770
+	
+	
+		Airspace Technology Demonstration
+		
+			2
+			February 28, 2018
+		
+	
+	Y. Jung, S. Engelland, A. Capps, R. Coppenbarger, B. Hooey, S. Sharma, L. Stevens, S. Verma, G. Lohr, E. Chevalley, V. Dulchinos, W. Malik, and L. Morgan Ruszkowski, "Airspace Technology Demonstration 2 (ATD-2) Phase 1 Concept of Use (ConUse)," NASA/TM-2018-219770, February 28, 2018.
+
+
+
+	
+		Airspace Technology Demonstration 2 (ATD-2) Phase 1 Concept of Use (ConUse) Addendum for Phase 2
+		
+			YJung
+		
+		
+		
+			July, 7, 2020
+		
+	
+	Y. Jung, "Airspace Technology Demonstration 2 (ATD-2) Phase 1 Concept of Use (ConUse) Addendum for Phase 2," Accessed on: July, 7, 2020. [Online]. Available: https://aviationsystemsdivision.arc.nasa.gov/publications/atd2/tech- transfers/1_High-Level_and_Project_Documents/1.1- 07%20ATD2_Phase_2_ConUse_20190918.pdf
+
+
+
+	
+		Field evaluation of the baseline Integrated Arrival, Departure, Surface Capabilities at Charlotte Douglas international airport
+		
+			YJung
+		
+		
+			WCoupe
+		
+		
+			ACapps
+		
+		
+			A
+		
+		
+			SEngelland
+		
+		
+			SSharma
+		
+	
+	
+		13th USA/Europe Air Traffic Management Research and Development Seminar
+		
+			June 2019
+			Vienna, Austria
+		
+	
+	Y. Jung, W. Coupe, A. Capps, A., S. Engelland, and S. Sharma, "Field evaluation of the baseline Integrated Arrival, Departure, Surface Capabilities at Charlotte Douglas international airport," 13th USA/Europe Air Traffic Management Research and Development Seminar, Vienna, Austria, June 2019.
+
+
+
+	
+		Predicting Gate Conflicts at Charlotte Douglas International Airport Using NASA ATD-2 Fused Data Sources
+		
+			WilliamJCoupe
+		
+		
+			HanbongLee
+		
+		
+			AndrewChurchill
+		
+		
+			IsaacRobeson
+		
+		10.1109/dasc50938.2020.9256641
+		ATD-2 Team
+		
+	
+	
+		2020 AIAA/IEEE 39th Digital Avionics Systems Conference (DASC)
+		
+			IEEE
+			May 27, 2020. July, 7, 2020
+		
+	
+	ATD-2 Team, "ATD-2 benefits mechanism," May 27, 2020. Accessed on: July, 7, 2020. [Online]. Available: https://aviationsystems.arc.nasa.gov/publications/2020/ATD2_Benefits_ Mechanism_v1_20200527.pdf
+
+
+
+	
+		Strategic Surface Metering at Charlotte Douglas International Airport
+		
+			IsaacRobeson
+		
+		
+			WilliamJCoupe
+		
+		
+			HanbongLee
+		
+		
+			YoonJung
+		
+		
+			LiangChen
+		
+		
+			LeonardBagasol
+		
+		
+			BobStaudenmeier
+		
+		
+			PeteSlattery
+		
+		10.1109/dasc50938.2020.9256580
+	
+	
+		2020 AIAA/IEEE 39th Digital Avionics Systems Conference (DASC)
+		
+			IEEE
+			
+		
+	
+	in press
+	I. Robeson, W. Coupe, H. Lee, Y. Jung, L. Chen, L. Bagasol, R. Staudenmeier, and P. Slattery, "Strategic surface metering at Charlotte Douglas international airport," 39th IEEE/AIAA Digital Avionics Systems Conference (DASC), in press.
+
+
+
+	
+		Evaluation of a tactical surface metering tool for Charlotte Douglas international airport via human-in-the-loop simulation
+		
+			SavitaVerma
+		
+		
+			HanbongLee
+		
+		
+			LynneMartin
+		
+		
+			LindsayStevens
+		
+		
+			YoonJung
+		
+		
+			VictoriaDulchinos
+		
+		
+			EricChevalley
+		
+		
+			KimJobe
+		
+		
+			BonnyParke
+		
+		10.1109/dasc.2017.8102046
+	
+	
+		2017 IEEE/AIAA 36th Digital Avionics Systems Conference (DASC)
+		St-Petersburg, FL
+		
+			IEEE
+			September 2017
+		
+	
+	S. Verma, H. Lee, L. Martin, L. Stevens, Y. Jung, V. Dulchinos, E. Chevalley, K. Jobe, and B. Parke, "Evaluation of a tactical surface metering tool for Charlotte Douglas international airport via human-in- the-loop simulation," 36th IEEE/AIAA Digital Avionics Systems Conference (DASC), St-Petersburg, FL, September 2017.
+
+
+
+	
+		Alternatives for Scheduling Departures for Efficient Surface Metering in ATD-2: Exploration in a Human-in-the-Loop Simulation
+		
+			BonnyKParke
+		
+		
+			LindsayK SStevens
+		
+		
+			WilliamJCoupe
+		
+		
+			HanbongLee
+		
+		
+			YoonCJung
+		
+		
+			DeborahLBakowski
+		
+		
+			KimberlyJobe
+		
+		10.1007/978-3-030-20037-4_28
+	
+	
+		Advances in Human Error, Reliability, Resilience, and Performance
+		Washington, D. C.
+		
+			Springer International Publishing
+			July 2019
+			
+		
+	
+	B. Parke, L. Stevens, W. Coupe, H. Lee, Y. Jung, D Bakowski, and K. Jobe, "Alternatives for scheduling departures for efficient surface metering in ATD-2: Exploration in Human-in-the-Loop Simulation," 10th international Conference on Applied Human Factors and Ergonomics, Washington, D. C., July 2019.
+
+
+
+	
+		Human Factors Impact of Different Ramp Controller Scheduling Advisories for ATD-2 Surface Metering in a Human-in-the-Loop Simulation
+		
+			BonnyParke
+		
+		
+			DeborahLBakowski
+		
+		
+			YoonCJung
+		
+		
+			HanbongLee
+		
+		
+			JeremyCoupe
+		
+		
+			LindsayKStevens
+		
+		10.2514/6.2020-2886
+	
+	
+		AIAA AVIATION 2020 FORUM
+		
+			American Institute of Aeronautics and Astronautics
+			June 2020
+		
+	
+	B. Parke, D. Bakowski, Y. Jung, H. Lee, W. Coupe, and L. Stevens, "Human factors impact of different ramp controller scheduling advisories for ATD-2 surface metering in a human-in-the-loop simulation," AIAA Aviation 2020 Forum, June 2020.
+
+
+
+	
+		ATD-2 Phase 3 Scheduling in a Metroplex Environment Incorporating Trajectory Option Sets
+		
+			WilliamJCoupe
+		
+		
+			YoonJung
+		
+		
+			LiangChen
+		
+		
+			IsaacRobeson
+		
+		10.1109/dasc50938.2020.9256509
+	
+	
+		2020 AIAA/IEEE 39th Digital Avionics Systems Conference (DASC)
+		13th USA/Europe Air Traffic Management Research and Development Seminar
+		Vienna, Austria
+		
+			IEEE
+			June 2019
+		
+	
+	W. Coupe, H. Lee, Y. Jung, L. Chen, and I. Robeson, "Scheduling improvements following the Phase 1 field evaluation of the ATD-2 Integrated Arrival, Departure, and Surface concept," 13th USA/Europe Air Traffic Management Research and Development Seminar, Vienna, Austria, June 2019.
+
+
+
+	
+		User Preference and Trajectory Options Sets (TOS) to Benefit Traffic Flow Management
+		
+			MichaelRobinson
+		
+		
+			SteveKamine
+		
+		10.2514/6.2019-3310
+	
+	
+		AIAA Aviation 2019 Forum
+		Forum, Dallas, TX
+		
+			American Institute of Aeronautics and Astronautics
+			2019. June 2019
+		
+	
+	M. Robinson, and S. Kamine. "User Preference and Trajectory Options Sets (TOS) to Benefit Traffic Flow Management." AIAA Aviation 2019 Forum, Dallas, TX, June 2019.
+
+
+
+	
+		Impact of Different Trajectory Option Set Participation Levels within an Air Traffic Management Collaborative Trajectory Option Program
+		
+			Hyo-SangYoo
+		
+		
+			ConnieBrasil
+		
+		
+			NancyMSmith
+		
+		
+			NathanBuckley
+		
+		
+			GitaHodell
+		
+		
+			ScottKalush
+		
+		
+			PaulULee
+		
+		10.2514/6.2018-3040
+	
+	
+		2018 Aviation Technology, Integration, and Operations Conference
+		Atlanta, GA
+		
+			American Institute of Aeronautics and Astronautics
+			June 2018
+		
+	
+	H. Yoo, C. Brasil, N. Buckley, G. Hodell, S. Kalush, P. Lee, and N. Smith, "Impact of different Trajectory Option Set participation levels within an Air Traffic Management Collaborative Trajectory Option Program," In 18th AIAA Aviation 2018 Forum, Atlanta, GA, June 2018.
+
+
+
+	
+		ATD-2 Phase 3 Scheduling in a Metroplex Environment Incorporating Trajectory Option Sets
+		
+			WilliamJCoupe
+		
+		
+			YoonJung
+		
+		
+			LiangChen
+		
+		
+			IsaacRobeson
+		
+		10.1109/dasc50938.2020.9256509
+	
+	
+		2020 AIAA/IEEE 39th Digital Avionics Systems Conference (DASC)
+		
+			IEEE
+			
+		
+	
+	in press
+	W. Coupe, Y. Jung, L. Chen, and I. Robeson, "ATD-2 Phase 3 scheduling in a metroplex environment incorporating Trajectory Option Sets," 39th IEEE/AIAA Digital Avionics Systems Conference (DASC), in press.
+
+
+				
+			
+		
+	

file142.txt

@@ -0,0 +1,408 @@
+
+	
+	
+		
+IntroductionA concept of advanced terminal airspace area operations is part of the Next Generation Air Transportation System (NextGen) efforts to accommodate expected increase in the demand and provide a higher level of throughput at the airports and within en route airspace [1,2,3].Scheduling optimization problems in dense terminal airspace area operations are drawing interest [4], since the increase of airspace capacity is becoming more important to accommodate large air traffic flows in that environment.The use of scheduling algorithms more efficient than the traditional FCFS approach is a way to increase a way to increase throughput and efficiencies in congested terminal airspace compared to the FCFS scheduler.On the other hand, it is known that the route topologies in the extended terminal airspace play an important role in the scheduling performance.Efficient scheduling and route assignment directly affect important performance metrics such as runway delays, throughput, fuel efficiency, and robustness to uncertainties in operations.However, despite the realization of the importance of the route topology, few extensive studies of the problem have been reported.The purpose of this paper is: 1) to investigate optimal and efficient scheduling algorithms for a dense terminal airspace operation that yield better performance compared to the traditional FCFS approach, and 2) to design an optimal route structure for the extended terminal area.A scheduling algorithm based on MILP has been successfully used in airport surface management due to its ability to optimize both the arrival/departure sequence and their scheduling [5].However, due to the expensive computational cost attributed to the branch and bound search algorithm, only a limited number of flights, typically less than 40, were allowed in the scheduling.The application of MILP-based scheduling algorithms to the dense terminal airspace operation has been less effective.Consideration of hundreds of flights in the span of a couple of hours typical of the dense terminal airspace operation dramatically increases the size of the MILP formulation and the corresponding computation burden easily exceeds the available computing resources.In a previous study, a heuristics-based MILP formulation was introduced in order to reduce computational burden and applied it to metroplex operations during their busiest operations [6].However, the cost for the branch and bound search is still expensive though not as prohibitive as in the original MILP formulation.The combination of GA-based heuristics and a Linear Programming (LP) method was proposed by Capozzi et al. [8] This is the method that is adopted in the current work.It explicitly separates the search for the optimum binary variables of route assignment and aircraft sequencing from the solution procedure for the continuous scheduling variables.GA-based MILP optimization scheme is able to find the optimal solution in significantly less computational time on the example problem considered.A central focus of the current work is the design of an optimal route topology in the extended terminal area using the GA-based MILP scheduler.This work will also use with flight trajectory uncertainty which is different from the previous work that held an unrealistic premise that detailed flight intent information including the transit times is known a priori.This would assume that one snapshot of the planning is sufficient to predict the scheduling performance.However, this is not true in the real-time simulation where uncertainties are present in flight trajectories and the schedules of the following flights are dynamically updated later in time based on the schedules of the previous flights.In order to overcome the above issues, we developed a dynamic planner framework that periodically updates the flight schedules to handle uncertainties.The dynamic planner consists of separate modules: a planner and a simulator.Uncertainty is implemented in the trajectory model of the simulator to account for aircraft arrival time errors.The planner adds an extra separation buffer at the scheduling points to cope with these inter-arrival errors.As a practical application of the heuristics-based, stochastic schedulers to dense terminal airspace simulation, a design of an optimal route structure in the extended terminal airspace area is carried out.Key design parameters are the number of merge points and their locations.First, we construct five distinct route structures with various merge topologies.The scheduling performance is evaluated for each topology using the stochastic FCFS heuristics-based scheduler and a dynamic planner framework is applied to each topology in order to validate the predicted scheduling performance.It is demonstrated that the largest separation amount required at all scheduling points is a dominant factor in scheduling performance.Finally, more generalized airspace topologies are considered.To investigate the sensitivities of the merge topology to the uncertainty modeling, three types of uncertainty distributions are considered: a constant, linear and quartic increment of the uncertainty per unit route length.The comparison of the corresponding scheduling results show that: 1) there exist optimal merge locations in the extended terminal airspace area, 2) the optimal merge locations tend to be positioned where large uncertainties are present so that pilot's control efforts reduce the local uncertainties at the control point.The rest of this paper is organized as follows: First, the basic formulation of the original MILP algorithms is explained, and the objectives and constraints for our route assignment and scheduling problem are defined.Second, a number of heuristics are introduced into the original MILP formulation as ways to reduce computational cost.Third, the dynamic planner framework with uncertainty modeling and extra separation buffers are detailed.Finally, a design of optimal route structure in terms of the merge topologies is discussed and followed by conclusions and future work.
+Problem FormulationWe apply a MILP-based formulation to the scheduling problem in the extended terminal airspace area, from the entry fixes to runway, and only the arrival portion of the scheduling will be considered.The plan consists of a time-constrained route for each aircraft in the demand set, with STA specifications at control points along each route, as dictated by separation rules, such that the resulting movement plan for all aircraft in the demand set is conflict-free.
+Basic Mixed Integer Linear ProgrammingA MILP scheduler has advantages in the scheduling problems over the traditional FCFS scheduler [9].The design variables can be freely chosen specifically to a problem as forms of binary variables and continuous variables.However, the computational cost of a branch and bound search algorithm is expensive and its scalability with the problem size is rather poor and a number of heuristics to reduce the computational burden will be explained in later sections.
+Initial Demand SetLet F, R, and P define a set of flights, available routes, and scheduling points, respectively, and N F , N R , and N P are the total number of flights, routes, and scheduling points in the demand set.F = {f j | 1 ≤ j ≤ N F } = {f 1 , f 2 , ..., f N F } R = {r j | 1 ≤ j ≤ N R } = {r 1 , r 2 , ..., r N R } P = {r j | 1 ≤ j ≤ N R } = {p 1 , p 2 , ...p N P }We further presume the existence of functions that select the subset of routes from R that are feasible for a given flight f ∈ F and the subset of points from P that are feasible for a given route r ∈ R.R f = {r k | all possible routes that flight (1) f can fly, k ∈ {1, 2, ..., N R (f )}} P r = {p k | ordered set of scheduling points (2)on the route r, k ∈ {1, 2, ..., N P (r)}}where N R (f ) is the number of all routes that flight f can fly, and N P (r) is the number of all the scheduling points on route r.Then, it can be easily inferred that a set of total routes and the points are the superset of R f and P r , respectively, and the relation is represented as follows:R = f ∈F R f and P = r∈R P r
+Decision VariablesGiven the notation in the previous section, the decision variables of interest for this problem can be defined:• A f,r -A binary route assignment variable that takes on the value of 1 if flight f is assigned to the route r and zero otherwise.• T f,r,p -A continuous variable representing the time that flight f is scheduled to cross the scheduling point p on route r, where f ∈ F, r ∈ R f , and p ∈ P r .• S f,f ,r,r ,p -A binary variable that takes on the value of 1 if flight f on route r is sequenced prior to flight f on route r at shared scheduled point p, where f ∈ F, r ∈ R f , r ∈ R f , and p ∈ P r ∩ P r = ∅.
+Objective and ConstraintsFor the purposes of this paper, the objective function is defined so as to minimize the total time required for all flights to reach the end of their route, i.e., the runway threshold:J = f ∈F r∈R f A f,r T f,r,p F(3)The problem constraints are as follows:• Assignment to Only One Route.Each flight can be assigned to one and only one route.• Crossing Time at Initial and Final Point.If a flight is assigned to a given route, then its start time on that route must be no earlier than the earliest feasible time on that route, T E f,r , and no later than the latest feasible time, T F f,r .• Ordering Constraint at Potentially Shared Scheduling Points.For each pair of flights and each pair of route assignment options that share a common scheduling point, the order in which the flights are sequenced at the common point must be uniquely specified.• Separation Constraint.At each common scheduling point, successive aircraft must be separated by a minimum time that is potentially intersection-dependent.• Transit Time Constraints.In order to be physically realizable, the travel time between scheduling points must be greater than the minimum possible travel time and should be bounded by the maximum "delayability" of the flight between the scheduling points.
+Heuristics-Based SchedulersOne of the drawbacks of the MILP-based scheduler is its prohibitive computational cost.However, observation of the branch and bound search procedure indicates a large portion of search time is spent on evaluating unrealistic combinations of route assignment and sequencing.Heuristics in the route assignment and the queuing can help eliminate some of the unrealistic search efforts and reduce the computational burden.In our study, we tried two types of heuristics.First, heuristics based on the FCFS scheduling behavior can eliminate a number of binary decision variables in the original MILP formulation.Second, the branch and bound search algorithm can be replaced by GA-based heuristics.
+Heuristics-Based Mixed Integer Linear ProgrammingBased on the observations of the FCFS scheduling strategy, assumptions are made on sequencing along certain route segments, such as merge portions.The details of the following heuristics were explained in our previous work on the metroplex operations and only brief summaries are listed here:• Precedence constraint heuristic.Sequences along the common segments of the routes do not change and follow the queuing from the upstream of the merging segments.• Windowing heuristics.A sequence change is not allowed for a pair with their earliest crossing times at the entry fixes separated by more than a certain amount, i.e., windowing value.Using windowing heuristics, resulting schedules are planned locally and subsequently corresponding to this value.• FCFS heuristics.The ordering at all scheduling points can be predetermined based on unimpeded transit times to a specific scheduling point, a runway or a entry fix.This strategy is equivalent to the FCFS scheduling ideas and the computational cost benefits are maximum.A scheduling performance almost equivalent to that of the original MILP formulation was obtained at only a fraction of the computation cost of the original MILP formulation.However, the cost for the branch and bound search is still expensive though not as prohibitive as in the original MILP formulation.
+Heuristics of Genetic AlgorithmsA GA is a stochastic search method widely used in numerous optimization areas and has ad-vantages in handling both discrete and continuous design variables [10].GAs are simple in mathematical formulation but are typically expensive in computation due to their stochastic search procedure.The idea of adopting GAs in the scheduling problem for determining the binary decision variables in lieu of the branch and bound searches has been suggested and used in surface management work [7] and metroplex operations [8].Once the binary variables of the route assignment and the sequencing are determined through GAs, computation of the STAs is carried out by an LP solver.Since the LP is very efficient in computation and takes less than a couple of seconds to schedule hundreds of flights, the idea of combining GAs with pure LP procedure is favored for the dense terminal airspace simulations.Advantages of the GA-based MILP planner include:• It allows for the solution to be seeded with a good initial guess, based on heuristics.• All individual candidate solutions are, by definition, feasible solutions -thus a usable solution is available at all times.• The solution tends to improve with computational time.• It naturally handles windowing heuristics.
+Individual Candidate Solution RepresentationEach individual candidate solution, or "an individual" in short, consists of two vectors: an assignment vector and a sequence vector.The length of each vector is equivalent to the number of flights in the demand set.Each element of the assignment vector represents a possible route assignment for a flight in the demand set.Given the structure of the route, a sequence of the scheduling points in the route is prescribed, and the sequence vector is defined at each scheduling point as a possible sequence of the flights passing through that point.Route assignments and the queuing are random based on the stochastic nature of the GAs.A constraint of no passing in the sequence vector along the common route segment is enforced to further reduce the computational cost.
+MutationTo maintain the diversity of the individual from one generation to the next, mutations are carried out at a specified probability in each generation to both vectors of route assignment and sequence.Due to the coupling between the assignment map and sequence map held within each individual, the mutation operators are applied sequentially.The assignment mutation simply consists of randomly replacing the route assigned for an individual with another value from the feasible set of routes for the flight.Then, the sequence map for each individual is updated based on the mutated route assignment map.A sequence mutation is applied to each scheduling point at a given route assignment.Swapping sequence is determined via probability.If a random number sample is less than the specified probability of mutating sequence, then the number of swaps are chosen from a (0, N max ), where N max is the maximum number of swaps.A random pair of flights that contain this scheduling point in their assigned routes swap their ordering relative to the current ordering.
+Fitness and SelectionFitness of each individual is defined as an objective function value and evaluated by solving the pure LP problem implied by its route assignment and sequence.The definition of the objectives and the imposition of the constraints are equivalent to those of the MILP formulation: earliest transit time limit, lower and upper bounds of the transit time via the specified speed controllability, and separation requirements specific to the aircraft type at the scheduling points.Once each individual in the population has been assigned a fitness, the selection of individuals to form the basis of the next generation is performed using a simple tournament selection scheme.A tournament scheme finds the best-performing P/2 number of individuals, where P is the size of the population, and those are selected as the basis of the next generation.These P/2 number of individuals are then mutated to form the population of size P to be evaluated.This cycle of fitnessselection-mutation is repeated until a specified number of generations are completed.
+Comparison of Computation Time and OptimalityThe optimality and the computation times are compared for the schedulers that were described in previous sections.The terminal route structure tested for comparisons is a binary route topology with double merges: 4 route options and 8 scheduling points (4 entry fixes, 2 merge points and 1 runway).The route topology is shown in Figure 1 with the entry fixes of WP 31 through WP 34 and the runway of WP 0 .The number of flights varies from 6 to 100.Computation times with respect to varying number of flights and the average delays of 8 flights are compared in Table 1.The expensive computational costs of the original MILP scheduler and the FCFS heuristics-based MILP scheduler prohibited computations for more than 8 flights and 40 flights, respectively and their values are represented as N/A in Table 1.Although the FCFS heuristics allows scheduling up to 40 flights, the computation time is not still satisfactory for the dense terminal airspace simulation.However, the speed-up of the computation time for the GAbased MILP planner is considerable even with 100 flights.It is observed from the additional scheduling of a larger number of flights using GA-based MILP planner that it is able to handle hundreds of flights in a few minutes.Therefore, it is concluded that the GA-based MILP scheduler is faster than the others and is more appropriate for the scheduling problem in the dense terminal airspace operation.
+Dynamic Planner FrameworkThe schedulers described in the previous section are deterministic.The transit time of any route segment was a function of aircraft type and the speed profile only, and the uncertainties along the routes were not considered in the planning.Even if the uncertainties are taken into account in the planner, the aforementioned schedulers are based on the premise that the uncertainties of each aircraft are known prior to the planning along any route segments.However, the STAs in the realtime simulation should be updated in a dynamic manner corresponding to the varying situations of weather, wind and off-nominal scenarios.A key to the realistic scheduling is a dynamic update of the STAs in a real-time trajectory model in consideration of the uncertainties in flight simulation.A dynamic planner framework is developed in our study by interactively integrating the trajectory simulation and the schedule planning for STA update.The framework consists of two components:• Simulator: This module is responsible for advancing time.It manages the creation of targets at specified location and time, and constructs the demand snapshot at a given instance of time.The module delegates to a trajectory model that handles the actual movement of flights along their most recent plan and blends motion between successive plans.Uncertainties in the transit time along the route segment and their propagations are implemented in the simulation module, which will be explained in later section.• Planner: This module is responsible for con-structing conflict-free plans for each aircraft in a given demand set.Each motion plan consists of a sequence of waypoints with an associated STA.Although any type of the scheduler can be used in the dynamic planner framework, the GA-based MILP scheduler is integrated into the current framework.Furthermore, to take into account the uncertainties in the trajectory simulation, an extra buffer additional to the desired separation is added to mitigate effect of uncertainties so that the resulting schedule remains conflictfree.The details of the additional buffer are discussed in later section.A specified amount of controllability is allowed in speed profile to maximize the scheduling performance and is implemented in both the planner and the trajectory model simulator.
+Trajectory Model SimulationFlight simulation of the trajectory model is made via subsequent communications with the planner.First, the dynamic planner starts from the pre-planning of the initial demand set.Given the speed profile and the ETA of each aircraft to the first schedule point, the initial STAs are computed by the scheduler at all scheduling points on the assigned route.The STAs are predicted such that they satisfy the constraints of the transit time bounds on the route segments and the desired separation at the scheduling points.The trajectory model periodically computes the distance from the current position to the next scheduling point.For our work, the update period is 60 seconds.With distance to the next scheduling point and the STA predicted by the planner, the target speed is calculated and checked whether it is bounded by the speed range specified in the original speed profile.Once the target speed is determined, then the trajectory simulator advances the flights by an update period.After the simulation, the aircraft position is recalculated and the earliest time to the scheduling point is updated.A subsequent planning cycle updates the STAs based on the most recent simulation results .This cycle of simulation and planning is iterated and advanced in time by the update period until all the flights arrive at the runway.An example of the cycles of simulation and planning is shown in Figure 2. Given a simple route structure shown in Figure 2(a), the STAs at the scheduling points of "Waypoint1" and "Run-way1" are updated at each update period of 60 seconds.Their convergence history is plotted in Figure 2
+ControllabilityTo delay or expedite an aircraft on its way to the next scheduling point, the controllability on any route segment is modeled such that it allows ±10 % speed variation.The corresponding transit time bounds are computed.This controllability is derived from the statistics of the aircraft flying with modern avionics and the onboard precision system [11].
+Uncertainty Modeling and PropagationAccurate prediction of uncertainties along an entire aircraft's trajectory is not trivial.It is a complicated function of space and time, which requires precise understanding of where and how much of the uncertainties are present and how they affect individual aircraft operations.However, the uncertainties in the runway arrival times are quantifiable from the statistics of the runway arrival times observed in a given duration of time.We can model the aircraft arrival time prediction errors at the runway by a normal distribution with the timeinvariant mean and standard deviation values.On the other hand, the uncertainty at the intermediate control points and route-merge points require a mathematical model of the uncertainty propagation mechanism along the routes.A simple linearized form is introduced in our simulation module: variance at each scheduling point is assumed to be proportional to the variance of the runway arrival time prediction errors when there are no control effort in between the scheduling point and runway.Uncertainty amount at any point on the route is scaled by the ratio of the intermediate route segment length to the entire route length from the entry fix to the runway.This presumes that the uncertainties grow longer along the longer routes since the flight is associated with longer transit time without control efforts.Based on the central limit theorem, we assume that the position error of aircraft at the scheduling point is approximately normally distributed.Then, the corresponding inter-arrival error of any pair is also normally distributed, and the mean and variance of inter-arrival errors are estimated by the following basic relationships: if X and Y are independent random variables that are normally distributed, then X + Y is also normally distributed, i.e., if X ∼ N (µ, σ 2 ) and Y ∼ N (ν, τ 2 ) and X and Y are independent, then aX + b ∼ N (aµ + b, a 2 σ 2 ) and X + Y ∼ N (µ + ν, σ 2 + τ 2 ).The means and the variances of the X and Y are the µ and ν, and σ and τ , respectively.The validity of the above relations holds best when the independence of two variables, X and Y , is relatively well guaranteed.The inter-arrival error of a pair of aircraft is a complicated function of many factors such as precision errors in navigation and weather including wind.We assume for simplicity in our trajectory simulation that the wind effect during the traffic simulation is rather constant.Direction and magnitude of the wind are relatively non-changing on each aircraft throughout the whole simulation, then we can treat the wind effect as a constant that is freely addable / subtractable to/from the standard deviation of the position error of each aircraft at all scheduling points, and the above relation holds relatively well.
+Additional Separation Buffer due to UncertaintiesAircraft arrival time errors and the corresponding inter-arrival error in a pair are likely to cause violations of the desired separation.In order to ensure desired separation is maintained in spite of arrival time error, extra buffers are added to the original desired separation [12].The amount of the additional buffers is determined from the probability of the inter-arrival errors and its normal distribution shown in Figure 3.If we choose a value, Z, for an additional buffer such that the cumulative probability corresponding to Z coincides with a specified confidence level, 90% for the current work, then we can say that the separation requirement in any pair will be satisfied under uncertainty with 90% confidence and be violated with 10% tol- erance.A brief graphical explanation is shown in Figure 3.This sets the additional buffer value as 1.645σ, where σ is the standard deviation of the inter-arrival error distribution.€ f 1 , σ 1 € f 2 , σ 2The above can be expressed mathematically as follows.Standard deviations of the arrival time errors of a leader and a follower in a given pair are denoted as σ leader and σ follower , respectively as in Figure 4.The amounts are scaled from runway standard deviation in proportion to the ratio of the local route segment to the entire route from runway.Assuming the probability of the position errors of a leader and a follower are independent of each other, the standard deviation of the interarrival error is assumed to be √ σ leader + σ follower .A corresponding additional buffer is set as 1.645 σ 2 leader + σ 2 follower for a 90 % confidence interval based on the normal distribution of the inter-arrival errors.A simple mathematical formulation of total amount of buffers is expressed in following Equation.
+Application: Optimal Route Structure Under UncertaintyA practical application of the developed schedulers and the dynamic planner framework is shown in this section.A design of an optimal route structure under uncertainty in the extended terminal airspace area is carried out to improve scheduling performance and, thus, to best utilize the limited airspace resources.First, key parameters for a route structure design are identified.A total of five example route structures are constructed with varying design parameters.A static FCFS heuristics-based MILP planner is used to analyze the scheduling performance of each notional route structure, but the results are validated by using the dynamic planner framework.Based on the analyses of the scheduling performance of the notional route structures, more general cases of various merge topologies are considered subsequently.
+ParameterizationA route structure consists of such parameters as the location and number of entry fixes, runways and merge points as well as route segment lengths.A demand set is also critical in scheduling performance.The demand set defines relevant flight information including the total number of flights, arrival time at entry to the route structure, and traffic duration time.In fact, the scheduling performance is very sensitive to the demand set.A fully saturated demand set, i.e., one which has no periods of low demands, is used to isolate the scheduling performance from the effects of the route structure alone.A fully saturated traffic flow is consistent with dense terminal airspace operation whereby demand exceeds capacity for extended periods of time.In this way, the runway capacity is always exhausted and the number of runways becomes no longer a parameter of the airspace topology.An entry fix topology, i.e., the total number of entry fixes and their locations, is also assumed to be given to facilitate the fully saturated traffic flow.Thus, the main parameter in our design study is the merge point topology, i.e., the location and number of the merge points.
+Numerical Test IFirst, a numerical experiment is performed on five route structures having different topologies with varying numbers and locations of the merge points.Figures 5 and6 have a single merge point whereas Figures 7 through 9 have two merge points.The locations of the merge points are moved in order to vary the ratios of route segment lengths in a given topology and thus vary the uncertainty distribution along the route.These topologies can also be defined by the parameters of the route segment length and the merge angle between two routes.For example, Figure 7 through 9 can be defined by varying the parameters of a, b, c, α, and β as shown in Figure 1, where a, b and c represent the route segment length, and α and β represent angles between two merging routes.A FCFS heuristics-based MILP is used for computing the scheduling performance of each topology.In the cost comparison of the schedulers explained in previous sections, a demand set of 80 flights is not trivial in scheduling even for the heuristicsbased MILP scheduler.Total computation time grows very quickly, especially for a stochastic case where hundreds or thousands of Monte Carlo simulations are performed.Thus, for this numerical experiment, the route is pre-assigned for each aircraft and the orders at the merge points are predetermined based on the the unimpeded transit times from the entry fix to the runway.For a stochastic scenario, an uncertainty model is directly implemented in the scheduler, and we do not employ a dynamic planner for this preliminary numerical experiment.However, a more realistic validation of these five airspace topologies is carried out by the dynamic planner and analysis results will be shown in the following section.Initially, a total of 80 flights pass through the entry fixes equally divided into four streams and follow their pre-assigned routes during a short period of 100 seconds.This short duration time ensures fully saturated air traffic.Four types of weight class categories are used: "heavy", "B757", "large", and "small".A majority of aircraft in our demand set, more than 90%, belong to the "large" type, indicative of today's operations.The total amount of separation required at each scheduling point is the sum of the minimum desired separation and the extra additional buffer to mitigate uncertainty.The quantification of this amount is done using the formulation of Equation 4. The amount of the desired separation is based on the weight classes of the leader and the follower in a given pair, and the values are specified in Table 2. Airspeed is assumed to be linearly decreasing along the routes in the extended terminal airspace area and its value at each scheduling point is summarized at the tables in Figure 5 through 9.Airspeeds at the entry fixes are fixed at 250 kts.Additionally, the desired separation of the entry fix was 5 nmi to model the separation required for en route airspace.Previously collected data indicates that a standard deviation of uncertainty (in the aircraft arrival time errors) at the entry fix is found to be approximately 30 seconds.The additional uncertanty-related separation buffer required at the entry fix in order to mitigate this uncertainty is calculated in a similar way for the merge points and runway.Given the prescribed uncertainty and its corresponding separation buffer, the FCFS heuristicsbased MILP scheduler simulates both the deterministic and stochastic scenarios for all airspace topologies shown in Figure 5 through 9.A total of 500 Monte Carlo simulations were carried out for the stochastic simulations.The separation buffer and total separation are computed for each case and summarized in the tables in Figure 5 through 9.The resulting average delays are shown in Table 3.The average delay is calculated as a difference between unimpeded transit time and actual transit time.The value of average delay is an artifact of the fully saturated traffic scenario.The key result is the performance improvement.First, for the deterministic scenarios where no consideration of uncertainty is made, the third column in Table 3   the double merge cases, differences in the average delays among all cases are very minor, less than 1%.This can be explained from the formulation in Equation 4, where the amount of total separation is solely a function of airspeed alone when there is no uncertainty.As airspeed gradually decreases towards the runway, the runway always requires the biggest separation of all the scheduling points.This makes the scheduling performance largely insensitive to the particular upstream route structure.The controllability of each aircraft does not affect the scheduling performance either, as most of the aircraft have to absorb delays propagated from the preceding aircraft.Second, for the stochastic scenarios, an improvement of 6.7 % is shown in Case 5, when comparing the single merge and the double merge topologies.Uncertainty creates perturbations in the transit times, and some of the aircraft can exploit their controllability to fill the gaps in a pair created by the uncertainty.The results from both deterministic and stochastic cases in Figure 5 and 6 are interesting and informative.The actual, average spacings between all pairs are extracted from the scheduling results for both cases and shown in blue in tables in Figure 5 and6.A careful comparison of five cases of the resultant average spacings and the required separation indicates that the average spacing at each scheduling point is dominated by the largest separation required along the route.This is shown in red in tables in Figure 5 and6.This observation suggests that in a fully saturated traffic flow, the traffic becomes steady with acceleration and deceleration adjusted by the controllability that is allowed in each route segment and incurs approximately the same amount of spacings in any pairs.
+Validation Using Dynamic PlannerThe scheduling results in Section of Numerical Test I are validated using the dynamic planner framework.A GA-based MILP is employed for the planner of the framework for a comparison with the FCFS heuristics-based MILP.Unlike the static planner used in Section of Numerical Test I, a dynamic planner involves iterative interactions between the planner and the simulator as time advances.An update period of 60 seconds was used in our validation.In each dynamic planning cycle, the simulator tries to track the STAs provided by the planner at all scheduling points, and the planner creates new schedules as a result of updated aircraft positions and ETAs from the simulator.A realistic demand set is critical in the dynamic planner so that the trajectory model can be simulated based on the operationally reasonable and realistic schedule plans.For this reason, rather than using fully saturated traffic as in the Numerical Test I, initially well-separated traffic flows, by 5 nmi in any pairs with random deviation ranging from -20% to +20%, are used for the dynamic planner.This results in hour-long traffic flows for the same number of flights.Also the initial departing point for all aircraft is fixed at about 100 nmi away from the runway and is almost aligned with the freeze horizon.The validation results are summarized in Table 4.An average delay value is, again, defined as a difference between the unimpeded transit time and the actual transit time from the entry fix to the runway.Compared to the values in Table 3 which used a fully saturated traffic flow, average delay values become more reasonable with the maximum value less than five minutes.It should be noted that the dynamic planner is more computation-intensive than the static scheduling planner, as the specified plan update period cannot be set too large in a real-time simulation.A wall-clock CPU time per dynamic planning takes about an hour with 60 seconds update period.The dynamic planning requires hundreds of cycles of simulation and planning for aircraft to travel the entire route of 100 nmi in length.Thus, the computation time of the dynamic planner for the stochastic case becomes very expensive with Monte Carlo simulation.The results for the stochastic cases shown in Table 4 are obtained from only 100 Monte Carlo simulations.More simulations are planned as part of the future work.It can be concluded from the average delay results shown in Table 4 that Case 5 is the bestperforming airspace topology and it has a performance improvement, compared to Case 1, of approximately 50% in the deterministic case and approximately 30% in the stochastic scenario.Although the sample size of the Monte Carlo simulations in the stochastic scenarios is not big enough to make the conclusion more trustworthy, the standard deviation of the 100 Monte Carlo samples are as little as 15 seconds resulting in 5% tolerance that is far less than our percentile performance improvement.It should also be noted that for a less saturated traffic flow, an improvement in the scheduling performance is more dramatic compared to the 6.7% in Table 3.This can be explained by the fact that in fully saturated traffic flow, all the aircraft quickly exhaust their controllability and are assigned their slowest speed profile.That is not the case with less saturated air traffic, and the benefits from the controllability are more dramatic in this case.It should also be noted that the same trend of the static planning in Numerical Test I is shown in the validation results: an airspace topology with small separation requirements at the scheduling points is more favorable in the scheduling performance.
+Numerical Test IIBased on the results in the previous sections that the scheduling performance is heavily dependent on the amount of maximum separation at the scheduling points, a simple numerical experiment is carried out to investigate more general variations of the airspace topology.Relationships among the component separation amounts at the scheduling points are analyzed when the merge points move around in the extended terminal airspace area.The airspace topology is simplified to a circle with 40 nmi radius as shown in Figure 10.The points, WP1 and WP2, represent the merge locations that can move circumferentially at a relatively constant radial distance away from the runway, and the routes can be merged at these locations.Radial distances, x and x + y, are also allowed to vary within radius bounds: 0 < x < R and 0 < y < R -x, where R is the radial distance from the runway to the entry fix.The movement of the merge points, WP1 and WP2, are shown in Figures 10(a) and 10(c), and the corresponding example airspace topologies are plotted in Figures 10(b) and 10(d), respectively.As WP1 and WP2 move about in the extended terminal airspace area, the entire route length in a particular airspace topology from the entry fix to the runway and the corresponding transit time may change.However, any large variations were excluded in the transit times by locating the entry fix such that the the entire route from the runway to the entry fix does not deviate much from the original radial lines.For the airspace topology with a single merge as in Figures 10(a) and 10(b), total separation amounts at the runway and the WP1, are computed from Equation 4 and plotted as solid lines with symbols in Figure 11(a).The two dashed lines represent the desired separation amount using the nominal speed profile, and the uncertaintyrelated, extra separation buffers are plotted as dotted lines.The traces of the maximum separation with respect to the varying merge locations are shown in Figure 11(b).Figure 11(b) implies that there exists an optimal merge location somewhere around 5 or 6 nmi away from the runway.The case of double merges inside the extended airspace area as in Figures 10(c) and 10(d) shows similar trend as the single merge case.At the given downstream merge point location, x, which can move from the runway to the entry fix, the upstream merge location, y, also traverses between downstream merge point and the entry fix.The plot of the maximum separation buffer of all scheduling points, i.e., runway, downstream and upstream merge points, are almost identical to the plot of Figure 11(b) and is omitted in the paper; how-   ever, this indicates that the separation at the runway typically requires the biggest amount and in other cases the downstream merge point requires a larger buffer than the upstream merge point.The exact optimal locations of merge points are less meaningful in our analysis, as they can vary depending on the underlying airspeed profile, uncertainty quantification and propagation model, and the assumed additional separation buffer due to uncertainty.Possible changes to the optimal merge locations can be inferred from Figure 11(a).Desired separation represented as the dashed lines in blue and red are rather non-changing as these are solely determined by the airspeed topology alone, and the assumption of monotonically decreasing airspeed in the extended terminal airspace area is reasonable.On the other hand, the modeling of uncertainty and its propagation in the extended terminal airspace area is still an area of active research that requires extensive studies on how the uncertainties are distributed or propagated along the routes.How temporal or spatial deviation from its predicted trajectory behavior is quantified over time and distance and translated into the time-based scheduler is a difficult and yet very important topic in the scheduling and real-time simulations.
+Various Uncertainty Propagation ModelsA brief sensitivity study of the optimal merge locations is carried out with respect to the different models of uncertainty distribution and propagation.The results in previous sections assumed that the increment of the uncertainty per unit route segment is constant and the resultant uncertainty is linearly proportional to the route segment length.The plot of constant uncertainty increment and the corresponding optimal merge location is plotted in Figure 12.The maximum of the two solid lines with the symbols represents the traces of the largest separation amount of all the scheduling points.The x-axis represents the location of the downstream merge point.Once the downstream merge point is located at the predicted optimum point, the separation amount at the upper merge point appear to be smaller than the one at the downstream merge point.On the other hand, if we put more weights in the uncertainty towards the entry fixes, different optimal merge locations are predicted.An assumption of a linear increment as in Figure 13 moves the optimal merge location slightly towards the entry fix, about 10 nmi away from the runway.If we put more weights in the entry fix boundary area as in Figure 14, at a quartic increase rate for example, the optimal merge location falls in the regions farther from the runway, about 23 nmi away.From these simple models of the uncertainty quantification and propagation, it is concluded that the optimal location of the merge point is where a large amount of uncertainties are present, and the merge point plays a role in reducing the uncertainties in that region by enforcing the pilot's control efforts to meet the suggested STA at that point.
+ConclusionsMILP-based optimization algorithms were used in our scheduling and route assignment problem, and a number of heuristics were introduced into the original formulation to keep its computational cost realizable in the dense terminal airspace operations.FCFS-heuristics and GA-based heuristics were adopted to reduce the computational burden, and the resultant computational costs and scheduling results were compared with the original MILP solutions.The GA-based MILP planner appears to be very efficient without loss of optimality.To take into account uncertainty propagation in the route structure, a dynamic planner framework is developed.The dynamic planner consists of the modules of the planner and the trajectory simulator.The STAs are updated in a dynamic manner via the interaction of the planner and the simulator throughout the whole simulation.An uncertainty model is implemented in the trajectory model of the simulator and the extra separation buffers are added at the scheduling points in the planner.As a practical application of the proposed schedulers, an investigation of the optimal route structure under uncertainties in the extended terminal airspace was carried out.A constant uncertainty increment was assumed along the route, ensuring the uncertainty amount grows linearly in proportion to the route segment length.After analyzing the airspace topologies with varying merge topologies, a route structure having the least maximum separation at the scheduling points has shown the best scheduling performance.These results were validated with a dynamic planner framework with a more reasonable demand set.Finally, airspace topologies with various uncertainty distribution models were tested: a constant, linear and quartic unitincrement of the uncertainty along the route segment.It is shown that the optimal merge point should be positioned to bound the growth of the uncertainty-related separation buffer such that the maximum total separation at any point along the route is minimized.This fact indicates that there exists a likely optimal merge topology.The optimal merge topology is still tuned for a of uncertainty propagation models while the exact topology did vary.Figure 1 .1Figure 1.Example route topology
+Trajectories of STAs and Actual Time of Arrival (ATA): STAs in solid lines and ATA as the last point of the line.
+Figure 2 .2Figure 2. Example of dynamic planner framework
+(b), and STA updates are shown by the triangles.Unlike the static planner, the STA values are updated at each update period and finally coincide with the Actual Time of Arrival (ATA) values at the scheduling points.The speeds along the routes change correspondingly in each update period.
+Figure 3 .3Figure 3. Probability distribution of position
+Figure 4 .4Figure 4. Uncertainties of position errors in a pair at the merge point (σ1=σ leader and σ2=σ follower .)
+Figure 10 .10Figure 10.Simplified airspace topology of the exterminal area with varying merge point locations
+Figure 12.Constant distribution of uncertainty
+Table 1 .1Computation times (in seconds) to schedule different number of flightsNumberAverageof67840100 delayfights(8 flights)MILP6.798.3 1258.4 N/A N/A 8.23MILP +FCFS2.133.2 419.3 2040 N/A 8.38heuristicsMILP +GA≤ 1.0 1.02.223.4 50.0 8.58heuristics
+t sep tot = t sep desired + t sep σ=d sep desired V+ 1.645 σ 2 leader + σ 2 follower ,where V is airspeed and t sep tot represents totalamount of separation requirement. Terms oft sep desired , d sep desired and t sep σ represent a desiredseparation in time, a desired separation in dis-tance, and an uncertainty-related, additional buffer,respectively.
+Table 2 .2Desired separation (nmi)leader PP P P P follower heavy B757 large small P P P Pheavy4555B7573334large3334small3333
+Table 3 .3Predicted average delays (seconds).AveragePerformanceAveragePerformanceTopology Casesdelayimprovementdelayimprovement(Deterministic) (w.r.t Case 1)(Stochastic) (w.r.t Case 1)singleCase 13363.41.0 %3834.81.0 %mergeCase 23370.8-0.2 %3918.2-2.1 %doubleCase 33348.10.45 %3571.66.8 %mergeCase 43343.60.6 %3564.87.0 %Case 53352.00.3 %3579.76.7 %
+Table 4 .4Average delays (seconds) validated by dynamic planner framework for stochastic case.Topology Cases Update DeterministicPerformanceUpdate StochasticPerformancecyclesscenarioimprovementcyclesscenarioimprovement(w.r.t. Case 1)(w.r.t. Case 1)singleCase 1304249.11.0 %350291.01.0 %mergeCase 2257235.05.6 %329283.02.7 %doubleCase 3228107.2656.9 %294228.221.6 %mergeCase 4232108.3156.5 %295213.526.6 %Case 5247109.1756.0%294199.331.0 %
+		
+		
+			
+
+				
+
+
+	
+		Air Traffic Controller Ability Requirements in the U.S. National Airspace System
+		
+			HSwenson
+		
+		
+			RBarhydt
+		
+		
+			MLandis
+		
+		10.4324/9781315242538-12
+		Ver. 6.0
+	
+	
+		Staffing the ATM System
+		Moffett Field, CA
+		
+			Routledge
+			2006
+			
+		
+	
+	Swenson, H., Barhydt, R., and Landis, M., "Next Gen- eration Air Transportation System (NGATS) Air Traffic Management (ATM)-Airspace Project," NASA Tech. Re- port, Ver. 6.0, Moffett Field, CA, 2006.
+
+
+
+	
+		Flexible, Performance-Based Route Planning for Super-Dense Operations
+		
+			JosephPrete
+		
+		
+			JimmyKrozel
+		
+		
+			JosephMitchell
+		
+		
+			JoondongKim
+		
+		
+			JasonZou
+		
+		10.2514/6.2008-6825
+	
+	
+		AIAA Guidance, Navigation and Control Conference and Exhibit
+		Honolulu, HI
+		
+			American Institute of Aeronautics and Astronautics
+			Aug. 2008
+		
+	
+	Prete, J., Krozel, J., Mitchell, J. S. B., Kim, J., and Zou, J., "Flexible, Performance-based Route Planning for Super- Dense Operations," AIAA Guidance, Navigation, and Con- trol Conference, Honolulu, HI, Aug. 2008.
+
+
+
+	
+		A Concept for Robust, High Density Terminal Air Traffic Operations
+		
+			DougIsaacson
+		
+		
+			JohnRobinson
+		
+		
+			HarrySwenson
+		
+		
+			DallasDenery
+		
+		10.2514/6.2010-9292
+	
+	
+		10th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference
+		Fort Worth, Texas
+		
+			American Institute of Aeronautics and Astronautics
+			Sep. 2010
+		
+	
+	Isaacson, D. R., Robinson III, J. E., Swenson, H., and Denery, D. G., "A Concept for Robust, High Density Termi- nal Air Traffic Operations," 10th AIAA Aviation Technol- ogy, Integration, And Operations (ATIO) Conference, Fort Worth, Texas, Sep. 2010.
+
+
+
+	
+		Evaluating Concepts for Metroplex Operations
+		
+			John-PaulClarke
+		
+		
+			LilingRen
+		
+		
+			EvanMcclain
+		
+		
+			DavidSchleicher
+		
+		
+			SebastianTimar
+		
+		
+			AdityaSaraf
+		
+		
+			DonaldCrisp
+		
+		
+			RichardGutterud
+		
+		
+			TarynLewis
+		
+		
+			TerenceThompson
+		
+		
+			CarolynCross
+		
+		
+			TarynLewis
+		
+		
+			MichaelMadson
+		
+		10.2514/6.2010-9249
+		NASA/CR-2010-000000
+	
+	
+		10th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference
+		
+			American Institute of Aeronautics and Astronautics
+			2010
+		
+	
+	Clarke, J. P., Ren, L., Schleicher, D., Thomson, T., Cross, C., Lewis, T. B., "Characterization of and Concepts for Metroplex Operations," NASA/CR-2010-000000 , 2010.
+
+
+
+	
+		An Optimization Model For Reducing Aircraft Taxi Times at the Dallas Fort Worth International Airport
+		
+			SRathinam
+		
+		
+			JMontoya
+		
+		
+			YJung
+		
+		
+			2008
+		
+	
+	26th International Congress of the Aeronautical Sciences
+	Rathinam, S., Montoya, J., and Jung, Y., "An Optimiza- tion Model For Reducing Aircraft Taxi Times at the Dal- las Fort Worth International Airport," 26th International Congress of the Aeronautical Sciences, 2008.
+
+
+
+	
+		Towards Optimal Routing and Scheduling of Metroplex Operations
+		
+			SAtkins
+		
+		
+			BCapozzi
+		
+		
+			SChoi
+		
+	
+	
+		th AIAA Aviation Technology, Integration, and Operations Conference (ATIO)
+		Hilton Head, South Carolina
+		
+			Sep. 2009
+		
+	
+	Atkins, S., Capozzi, B., and Choi, S., " Towards Opti- mal Routing and Scheduling of Metroplex Operations ,"9th AIAA Aviation Technology, Integration, and Operations Conference (ATIO), Sep. 2009, Hilton Head, South Car- olina.
+
+
+
+	
+		Planning techniques for airport ground operations
+		
+			JGarcia
+		
+		
+			ABerlanga
+		
+		
+			JMMolina
+		
+		
+			JABesada
+		
+		
+			JRCasar
+		
+		10.1109/dasc.2002.1067902
+	
+	
+		Proceedings. The 21st Digital Avionics Systems Conference
+		The 21st Digital Avionics Systems Conference
+		
+			IEEE
+			2005
+			18
+			
+		
+	
+	Garcia, J. Berlanga, A., Molina, J. M., Casar, J. R., "Op- timization of airport ground operations integrating genetic and dynamic flow management algorithms," Al Communi- cations, Vol. 18, N. 2, pp. 143-164, 2005.
+
+
+
+	
+		Hybrid Optimization Approach To Traffic Management
+		
+			SAtkins
+		
+		
+			BCapozzi
+		
+	
+	
+		10th AIAA Aviation Technology, Integration, and Operations Conference (ATIO)
+		Fort Worth, Texas
+		
+			Sep. 2010
+		
+	
+	Atkins, S., and Capozzi, B., " Hybrid Optimization Ap- proach To Traffic Management ,"10th AIAA Aviation Tech- nology, Integration, and Operations Conference (ATIO), Sep. 2010, Fort Worth, Texas.
+
+
+
+	
+		A mixed integer linear programming model for dynamic route guidance
+		
+			DavidEKaufman
+		
+		
+			JasonNonis
+		
+		
+			RobertLSmith
+		
+		10.1016/s0191-2615(98)00013-7
+	
+	
+		Transportation Research Part B: Methodological
+		Transportation Research Part B: Methodological
+		0191-2615
+		
+			32
+			6
+			
+			1998
+			Elsevier BV
+		
+	
+	Methodology
+	Kaufman, D. E., Nonis, J., and Smith, R. L., "A Mixed Integer Linear Programming Model for Dynamic Route Guidance," Transportation Research: Part B, Methodology, Vol. 21 pp. 431-440, 1998.
+
+
+
+	
+		Genetic algorithms in search, optimization, and machine learning
+		
+			D-EGoldberg
+		
+		10.5860/choice.27-0936
+	
+	
+		Choice Reviews Online
+		Choice Reviews Online
+		0009-4978
+		1523-8253
+		
+			27
+			02
+			27-0936-27-0936
+			1989
+			American Library Association
+		
+	
+	Goldberg, D-E., "Genetic Algorithms in Search, Opti- mization and Machine Learning," Addison-Wesley, 1989.
+
+
+
+	
+		Airborne-Managed Spacing in Multiple Arrival Streams
+		
+			BBarmore
+		
+		
+			TAbbott
+		
+		
+			KKrishnamurthy
+		
+	
+	
+		24th International Congress of the Aeronautical Sciences
+		Yokohama, Japan
+		
+			Sep. 2004
+		
+	
+	Barmore, B., Abbott, T, and Krishnamurthy K, " Airborne-Managed Spacing in Multiple Arrival Streams ,"24th International Congress of the Aeronautical Sciences, Sep. 2004, Yokohama, Japan.
+
+
+
+	
+		Airport Arrival Capacity Benefits Due to Improved Scheduling Accuracy
+		
+			LarryAMeyn
+		
+		
+			HeinzErzberger
+		
+		10.2514/6.2005-7376
+		AIAA 2005-7376
+	
+	
+		AIAA 5th ATIO and16th Lighter-Than-Air Sys Tech. and Balloon Systems Conferences
+		Arlington, Virginia
+		
+			American Institute of Aeronautics and Astronautics
+			September 2005
+		
+	
+	Meyn, Larry. A., and Erzberger, H., " Airport Arrival Capacity Benefits Due to Improved Scheduling Accuracy ", AIAA Aviation, Technology, Integration and Operations Conference (ATIO), Arlington, Virginia, September 2005, AIAA 2005-7376.
+
+
+
+	
+		TRAC Trial Planning and Scenario Generation to Support Super-Density Operations Studies
+		
+			ToddJCallantine
+		
+		10.2514/6.2009-5836
+		AIAA 2009-5836
+	
+	
+		AIAA Modeling and Simulation Technologies Conference
+		Chicago, IL
+		
+			American Institute of Aeronautics and Astronautics
+			August 2009
+		
+	
+	Callantine, T. J., "TRAC Trial Planning and Scenario Generation to Support Super-Density Operations Studies," AIAA Modeling and Simulation Technologies Conference, Chicago, IL, August 2009, AIAA 2009-5836.
+
+
+				
+			
+		
+	

file143.txt

@@ -0,0 +1,1190 @@
+
+	
+	
+		
+IntroductionGround-based flight simulation has a variety of aeronautics applications such as training, research and development, and accident investigations.Safety and cost savings relative to flight test are the most appealing virtues of using groundbased flight simulation.With the advanced technology in digital computing and image generation, the realism and fidelity of today's flight simulators have improved significantly from the old blue box of the thirties.However, the effectiveness of ground-based flight simulation is difficult to determine.Simulation may be physically similar to flight, e.g., same cockpit layout, control force feel, and tasks.But the fundamental human/machine interaction, specifically in visual-motion interactions, is often very different.In the extreme, the specific force cues are missing from fixed'base flight simulators.Motion-based flight simulators do provide onset specific force cues but can have visual-motion cueing conflicts due to limited travel.Pilots, therefore, must adjust their strategy in using the simulation cues to perform the given tasks.Since humans are adaptive and optimizing in nature, unless these characteristics can be quantified, the effectiveness of flight simulation, e.g., transfer of training and transfer of handling qualities results, with respect to simulator missions cannot be predicted.The presumption is that, if one can develop a comprehensive understanding of how pilots perceive aircraft states and task parameters from available simulation cues, and how they process and react to that information in given tasks, an analytical methodology can be developed to characterize that behavior process.It may then be used to interpolate and extrapolate results learned from ground-based flight simulations.Thus the effectiveness issue can be determined.This paper reviews several critical elements associated with ground-based flight simulation's visual and motion cues that are most influential to the human/machine interface.The objectives are to summarize significant results from past studies and to identify future research directions for determining ground-based flight simulation effectiveness.
+Visual Cues. Visual cues are the single most important simulation cues in all ground-based flight simulators for determining the orientation and position of the simulated aircraft.From the out-the-window (OTW) scene, instruments, and perhaps a Head-Up Display (HUD) and/or Head-Down Display (HDD), pilots observe the simulated aircraft states to develop appropriate actions to perform the tasks.
+Transport DelayTransport delay has been a critical factor in visual cueing perception.The delay reflects how fast the image generator or displays can present the simulated aircraft's response due to pilot's control inputs.Time delay has been found to have significant effects on pilot workload in several studies. 1 ' 2 FAA Advisory Circulars have suggested no more than 150 msec delay for transport flight simulators, 3 and 100 msec delay for helicopter flight simulators 4 where delay is defined as the time interval taken from the control input to change in the OTW scene.The current technology has improved the transport delay contributed by the image generator alone to under 50 msec (e.g., about 50 msec for E&S ESIG 4530 5 and about 25 msec for SGI Onyx 6 ).Time delay due to integration steps in the real-time digital computer, i.e., from accelerations to rates, and rates to displacement, has also improved significantly due to using predictive algorithms 7 and faster real-time computer processors.The technology allows modern simulators to easily meet those recommended criteria.
+Visual ResolutionThe ability to distinguish and recognize an object or a target from OTW is primarily dependent on the contrast and resolution of the displayed objects and targets.Level of contrast depends on the display system technology, e.g., collimation through lens, and projection through light valves.The limiting factor for resolution is the number of polygons that image generators can generate, and the performance and efficiency of all visual system components in the pipeline.The resolution requirement also depends on the distance (range) and flight tasks.Brown 8 showed a process to determine the required resolution for a TA-4J in an aerial combat maneuver.Larsen 9 used Johnson's Criteria, 10 which are dependent on task level, i.e., detection, shape orientation, shape recognition, and detail recognition, to develop the required resolution in line pairs for an air combat training.Polygon count, though convenient, is not a good measurement of the resolution nor provides a good comparison between systems since each manufacturer has its own polygon definition.A recommended measurement common in industry is to use the Modulation Transfer Function (MTF) 11 which combines the contrast and resolution as a single parameter to determine the entire display pipeline performance, 12 i.e., from image generator to display.Therefore, a logical recommendation to quantify the display system resolution performance is to develop a standard test pattern and measuring procedure, and then use MTF as an objective measurement.
+Scene ContentOut-the-window scene content plays an important role in pilot's perception in estimating position, attitude, and their rate of change.Lintern, 13 in a simulation bombing training study with 42 student pilots, compared results in dusk condition with limited scene features and from day light with extensive scene features.He found that scene content produced significant effects in pilots' bombing error performance.Lintern has also found that training effectiveness improves with increases in visual scene detail. 13" 15 In a separate bombing study with 32 pilot subjects, Lintern 16 found that scene content, i.e., landscape vs. grid pattern, has a significant effect on pilots pitch control performance and transfer of training, all in favor of the landscape case.The shape of objects and application of texture also play significant roles.Kleiss,17 in his discussions of visual scene properties for low-altitude flight, found that change of global optic flow rate and change of optical edge rate are useful for perceiving change in speed.In a visual environment at a speed of 600 knots and 150 ft above ground with 21 A-10 pilots, DeMaio 18 found objects are effective for estimating altitude.He suggests that a density of about 12 to 15 objects per square mile is necessary and sufficient for maintaining altitude.The same study also finds equivalent cueing effectiveness can be provided by a two-dimensional texture pattern.Kellogg 19 in his investigation with 10 experienced C-130 pilots found that texture had a significant and positive effect in centerline positioning and altitude control in an assault landing task.That conclusion is consistent with findings from DeMaio 18 and Kraft. 20ditional studies have been recommended by DeMaio to develop better understanding of what types of texture patterns contribute to effective altitude cueing.Kleiss indicates variations in terrain shape and object size or spacing are important parameters for the simulator designer, and suggests further investigation to determine level of ' terrain resolution requirements.Visual Field-Of-View (FOV1 The effectiveness of FOV is a very practical issue for ground-based flight simulators.For realism purposes, one would naturally keep the visual cueing environment as close to the simulated aircraft as possible, i.e., wide FOV for most of simulated aircraft.From the visual self-motion perspective, peripheral vision is also important. 21However, wide FOV can be an expensive proposition.It typically demands a high cost in image generation systems and monitors even if added weight and space are not issues.In a single roll degree-of-freedom (DOF), Moriarty 22 has shown peripheral vision has significant effects in a compensatory tracking task when subjects using a sidestick to control higher order control element dynamics (~k/s 3 ).With peripheral vision, results showed that pilots were able to provide more phase lead in the frequency range below the crossover frequency. 23In the same study, however, he did not find peripheral vision had a significant effect when a lower order control element (~k/s 2 ) was used.This suggests that wide FOV may have significant benefit when the simulated aircraft dynamic characteristics become higher in order.A review of the effectiveness of wide FOV in multiple degrees-of-freedom flight simulations has produced mixed results.Several studies 13 ' 19 ' 24 " 27 have been identified which cover a range of tasks and types of aircraft.These investigations all have used a large number of test subjects and used statistical analysis to determine the significance of their results, as summarized in Table 1.As shown, results from the same flight simulator differ as tasks and test subjects varied which suggest more systematic investigation in determining the effectiveness of FOV is required.
+Man/Machine InteractionEffectiveness of motion vs. no-motion in ground-based flight simulations is a heatly debated issue.Platform motion has been shown to improve pilot-vehicle performance when compared with fixed-based flight simulators.Using a roll attitude stabilization task in hover, Stapleford 28 found that motion cues increased pilot phase lead and led to higher pilot crossover frequency and gain.In a dogfight scenario investigating the effects of motion vs. no-motion, Jex 29 found that under the full motion case test subjects were able to provide more phase lead at low American Institute of Aeronautics and Astronautics frequency which helped avoid drifts and overshoots in target tracking, and to provide higher gain (a factor of 1.6 over nomotion condition) in disturbance rejection.These results support the applications of motion platforms in groundbased flight simulations.For training effectiveness, however, no significant transfer of training due to motion was found in several military studies 25 ' 27 even though motion cues were found to have significant effects to improve pilots performance in some measurements and tasks. 27A comprehensive understanding of man/machine interaction involving visual and motion cues is therefore required to determine the effectiveness of the ground-based flight simulator.
+PsychophysicsIn fixed-based simulators, even without a motion device, visual cues generate self-induced motion.The self-motion is dependent on the peripheral visi&n, spatial frequency, and background of the scene. 30The approximate frequency response of the visually induced motion bears a first order characteristic which falls off at 0.1 Hz. 21This indicates a significant delay in integrating the acceleration to rate and/or position to perform the task if the acceleration information is solely derived from visual cues.To determine the simulation cueing effects one approach is to develop a structured model such that pilot/vehicle interaction can be analyzed.It is desired that a closed-loop mathematical structure can represent pilot's physical interaction with controls, simulation cues, and the task.A representative structure developed from manual flight control concepts is shown in Figure I. 31 If each key element in this closed-loop structure can be characterized and quantified, the complicated man/machine interface with simulation cues in ground-based flight simulations may be explained analytically.The human's motion sensing mechanism primarily comes from vestibular system, and proprioceptive feedback via organ, limbs, and surface pressure.Gum 32 discussed these sensing devices characteristics and developed mathematical models for each sensing mechanism.Peters 33 did a summary review on both angular and translational motion sensing studies in 1969, followed by another extensive review by Zacharias 34 in 1978.Both reviews identified a wide range of studies and results in specific human sensory characteristics and modeling.Most of the results, however, have been found in a single degree-of-freedom only.The established understanding indicates that angular rates are sensed by semicircular canals in the vestibular system, 34 ' 35 low-frequency linear accelerations are sensed by the otoliths, and highfrequency linear accelerations are sensed by other tactile mechanisms, including the neck muscles and receptor in a pilot's seat-of-pants. 34" 36 A clear and brief summary including block diagrams of key motion sensory characteristics models is presented by Schroeder 36 in his 1999 report.Threshold is one of the nonlinear human sensing characteristics of particular interest since it is directly related to the time delay in sensing the onset acceleration and the perception of smoothness of motion cues.Table 2 summarizes findings from several representative investigations. 34" 35 ' 37 ' 38 The range of variations reflects empirical effects due to different test subjects, test apparatus, and methods.In addition, as a common practice in motionbased flight simulators, low frequency longitudinal and lateral accelerations are generated by tilting the platform, e.g., a x = g sin0.The translational acceleration threshold, therefore, has an effect on angular tilt threshold.Similarly, the angular rate threshold also has a direct impact on the tilting motion which may lead to a conflict with visual perception and a sensation of vertigo due to pilot sensing uncommanded rotational cues.A lot of work has been done in this area but knowledge of human sensing characteristics is still incomplete.Understanding of otolith characteristics is limited to the longitudinal DOF only.The tactile model needs more refinement and validation.Angular motion sensing characteristics are mainly developed from pure rotational motion alone.Data have shown significant angular rate threshold increases when translational motion is added which suggests there is a dependency in angular motion sensing characteristics on otolith sensing. 35Most importantly, most of the past works are done in single DOF.The need to develop an integrated cueing model for multiple DOF as recommended by Zacharias 34 still exists.
+Pilot ModelingWith pilot-in-the-loop ground-based flight simulation, a feedback loop is formed with the pilot closing the control loop with a task using the perceived simulated aircraft response via visual and motion cues.The goal is to utilize a structured approach for human characteristics and behavior to determine the effectiveness of given flight simulation cues.If such loop structure and simulation feedback cueing characteristics can be identified, criteria can then be developed to determine and predict the simulation effectiveness based on the missions.McRuer 23 investigated such a logical approach by formulating a pilot model based on plant characteristics in a tracking task with fixed-based flight simulations.One important aspect from his investigation was developing a crossover model, which relates the operator (pilot) and controlled element (simulated aircraft) transfer characteristics in the frequency domain.This model has been widely used among the researchers and investigators with its key parameters, crossover frequency and phase margin, to measure pilot's response due to specific variations in a closed-loop system.One specific application using the pilot crossover characteristics to determine the simulation cueing effectiveness with a closed-loop structural pilot model is by American Institute of Aeronautics and Astronautics Hess. 39In a series of studies, Hess investigated a single loop maneuver, i.e., vertical (bob-up and bob-down), and a multi-loop maneuver, i.e., roll-lateral (a sidestep), by comparing simulator data and flight test data from an Army UH-60 Black Hawk. 40n a closed-loop system representation, a structural pilot model was developed based on psychophysics characteristics that included central nervous system and neuromuscular inner loop modeling, and a procedure using pilot crossover parameters to determine the loop closure performance was developed to determine simulation fidelity.This approach shows promise, but has not been fully validated.Another approach in analyzing and determining simulation cueing effectiveness is through application of optimal control theory. 41Levision and Junker 42 investigated a structured closed-loop model which applied bank angle error and roll acceleration in a cost function for a roll tracking task and a disturbance rejection task.They found that motion cues were much more effective in the disturbance task than in the tracking task, and led to significant increase in gain-crossover frequency of pilots.This is consistent with findings from Stapleford 28 and Jex. 29In addition, to check the general application of the model, a typical set of pilot parameter values were chosen and remained fixed, which included adding control rate to the cost function, to be tested in eight different test conditions.The model results showed good agreement with experimental measures, i.e., RMS tracking error.In the same investigation, efforts were made to include vestibular sensor dynamics to determine the significance of the sensory characteristics in the disturbance rejection task.The results did not find significant differences compared with the simple informational representation.Structured pilot model approaches have shown promise in providing analytical ways of characterizing and estimating man/machine interaction with simulation cues.The findings, however, have been limited to small samples of control tasks and limited degrees-of-freedom.The interaction between the visual cues and motion cues are not fully understood.
+Motion Cueing CriteriaMotion cues have been shown to improve pilot performance.False cues due to limited motion travel, however, could have severe impact on the effectiveness of the motion. 43It should be noted that motion cues are a combination of the motion system dynamics and the motion drive algorithms, i.e., washout filters.Therefore, the characteristics of both must be considered in evaluating motion cueing effectiveness.For motion system dynamics, AGARD-AR-144 44 has identified five key system characteristics.They are: excursion limits for single DOF, describing function, linearity and acceleration noise, hysteresis, and dynamic threshold.However, no objective performance criteria were recommended.FAA AC 120-63" proposes a minimum describing function requirement in the frequency domain for helicopter simulators, Figure 2, and is supported by an investigation using a 20-ft sidestep with motion cues fully matching the visual cues. 5Logically, the linearity and acceleration noise criteria can be developed from the human's motion sensing threshold.To determine the motion cueing fidelity requirement due to washout filter applications, Sinacori 45 first developed criteria using the magnitude and phase of motion cues at 1 rad/sec for angular rate and specific force, Figure 3, to correlate with pilots' subjective perception of motion cues in an "S" maneuver at 60 knots with a high performance helicopter simulation.High, Medium, or Low motion fidelity region was established based on motion sensation relative to visual flight (as perceived through the use of the visual display).Jex 46 developed a lateral washout filter criterion, also shown in Figure 3, based on four pilots comments using an air-toair gunnery type evasive maneuver and a roll washout filter of s/(s+0.4).Schroeder 36 refined Sinacori's criterion based on his work in yaw and vertical motion DOF with helicopter tasks.White 47 takes a different approach in defining motion fidelity criterion that is dependent on the magnitude of false specific force cues, Figure 4.This approach is justified based on human motion sensory threshold characteristics.There are two specific motion drive components that typically are overlooked by simulator operators but have significant effects in cueing conflict with visual cues.One relates to translational motion relative to the angular motion, and the other is the tilting.Translational travel that is required to fully coordinate with roll and pitch angular motion is normally heavily attenuated due to available travel.The resulting specific force false cue has been found to significantly affect pilots' perception of motion and their workload. 48A roll-lateral coordination criterion 49 was developed independently for this specific cueing application from a sidestep task.Tilting is another visual-motion cueing conflict that bears a significant effect.Usually, low frequency longitudinal and lateral specific force cues are generated by tilting the cab as discussed previously.Excessive angular rate can easily lead to severe visual-motion cueing conflict.A rate limit tied to human angular-rate sensing threshold is recommended.The criteria being reviewed provide some guidelines to the flight simulation community that may affect the effectiveness of motion-based flight simulators.However, these criteria are developed from limited empirical data with selected tasks, and from single DOF and two degrees-offreedom investigations.Extending the investigation into multiple degrees-of-freedom, and developing correlation with visual cueing parameters, e.g., FOV and delay, and pilot crossover characteristics, which are simulated aircraft dynamics and task dependent, are recommended.
+American Institute of Aeronautics and Astronautics
+SummaryA brief summary is presented as follows, Transport delay: Modern technology can meet current FAA specifications.Visual resolution: Guideline for required visual resolution relative to task level exists.Developing a universal procedure to measure the visual resolution is recommended.Scene content: Scene content has significant effects in transfer of training and pilot performance.Future studies in texture patterns, terrain shape, and object size and spacing are recommended.Field-of-view:Large FOV has been shown to have significant effects with higher-order simulated aircraft dynamics.Results from various transfer of training studies were mixed.More empirical data with a range of tasks and simulated aircraft characteristics are recommended to establish the FOV effect.Psychophysics:Human angular motion sensing characteristics have been established.Translational motion sensing characteristics from the otoliths are limited to the longitudinal DOF.The tactile model needs refinement and validation.Future studies in interaction between multiple sensing mechanisms and integrated cueing model in multiple degrees-of-freedom are recommended.Pilot modeling: The existing approaches to determine simulation effectiveness in limited DOF studies have shown promises.More empirical data from a variety of tasks, simulated aircraft, and visual and motion cueing conditions are recommended to improve the modeling techniques and to validate the approach.Motion cueing criteria: Developing a more comprehensive motion system dynamic specification is recommended.More empirical data to support the established motion fidelity criteria and expand the criteria to multiple degrees-of -freedom are recommended.
+Concluding RemarksThis review covers only a small but important part of issues related to ground-based flight simulation effectiveness.Extensive work has been done and quite a bit knowledge has been gained in past decades yet few definite answers are offered to determine the effectiveness of the simulation.The statement reflects limited knowledge in man/machine interaction using simulation cues and suggests additional research is required.In addition to preceding recommendations and summarized future work, additional recommendations are presented for future research.(c)2000 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.
+(c)2000 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s) 1 Sponsoring Organization.
+(c)2000 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s) 1 Sponsoring Organization.
+FigureFigure 2 .2Figure I.A representative man-in-the-loop manual flight control structure
+Figure 3. Recommended motion fidelity criterion
+Figure 4 .4Figure 4.The permissible values of nonlinear distortion (Reference 47)
+Table 2 .235mmary of motion sensing threshold Hosman 38 sine wave, 1-14 rad/sec Zaichik et al35, sine wave, 0.5 -8 rad/sec 8 American Institute of Aeronautics and Astronautics1. A more organized effort in following recommendationssuggested by past investigators and researchers to fill inthe blanks.2. A universal test procedure that documents the keysimulation cueing characteristics and effects thatinclude, but not limited to, simulated aircraft, visual
+			American Institute of Aeronautics and Astronautics
+		
+		
+			
+Table 1. Summary of Field-of-View (FOV) effectiveness resultsIrish 24 Nataupsky 25 Kellogg 19 Dixon 26 Westa 27 Lintern 13 Advanced Simulator for Pilot Training (ASPT), T-37 Summary: Used three levels of FOV, 36°V.Found FOV to be significant i controlled approach, and 360°			
+			
+
+				
+
+
+	
+		The Effect of Visual-Motion Time Delays on Pilot Performance in a Simulated Pursuit Tracking Task
+		
+			GKMiller
+		
+		
+			Jr
+		
+		
+			DRRiley
+		
+		
+			March 1977
+			NASA TN D-8364
+		
+	
+	Miller, G. K.,Jr.; and Riley, D. R.: "The Effect of Visual-Motion Time Delays on Pilot Performance in a Simulated Pursuit Tracking Task," NASA TN D-8364, March 1977.
+
+
+
+	
+		Effects of vehicle bandwidth and visual spatial-frequency on simulation cueing synchronization requirements
+		
+			WilliamChung
+		
+		
+			JefferySchroeder
+		
+		
+			WalterJohnson
+		
+		
+			WilliamChung
+		
+		
+			JefferySchroeder
+		
+		
+			WalterJohnson
+		
+		10.2514/6.1997-3655
+		AIAA-97-3655
+	
+	
+		22nd Atmospheric Flight Mechanics Conference
+		New Orleans, LA
+		
+			American Institute of Aeronautics and Astronautics
+			1997
+		
+	
+	Chung, W.; Schroeder, J.A.; and Johnson, W.W.: "Effects of Vehicle Bandwidth and Visual Spatial-Frequency on Simulation Cueing Synchronization Requirements," AIAA Atmospheric Flight Mechanics Conference, New Orleans, LA, AIAA-97-3655, 1997.
+
+
+
+	
+		Large and Medium Hub: Aviation Forecast, 1966-1982. Aviation Forecast Division, Office of Aviation Economics, Federal Aviation Administration, Department of Transportation, Washington, D.C. July 1971. 40p. $3
+		10.1177/004728757201100120
+		AC-120-40B
+	
+	
+		Journal of Travel Research
+		Journal of Travel Research
+		0047-2875
+		1552-6763
+		
+			11
+			1
+			
+			July 1991
+			SAGE Publications
+		
+	
+	AC-120-40B, Airplane Simulator Qualification, U.S. Department of Transportation, Federal Aviation Administration, July 1991.
+
+
+
+	
+		Effects of vehicle bandwidth and visual spatial-frequency on simulation cueing synchronization requirements
+		
+			WilliamChung
+		
+		
+			JefferySchroeder
+		
+		
+			WalterJohnson
+		
+		
+			WilliamChung
+		
+		
+			JefferySchroeder
+		
+		
+			WalterJohnson
+		
+		10.2514/6.1997-3655
+	
+	
+		22nd Atmospheric Flight Mechanics Conference
+		
+			American Institute of Aeronautics and Astronautics
+			1997
+		
+	
+	Chung, W.; and Schroeder, J.A.: " Visual and Roll-Lateral Motion Cueing Synchronization Requirements for Motion- Based Flight Simulations, " AHS 53 rd Forum, 1997.
+
+
+
+	
+		Investigation of roll-lateral coordinated motion requirements with a conventional hexapod motion platform
+		
+			WilliamChung
+		
+		
+			DougRobinson
+		
+		
+			JasonWong
+		
+		
+			DucTran
+		
+		10.2514/6.1998-4172
+		AIAA-98-4172
+	
+	
+		AIAA Modeling and Simulation Technologies Conference and Exhibit
+		
+			American Institute of Aeronautics and Astronautics
+			1998
+		
+	
+	Chung, W.; Robinson, D.; Wong, J.; and Tran, D.: "Investigation of Roll-Lateral Coordinated Motion Requirements with a Conventional Hexapod Motion Platform," AIAA Modeling and Simulation Technologies Conference, AIAA-98-4172, 1998.
+
+
+
+	
+		
+			REMcfarland
+		
+		CGI Delay Compensation," NASA TM-86703
+		
+			1985
+		
+	
+	McFarland, R.E.: "CGI Delay Compensation," NASA TM-86703, 1985.
+
+
+
+	
+		Flight Simulator Visual Display Requirements for Aerial Combat Manuevering
+		
+			LBrown
+		
+		
+			CaptBrunderman
+		
+		
+			JCapt
+		
+		AFHRL-TR-85-39
+		
+			1985
+		
+	
+	Brown, L, Capt.; Brunderman, J., Capt.: "Flight Simulator Visual Display Requirements for Aerial Combat Manuevering," AFHRL-TR-85-39, 1985.
+
+
+
+	
+		A Visual System Display for Full-Mission Flight Simulator Training
+		
+			MLarsen
+		
+		
+			FGruendell
+		
+	
+	
+		Image VII Conference
+		
+			June 1994
+		
+	
+	Larsen, M.; Gruendell, F.:"A Visual System Display for Full-Mission Flight Simulator Training," Image VII Conference, June 1994.
+
+
+
+	
+		A new electron gun for the vacuum evaporation of metals and dielectrics by R. Thun and J. B. Ramsey, U.S. Army Engineer Research and Development Laboratories, Fort Belvoir, Virginia
+		
+			JJohnson
+		
+		10.1016/0042-207x(59)90458-0
+	
+	
+		Vacuum
+		Vacuum
+		0042-207X
+		
+			9
+			5-6
+			300
+			October 1958
+			Elsevier BV
+			Virginia
+		
+	
+	Analysis of Image Forming Systems
+	Johnson, J.: "Analysis of Image Forming Systems," Image Intensifier Symposium, US Army Engineer Research and Development Laboratories, Ft. Belvoir, Virginia, October 1958, pp.249-273.
+
+
+
+	
+		THE POWER SYSTEMS COMMITTEE OF THE AMERICAN ROCKET SOCIETY: (Predecessor of American Institute of Aeronautics and Astronautics)
+		
+			SRBlack
+		
+		
+			PLyon
+		
+		10.1016/b978-0-12-395680-4.50005-8
+	
+	
+		Power Systems for Space Flight
+		
+			Elsevier
+			1995
+			
+		
+	
+	Course Reference Material
+	Black, S.R.; and Lyon, P: "Fundamentals of Visual Simulator Display Systems," Course Reference Material, Image Society Professional Development Course, 1995. American Institute of Aeronautics and Astronautics (c)2000
+
+
+
+	
+		Subject and Author Indexes of Technical Papers Published in the AIAA Journals, Progress in Astronautics and Aeronautics, and Astronautics & Aeronautics in 1974
+		10.2514/3.49613
+	
+	
+		AIAA Journal
+		AIAA Journal
+		0001-1452
+		1533-385X
+		
+			12
+			12
+			
+			
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.
+
+
+
+	
+		Visual System Resolution: Where Have All My Pixels Gone?
+		
+			JDClevenger
+		
+		
+			TMNelson
+		
+	
+	
+		1998 IMAGE Conference
+		Scottsdale, Arizona
+		
+			August, 1998
+		
+	
+	Clevenger, J.D.; and Nelson, T.M.: "Visual System Resolution: Where Have All My Pixels Gone?", 1998 IMAGE Conference, Scottsdale, Arizona, August, 1998.
+
+
+
+	
+		Simulator Design and Instructional Features for Air-to-Ground Attack: A Transfer Study
+		
+			GavanLintern
+		
+		
+			DanielJShepard
+		
+		
+			DonnaLParker
+		
+		
+			KarenEYates
+		
+		
+			MargaretDNolan
+		
+		10.1177/001872088903100107
+	
+	
+		Human Factors: The Journal of the Human Factors and Ergonomics Society
+		Hum Factors
+		0018-7208
+		1547-8181
+		
+			31
+			1
+			
+			1989
+			SAGE Publications
+		
+	
+	Lintern, G; Sheppard, D.J.; Parker, D.L.; Yates, K.E.; and Nolan, M.D.: "Simulator Design and Instructional Features for Air-to-Ground Attack: A Transfer Study," Human Factors, 1989, 31(1), 87-99.
+
+
+
+	
+		Visual Augmentation and Scene Detail Effects in Flight Training
+		
+			GavanLintern
+		
+		
+			JeffersonMKoonce
+		
+		10.1207/s15327108ijap0204_4
+	
+	
+		The International Journal of Aviation Psychology
+		The International Journal of Aviation Psychology
+		1050-8414
+		1532-7108
+		
+			2
+			4
+			
+			1992
+			Informa UK Limited
+		
+	
+	Lintern, G.; and Koonce, J.M.: "Visual Augmentation and Scene Detail Effects in Flight Training," International Journal of Aviation Psychology, 1992, 2, 281-301.
+
+
+
+	
+		Scene Content and Runway Breadth Effects on Simulated Landing Approaches
+		
+			GavanLintern
+		
+		
+			MichaelBWalker
+		
+		10.1207/s15327108ijap0102_3
+	
+	
+		The International Journal of Aviation Psychology
+		The International Journal of Aviation Psychology
+		1050-8414
+		1532-7108
+		
+			1
+			2
+			
+			1991
+			Informa UK Limited
+		
+	
+	Lintern, G; and Walker, M.B.: "Scene Content and Run- way Breadth Effects on Simulated Landing Approaches," InternationalJournal of Aviation Psychology, 1991, 1, 1 IT- 132.
+
+
+
+	
+		Content, Variety, and Augmentation of Simulated Visual Scenes for Teaching Air-to-Ground Attack
+		
+			GavanLintern
+		
+		
+			KarenEThomley-Yates
+		
+		
+			BrianENelson
+		
+		
+			StanleyNRoscoe
+		
+		10.1177/001872088702900105
+	
+	
+		Human Factors: The Journal of the Human Factors and Ergonomics Society
+		Hum Factors
+		0018-7208
+		1547-8181
+		
+			29
+			1
+			
+			1987
+			SAGE Publications
+		
+	
+	Lintern, G.; Thomley-Yates, K.E.; Nelson, B.E.; Roscoe, S.N.: "Content, Variety, and Augmentation of Simulated Visual Scenes for Teaching Air-to-Ground Attack," Human Factors, 1987, 29(1), 45-59.
+
+
+
+	
+		Visual Scene Properties Relevant for Simulating Low-Altitude Flight: A Multidimensional Scaling Approach
+		
+			JamesAKleiss
+		
+		10.1518/001872095778995607
+	
+	
+		Human Factors: The Journal of the Human Factors and Ergonomics Society
+		Hum Factors
+		0018-7208
+		1547-8181
+		
+			37
+			4
+			
+			1995
+			SAGE Publications
+		
+	
+	Kleiss, J.A.: "Visual Scene Properties Relevant for Simulating Low-Altitude Flight: A Multidimensional Scaling Approach," Human Factors, 1995, 37(4), 711-734.
+
+
+
+	
+		Visual cueing effectiveness: Comparison of perception and flying performance
+		
+			JoeDe Maio
+		
+		
+			EdwardJRinalducci
+		
+		
+			RebeccaBrooks
+		
+		
+			JohnBrunderman
+		
+		10.1037/e574072012-011
+	
+	
+		5 th Annual Interservice/Industry Training Equipment Conference
+		Arlington, VA
+		
+			American Psychological Association (APA)
+			1983
+		
+	
+	DeMaio, J.,; Rinalducci, E.J.; Brooks, R.; and Brunderman, J.: "Visual Cueing Effectiveness: Comparison of Perception and Flying Performance," 5 th Annual Interservice/Industry Training Equipment Conference, Arlington, VA, 1983.
+
+
+
+	
+		Field-Of-View Variations and Stripe-Texturing Effects on Assault Landing Performance in the C-130 Weapon System Trainer
+		
+			RobertSKellogg
+		
+		
+			DavidCHubbard
+		
+		
+			MichaelJSieverding
+		
+		10.1037/e520422006-001
+		AFHRL-TR-89-3
+		
+			1989
+			American Psychological Association (APA)
+		
+	
+	Kellogg R.S.; Hubbard, D.C.; Sieverding, M.J.: "Field- Of-View Variations and Stripe-Texturing Effects on Assualt Landing Performance in the C-130 Weapon System Trainer," AFHRL-TR-89-3, 1989.
+
+
+
+	
+		Peripheral Cues and Color in Visual Simulation
+		
+			ConradLKraft
+		
+		
+			CharlesDAnderson
+		
+		
+			CharlesLElworth
+		
+		10.1177/154193128202601022
+	
+	
+		Proceedings of the Human Factors Society Annual Meeting
+		Proceedings of the Human Factors Society Annual Meeting
+		0163-5182
+		
+			26
+			10
+			
+			
+			SAGE Publications
+		
+	
+	Kraft, C.L.; Anderson, C.D.; and Elworth, C.L.: "Peripheral Cues and Color in Visual Simulation," Proceedings of the Human Factors Society 26 th Annual Meeting, p. 906.
+
+
+
+	
+		Visually Induced Motion in Flight Simulation
+		
+			LRYoung
+		
+		
+			April 1978
+			Belgium
+		
+	
+	AGARD-CP-249. Piloted Aircraft Environment Simulation Techniques
+	Young, L.R.: "Visually Induced Motion in Flight Simulation," AGARD-CP-249, Piloted Aircraft Environment Simulation Techniques, Belgium, April 1978.
+
+
+
+	
+		Roll axis tracking improvement resulting from peripheral vision motion cues
+		
+			ThomasEMoriarty
+		
+		
+			AndrewMJunker
+		
+		
+			DonRPrice
+		
+		10.1037/e506152009-059
+	
+	
+		Conference Proceedings of the 12* Annual Conference on Manual Control, NASA TM-X 73
+		
+			American Psychological Association (APA)
+			1976
+			170
+		
+	
+	Moriarty, T.E.; Junker, A.M.; and Price, D.R.: "Roll Axis Tracking Improvement Resulting from Peripheral Vision Motion Cues," Conference Proceedings of the 12* Annual Conference on Manual Control, NASA TM-X 73, 170, 1976.
+
+
+
+	
+		DYNAMIC RESPONSE OF HUMAN OPERATORS
+		
+			DuaneTMcruer
+		
+		
+			EzraSKrendel
+		
+		10.21236/ad0110693
+		AGARD-AG-188
+		
+			1974
+			Defense Technical Information Center
+		
+	
+	McRuer, D.T.; and Krendel, E.S.: "Mathematical Models of Human Pilot Behavior," AGARD-AG-188, 1974.
+
+
+
+	
+		Effects of platform motion, visual and G-seat factors upon experienced pilot performance in the flight simulator.
+		
+			PhilipAIrish
+		
+		
+			GeorgeHBuckland
+		
+		10.1037/e440852004-001
+		AFHRL-TR-78-9
+		
+			
+			American Psychological Association (APA)
+		
+	
+	Irish, P.A.; and Buckland, G.H.: "Effects of Platform Motion, Visual, and G-Seat Factors Upon Experienced Pilot Performance 1978. in the Flight Simulator," AFHRL-TR-78-9,
+
+
+
+	
+		Platform Motion Contributions to Simulator Training Effectiveness: Study III- Interaction of Motion with Field-of-View
+		
+			MarkNataupsky
+		
+		
+			WayneLWaag
+		
+		
+			DouglasCWeyer
+		
+		
+			RobertWMcfadden
+		
+		
+			EdwardMcdowell
+		
+		10.1037/e542792011-001
+		AFHRL-TR-79-25
+		
+			1979
+			American Psychological Association (APA)
+		
+	
+	Nataupsky, M.; Waag, W.L.; Weyer, D.C.; McFadden R.W.; McDowell, E.: "Platform Motion Contributions to Simulator Training Effectiveness: Study III -Interaction of Motion with Field-Of-View," AFHRL-TR-79-25, 1979.
+
+
+
+	
+		Field-of-View Assessment of Low-Level Flight and an Airdrop in the C-130 Weapon System Trainer (WST)
+		
+			KevinWDixon
+		
+		
+			ElizabethLMartin
+		
+		
+			VictoriaARojas
+		
+		
+			DavidCHubbard
+		
+		10.1037/e530252006-001
+		AFHRL-TR-89-9
+	
+	
+		Field-of-View Assessment of Low-Level Flight and an Airdrop in the C-130 Weapon System Trainer (WST)
+		
+			American Psychological Association (APA)
+			1989
+		
+	
+	Dixon, K.W.; Martin, E.L.; Rojas, V.A.; and Hubbard, D.C.: "Field-of-View Assessment of Low-Level Flight and an Airdrop in the C-130 Weapon System Trainer (WST)," AFHRL-TR-89-9, 1989.
+
+
+
+	
+		Simulator design features for carrier landing: II. In-simulator transfer of training.
+		
+			DanielPWestra
+		
+		10.1037/e450922004-001
+		NAVTRAEQUIPCEN 81-C-0105-1
+		
+			1982
+			American Psychological Association (APA)
+		
+	
+	Technical Report
+	Westra, DP.: "Simulator Design Features for Carrief Landing: II. In-Simulator Transfer of Training," Technical Report NAVTRAEQUIPCEN 81-C-0105-1, 1982.
+
+
+
+	
+		Experiments and a Model for Pilot Dynamics with Visual and Motion Inputs
+		
+			RLStapleford
+		
+		
+			RAPeters
+		
+		
+			Alex
+		
+		
+			FR
+		
+		NASA CR-1325
+		
+			1969
+		
+	
+	Stapleford, R.L.; Peters, R.A.; and Alex, F.R.: "Experiments and a Model for Pilot Dynamics with Visual and Motion Inputs, " NASA CR-1325, 1969.
+
+
+
+	
+		Roll Tracking Effects of G-Vector Tilt and Various Types of Motion Washout
+		
+			HRJex
+		
+		
+			REMagdalene
+		
+		
+			AMJunker
+		
+	
+	
+		NASA CP-2060
+		
+			November 1978
+			
+		
+	
+	Jex, H.R.; Magdalene, R.E.; and Junker, A.M.: "Roll Tracking Effects of G-Vector Tilt and Various Types of Motion Washout," NASA CP-2060, pp. 463-502, November 1978.
+
+
+
+	
+		Foreground and background in dynamic spatial orientation
+		
+			ThomasBrandt
+		
+		
+			EugeneRWist
+		
+		
+			JohannesDichgans
+		
+		10.3758/bf03203301
+	
+	
+		Perception & Psychophysics
+		Perception & Psychophysics
+		0031-5117
+		1532-5962
+		
+			17
+			5
+			
+			1975
+			Springer Science and Business Media LLC
+		
+	
+	Brandt, Th.; Wist, E.T.; and Dichgans, J.M.: "Foreground and Background in Dynamic Spatial Orientation", Perception andPsychophysics, 17 (1975) 497- 503.
+
+
+
+	
+		SEMI-ANNUAL REPORT ON ACCELERATORS , JANUARY 1-JUNE 30, 1964
+		10.2172/4010640
+	
+	
+		Second Semi-Annual Status Report on NASA Grant NsG-577, Man-Vehicle Control Laboratory, Center for Space Research, M.I.T
+		Cambridge, Mass
+		
+			Office of Scientific and Technical Information (OSTI)
+			December 1964
+		
+	
+	Second Semi-Annual Status Report on NASA Grant NsG-577, Man-Vehicle Control Laboratory, Center for Space Research, M.I.T, Cambridge, Mass., December 1964.
+
+
+
+	
+		Modeling of the human force and motion-sensing mechanisms.
+		
+			DonRGum
+		
+		10.1037/e438762004-001
+		AFHRL-TR-72-54
+	
+	
+		Modeling of the Human Force and Motion-Sensing Mechanisms
+		
+			American Psychological Association (APA)
+			June 1973
+		
+	
+	Gum, D.R.: "Modeling of the Human Force and Motion-Sensing Mechanisms", AFHRL-TR-72-54, June 1973.
+
+
+
+	
+		Spatial Disorientation—Cues, Illusions and Misperceptions
+		
+			RAPeters
+		
+		10.4324/9781315568584-13
+	
+	
+		Aviation Visual Perception
+		
+			Routledge
+			1969
+			
+		
+	
+	NASA CR-1309
+	Peters, R.A.: "Dynamics of the Vestibular System and Their Relation to Motion Perception, Spatial Disorientation, and Illusions," NASA CR-1309, 1969.
+
+
+
+	
+		Aerospace Medical Research Laboratory, Wright-Patterson Air Force Base, Ohio, AMRL-TR-70-21
+		
+			GLZacharias
+		
+		10.1016/0022-460x(72)90496-8
+		AMRL-TR-78-2
+	
+	
+		Journal of Sound and Vibration
+		Journal of Sound and Vibration
+		0022-460X
+		
+			25
+			4
+			656
+			March 1978
+			Elsevier BV
+			Wright-Patterson Air Force Base, Ohio
+		
+	
+	Zacharias, G.L.: "Motion Cue Models for Pilot-Vehical Analysis," AMRL-TR-78-2, Wright-Patterson Air Force Base, Ohio, March 1978.
+
+
+
+	
+		Acceleration perception
+		
+			LEZaichik
+		
+		
+			VVRodchenko
+		
+		
+			IVRufov
+		
+		
+			YPYashin
+		
+		
+			ADWhite
+		
+		10.2514/6.1999-4334
+	
+	
+		Modeling and Simulation Technologies Conference and Exhibit
+		Portland, Oregon
+		
+			American Institute of Aeronautics and Astronautics
+			1999
+		
+	
+	Zaichik, L.E.; Rodchenko, V.V.; Rufov, I.V.; Yashin, Y.P.; and White, A.D.: "Acceleration Perception," AIAA- 99-4334, Modeling and Simulation Technologies Conference, Portland, Oregon, 1999.
+
+
+
+	
+		Spatial frequency and platform motion effects on helicopter altitude control
+		
+			JefferySchroeder
+		
+		
+			WilliamChung
+		
+		
+			RonaldHess
+		
+		10.2514/6.1999-4113
+	
+	
+		Modeling and Simulation Technologies Conference and Exhibit
+		
+			American Institute of Aeronautics and Astronautics
+			1999-208766, July 1999
+		
+	
+	Schroeder, J. A.: "Helicopter Flight Simulation Motion Platform Requirements, " NASA TP-1999-208766, July 1999.
+
+
+
+	
+		AIAA SPECIALISTS CONFERENCE ON RANKINE SPACE POWER SYSTEMS, NASA LEWIS RESEARCH CENTER, CLEVELAND, OHIO, OCTOBER 26-28, 1965. VOLUME I
+		
+			JLMeiry
+		
+		10.2172/4524921
+		
+			1966
+			Office of Scientific and Technical Information (OSTI)
+			2000
+		
+	
+	NASA CR-628
+	Meiry, J.L.: "The Vestibular System and Human Dynamic Space Orientation, " NASA CR-628, 1966. American Institute of Aeronautics and Astronautics (c)2000
+
+
+
+	
+		Thresholds of motion perception measured in a flight simulator
+		
+			RJ A WHosman
+		
+		
+			JCVan Der Vaart
+		
+		10.1037/e506152009-064
+	
+	
+		NASA TM X-73,170 Twelfth Annual Conference on Manual Control
+		
+			American Psychological Association (APA)
+			May 1976
+			38
+		
+	
+	Thresholds of Motion Perception Measured in a Flight Simulator
+	American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization. 38. Hosman, R.; and van der Vaart, J.C.: "Thresholds of Motion Perception Measured in a Flight Simulator," NASA TM X-73,170 Twelfth Annual Conference on Manual Control, May 1976.
+
+
+
+	
+		A preview control model of driver steering behavior
+		
+			RAHess
+		
+		
+			AModjtahedzadeh
+		
+		10.1109/icsmc.1989.71347
+	
+	
+		Conference Proceedings., IEEE International Conference on Systems, Man and Cybernetics
+		
+			IEEE
+			1985
+			2
+			
+		
+	
+	Hess, R.A.: "A Model-Based Theory for Analyzing Human Control Behavior," Advances in Man-Machine Systems Research, Volume 2, pages 129-175, 1985.
+
+
+
+	
+		Flight simulator fidelity assessment in a rotorcraft lateral translation maneuver
+		
+			RAHess
+		
+		
+			TMalsbury
+		
+		
+			AAtencio, Jr.
+		
+		10.2514/6.1992-4424
+	
+	
+		Guidance, Navigation and Control Conference
+		
+			American Institute of Aeronautics and Astronautics
+			January-February 1993
+			16
+		
+	
+	Hess, R.A.; Malsbury, T.; and Atencio Jr., A.: "Flight Simulator Fidelity Assessment in a Rotorcraft Lateral Translation Maneuver," Journal of Guidance, Control, and Dynamics, Vol. 16, No.l, January-February 1993.
+
+
+
+	
+		A control theoretic approach to manned-vehicle systems analysis
+		
+			DLKleinman
+		
+		
+			SBaron
+		
+		
+			WLevison
+		
+		10.1109/tac.1971.1099842
+	
+	
+		IEEE Transactions on Automatic Control
+		IEEE Trans. Automat. Contr.
+		0018-9286
+		
+			16
+			6
+			
+			June 1971. -42
+			Institute of Electrical and Electronics Engineers (IEEE)
+		
+	
+	NASA CR-1753
+	Kleinman, D.L.; and Baron, S.: "Manned Vehicle Systems Analysis by Means of Modern Control Theory," NASA CR-1753, June 1971. - 42.
+
+
+
+	
+		A Model for the Pilot's Use of Motion Cues in Roll-Axis Tracking Tasks
+		
+			WHLevison
+		
+		
+			AMJunker
+		
+	
+	
+		Thirteenth Annual Conference on Manual Control, MIT
+		Cambridge, Mass
+		
+			June 1977
+		
+	
+	Levison, W.H.; and Junker, A.M.: "A Model for the Pilot's Use of Motion Cues in Roll-Axis Tracking Tasks," Thirteenth Annual Conference on Manual Control, MIT, Cambridge, Mass., June 1977.
+
+
+
+	
+		Motion washout filter tuning - Rules and requirements (expert systems flight simulators)
+		
+			PeterGrant
+		
+		
+			LloydReid
+		
+		10.2514/6.1995-3408
+		AIAA-95-3408
+	
+	
+		Flight Simulation Technologies Conference
+		
+			American Institute of Aeronautics and Astronautics
+			August 1995
+		
+	
+	Grant, P.R.; and Reid, L.D.: "Motion Washout Filter Tuning: Rules and Requirements," ALAA Flight Simulation Technologies Conference, AIAA-95-3408, August 1995.
+
+
+
+	
+		NATO AGARD Night Vision Systems Testing
+		
+			BrucePHunn
+		
+		10.21236/ada426326
+		No. 144
+		
+			September 1979
+			Defense Technical Information Center
+			AGARD, NATO, Neuilly sur Seine, France
+		
+	
+	AGARD Advisory Report
+	AGARD Advisory Report No. 144: "Dynamic Characteristics for Flight Simulator Motion Systems,", AGARD, NATO, Neuilly sur Seine, France, September 1979.
+
+
+
+	
+		The Determination of Some Requirements for a Helicopter Flight Research Simulation Facility
+		
+			JBSinacori
+		
+	
+	
+		NASA CR 152066
+		
+			September 1977
+		
+	
+	Sinacori, J. B.: "The Determination of Some Requirements for a Helicopter Flight Research Simulation Facility," NASA CR 152066, September 1977.
+
+
+
+	
+		Modeling Biodynamic Effects of Vibration. Fifth Year
+		
+			HenryRJex
+		
+		
+			RaymondEMagdaleno
+		
+		10.21236/ada073819
+		AFFDL- TR-79-3134
+		
+			1979
+			Defense Technical Information Center
+			
+		
+	
+	Jex, H.R., Jewell, W.F., Magdaleno, R.E., and Junker, A.M., "Effects of Various Lateral-Beam Washouts on Pilot Tracking and Opinion in the Lamar Simulator," AFFDL- TR-79-3134, pp. 244-266, 1979.
+
+
+
+	
+		Motion fidelity criteria based on human perception and performance
+		
+			AlanWhite
+		
+		
+			VictorRodchenko
+		
+		10.2514/6.1999-4330
+	
+	
+		Modeling and Simulation Technologies Conference and Exhibit
+		Portland, Oregon
+		
+			American Institute of Aeronautics and Astronautics
+			August 1999
+		
+	
+	White, A.D.; and Rodchenko, V.V.: "Motion Fidelity Criteria Based on Human Perception and Performance," AIAA-99-4330, Modeling and Simulation Technologies Conference, Portland, Oregon, August 1999.
+
+
+
+	
+		Simulation Motion Requirements for Coordinated Maneuvers
+		
+			JefferyASchroeder
+		
+		
+			WilliamWChung
+		
+		10.4050/jahs.46.175
+	
+	
+		Journal of the American Helicopter Society
+		J. Am. Helicopter Society
+		0002-8711
+		
+			46
+			3
+			175
+			May 1997
+			American Helicopter Society
+			Virginia Beach, Virginia
+		
+	
+	Schroeder, J.A.; Chung, W.W.; and LaForce, S.: "Effects of Roll and Lateral Flight Simulation Motion Gains on a Sidestep Task," American Helicopter Society 53 rd Annual Forum, Virginia Beach, Virginia, May 1997.
+
+
+
+	
+		Motion fidelity criteria for roll-lateral translational tasks
+		
+			JulieMikula
+		
+		
+			DucTran
+		
+		
+			WilliamChung
+		
+		10.2514/6.1999-4329
+	
+	
+		Modeling and Simulation Technologies Conference and Exhibit
+		Portland, Oregon
+		
+			American Institute of Aeronautics and Astronautics
+			1999
+		
+	
+	Mikula, J.; Chung, W.W.; and Tran, D.: "Motion Fidelity Criteria for Roll-Lateral Translational Tasks," AIAA 99-4329 Modeling and Simulation Technologies Conference, Portland, Oregon, 1999.
+
+
+				
+			
+		
+	

file144.txt

@@ -0,0 +1,211 @@
+
+	
+	
+		
+INTRODUCTIONIn 2015, NASA Ames Research Center conducted the "Big Data, Data Analytics, and Net-enabled ATM and Airspace Operations Project" to identify air traffic management (ATM) functions that can benefit from networked, net-enabled, and/or cloud-based architecture.The objective of this effort was to investigate methodologies that could reduce duplication; reduce cost of operations and upgrades related to air traffic management, airline operations, and flight operations; and provide data analytics.Additionally, potential "fee for service" mechanisms for funding the development and operation of a solution application were to be considered.A subject matter experts (SME) team with experience in ATM, airline operations, airport operations, and big data architectures was assembled to tackle this challenge.Five topics were developed by the team related to the National Airspace System (NAS) performance, they are i) Weather Impacts to Aviation, ii) Big Data and Modeling Infrastructure, iii) Integrated Gate Turnaround Management, iv) Flight Operations Control Management System, and v) Departure and Arrival Management.After a team review, the Integrated Gate Turnaround Management (IGTM) was selected to proceed with a prototype demonstration with the following guidelines.• Use of simulation in the ATM environment • Focus on airline side and ramp operations o Desire to leverage data to manage the uncertainties in nominal and non-nominal, and o Provide performance monitoring and collaborative decision tools to promote integrated gate turnaround operations to meet the on-tie and/or predictable performance upon arrival and departure at the terminal
+CONCEPT AND BENEFITS OF THE INTEGRATED GATE TURNAROUND MANAGEMENTThe concept of Integrated Gate Turnaround Management (IGTM) is to leverage i) the data analytics technologies with multiple historical database and live data to establish bounds of uncertainties of dependent parameters associated with NAS performance, [1,2] ii) a distributed data network shared by stakeholders, [3] and iii) collaborative decision tools [4,5] to optimize the arrival and departure performance at an airport through en route, terminal, and the gate.The IGTM concept is shown in Figure 1.By providing predictable traffic performance information under nominal and off-nominal conditions, e.g., adverse weather, high/low traffic density, runway condition due to rain and/or ice, available flight crew, available ground crew, ground stop orders, and number of transient passengers, collaborative decision tools can be developed for individual stakeholders to optimize the performance under an integrated environment.This coordinated optimization will prevent localized optimization that could lead to system-wide delay.
+THE IGRM PROTOTYPEThe IGTM prototype was defined for the service space starting from the final metering fix upon arrival at an airport to takeoff or wheel-off.This will allow the prototype to be integrated with automation tools for terminal and en route airspace operations for future applications.The IGTM prototype covers the flight and ground operations from approach to runway, landing and wheels-down, to taxi-in from the active movement area to the in-bound transfer spot for handover by ATC to ramp control, lastly, to taxi in the ramp area to arrive at the assigned gate.The aircraft, once parked at the gate, would then undergo gate turnaround operations after which the aircraft reverses the process with gate pushback, taxi-out to the out-bound transfer spot from ramp control to ATC, taxi-out to the departure queue, and ending at takeoff to meet the departure slot time.This complex process involves the controlling authorities and decision stakeholders which includes the ATC ground controllers, that have authority for the airport movement area, and the airport/airline ramp controllers, that have authority for the airport non-movement area.The airline operators, airline dispatchers, pilots and ground handling personnel who determine the readiness of the airplane are also stakeholders in this process.The system architecture of the IGTM prototype is shown in Figure 2. The prototype has eight major components, listed below.Descriptions of these components are discussed in following paragraphs.1. IGTM Model/Controller, which controls the IGTM simulation and collaborative decision tools.  5. SOSS JMS Service Adaptor, to translate data between the SOSS and the IGTM Controller.6. IGTM Graphic User Interface (GUI) including the Dashboard, which provides the interface to the stakeholders.7. ActiveMQ, an open source broker with a Java Messaging Service (JMS), provides the communications network for servicing the data exchanges.8. Mongo server or MongoDB, an open source database server for the IGTM prototype.
+IGTM Host System/Launch ScriptsThe IGTM prototype is entirely written in JAVA and works with products, which are OS agnostic and can run on Linux, Windows or a MAC system.The prototype is a collection of 8 processes (See Figure 2, excluding the scenario configuration) that can be configured through XML to run on any number of systems.Scenario configuration is completed pre-simulation, by loading the desired simulated case files into MongoDB.The prototype was designed to allow any number of GUI/Dashboards users from multiple stakeholders and Model/Controllers exchange messages through the ActiveMQ Bus.By modifying the host, port, database and transport prefix in the various XML files one can reconfigure the process home system.Mongo itself can be configured to run on a clustered system solely by making changes in its XML files and the configuration files of the Model/Controller/BAI and the AOC components.These modifications can be made to increase the overall application performance.Modifications to the ActiveMQ configuration can increase the number of throughput channels and the volume of traffic allowed to pass through each channel.The launch, configuration, and control of the IGTM processes were accomplished with several Linux Bash scripts, which were called from a master script.All major events are captured in log files and are visible in the standard output.Data collection for the events are captured on the MongoDB server.
+IGTM Model/ControllerThe IGTM Model/Controller, or Controller for short, is an event driven process.The Controller used the data from AOC App, which provided the airline scheduled data, BAI App, which provided the estimated event performance (in this demonstration the mean and standard deviation, σ, of a specified event performance from historical data was used), SOSS for actual flight event data, and commands from the GUI, which provides inputs from the stakeholders for collaborative decision making or gaming on a specific event.The events modeled in the prototype are listed below.
+i.Final metering fix ii.Wheels Down (or touchdown) iii.Inbound spot iv.Gate Arrival/Parked v.Gate turnaround (based on completion of all the gate activities) vi.Push back vii.Outbound spot viii.Departure queue ix.Wheels Up (Takeoff)For each flight, three sets of time performance related data, Scheduled Time, Estimated Time, and Actual Time, are required for each event as shown in Figure 3. Scheduled event time, T i_Scheduled , for an Event i as listed above for individual flights was drawing from the AOC App.Actual event time, T i_Actual , came from the Surface Operations Simulator and Scheduler (SOSS).Estimated event completion time, T i_Estimated was calculated based on the mean and standard deviation (σ) from the BAI App on a specific event as shown in Equation 1.T Estimated_i = T Estimated_i-1 + t Estimated_i-1_to_i(1) where T is the simulation time, e.g., in zulu in seconds i denotes an event, e.g., arrival at the gate i-1 denotes a previous event, e.g., inbound spot t is the duration rom Event i-1 to Event iThe time to take from Event i-1 to Event i is defined in Equation 2.t Estimated_i-1_to_i = Size of the Event/Rate of the Event (2)An example for the Inbound Spot Time, "Size of the Event" would be the distance from the touchdown spot to the inbound spot, and the "Rate of the Event" would be the average (or mean) taxi speed.Distance from the touchdown spot to the inbound spot is given by SOSS, and the average taxi speed is provided by the BAI App.At the gate, the "Turnaround Time," where applicable, determines the "Pushback Time" in Figure 3, is dictated or triggered by the completion of following gate activities i.Deplaning of passengers ii.Baggage unloading iii.Fueling iv.Cabin services v.Catering services vi.Baggage loading vii.Flight crew availability viii.Cabin crew availability ix.Maintenance x.
+Boarding of passengersFor the gate turnaround events, Equation 1 applies to all these events with the same three sets of time performance, i.e., Scheduled, Actual, and Estimated.In these instances, the Deplaning Time was determined by the number of passengers (Size of the Event), and the deplane rate (Rate of the Event).Number of passengers is given by the AOC App, and the average deplane rate is provided by the BAI App.For the turnaround time at the gate, assumptions of critical paths were developed as shown in Table 1.Times for critical paths were calculated, and the estimated turnaround time was determined by the critical path that took the longest time.Additional time delay due to door-close to brake-release, late arrival of the flight crew, and/or cabin crew, and time to receive the clearance for a pushback were included in the turnaround time to determine the Estimated pushback time.OR# OR# OR# OR# OR# OR# OR# Σ# Σ# Σ# Σ# Σ# Σ# OR# TOD# Time# Touchdown# Time# Arrival# Time# Pushback# Time# Outbound## Spot#Time# Departure# Slot#Time# Takeoff# Time# Actual# TOD# Time# Touchdown# Time# Arrival# Time# Pushback# Time# Outbound## Spot#Time# Departure# Slot#Time# Takeoff# Time# Es?mated#by# Computa?on# Descent# Time# Es?mated#by# Analysis# Ramp# Time# TOD# Time# Touchdown# Time# Arrival# Time# Pushback# Time# Outbound## Spot#Time# Departure# Slot#Time# Takeoff# Time# Scheduled# Turnaround# Time# Schedule# Op?miza?on# Normal# Clearance# Op?mal# Clearance# Taxi# Time# Taxi# Time# Normal# Delay# Op?mal# Clearance# OR# Σ# Inbound# Spot#Time# Inbound# Spot#Time# Taxi#In# Time# Inbound## Spot#Time# FMF# Time# FMF# Time# FMF# Time# FMF#-#Final#Metering#Fix#
+Collaborative Decision ToolsOne of the key concepts of the IGTM is applying collaborative decision-making among stakeholders to gain the integrated NAS performance improvements than the optimization of a local event.Thus, the effects of a single decision-making must be propagated through the NAS beyond the local decision-making domain, and promote the coordinated or collaborative decision-making on scheduling the events.For the IGTM prototype, simple uses cases were developed to demonstrate the tools to mitigate unexpected early and late arrival situations.
+Airline Operational Control Center (AOC) AppThe AOC App was developed to simulate the data from Airline Operational Control Center (AOC), which consisted of scheduled flight time information at the Final Metering Fix (FMF), wheels-down (touchdown), inbound spot, arrival at the gate, pushback, outbound spot, and wheels-up (takeoff).The AOC data were developed based on the arrivals and departures at Terminal A of Ft.Worth International Airport (DFW).For the prototype, all the AOC data were stored on the Mongo Server to be accessed by the AOC App.
+BAI AppThe Big Data/Analytics Input (BAI) App simulates the applications of the Big Data and Analytics technologies, which develop predictable results based on the past and current NAS performance according to selected decision trees through a User Interface (UI).The prototype was developed to demonstrate the potential benefits of the BAI data through the UI.For the IGTM prototype, arrival and departure time of flights in December 2013 at Terminal A of DFW were analyzed based on data from the NASA data warehouse.[6] The mean and standard deviation of the turnaround time were identified by analyzing Actual gate-in and gate-out time at each gate per type of airplanes.Estimated time events based on predicted results, in this case, the mean and standard deviation from historical data, were generated for each events defined in the IGTM Controller.For example, "Rate of Event" such as the flight speed between the FMF and the touchdown was defined by SOSS of a specific type of airplane, e.g.B737-800, with a specified speed variation.Rate of Events for gate turnaround, e.g., passenger deplaning, cabin services, and boarding, were obtained from References [7,8].BAI data were also stored on the Mongo Server to be accessed by the BAI App.
+Surface Operations Simulator and Scheduler (SOSS)SOSS is a NASA developed Fast-Time simulator, which was used to simulate flights arriving and departing from Terminal A of DFW, with arrivals on runway 17C and departures on 17R.SOSS was connected through a Java Messaging Service (JMS) Interface which provided the Actual events and the simulated clock time.
+SOSS AdapterThe SOSS Adapter is the fast-simulation translator, while the SOSS component is the Fast-simulation message emitter/consumer that reads/writes in scenario data from files.The SOSS Adapter translates messages from a proprietary socket connection structured data stream and converts each packet into JSON or Serialize Data, which is then transported over the ActiveMQ message bus.
+IGTM GUIThe Control Panel of the IGTM GUI is shown in the top of Figure 4, which allows the user to configure the environment to observe flight data associated with the Airport, Terminal, Spot, and Departure Runway.The Control Panel also provides the user with Collaborative Decision Making (CDM) use case options.The Dashboard, which displays flight status and predicted time events, is shown in the bottom of Figure 4.The color code on the right describes the estimated bounds based on the standard deviation derived by the BAI.A light Green of an event represents a likelihood of an event would be completed within one standard deviation of the scheduled event time or about a 68 percent of successful probability.A light Yellow indicates the event would be completed between 1 and 2 standard deviations or about a 95 percent of successful probability.A Red indicates the event would be completed greater than 2 standard deviations or there is only an about 5 percent of chance to meet the scheduled time.The purpose of the color code is to give users a direct implication of the success rate of completing a given event under the uncertainties the BAI data are generated.Therefore, performance bounds can be established based on the dependencies among performance parameters and specified uncertainties.For gate turnaround performance at the gate, a Gate Status display can be selected by the users from the Dashboard for a given flight as shown in Figure 5.The display shows the same color code topology and the standard deviation (σ or Sigma) in minutes.Users can then associated the color code to the time objectively.
+ActiveMQA messsage 'event' can be defined as "a change in state."In the IGTM application events are as simple as 'an aircraft touches down', 'the aircraft reaches the gate', 'another aircraft has all baggage removed," or various other changes in an aircraft's state of location or activity as shown in Figure 3 in the Gate-Turn Model.IGTM software components handle numerous events.An ATM's system architecture may treat this state change as an event whose   Gate""""A16"PushBack DeNicing Outbound2Taxi2SpotDeparture2Slot TakeNoff occurrence can be made known to other processes within the application architecture.From a formal perspective, what is produced, published, propagated, detected, or consumed is a (typically asynchronous) message called the event notification, and not the event itself, which is the state change that triggered the message emission.Events do not 'travel', they just occur.However, the term event is often used metonymically to denote the notification message itself, which may lead to some confusion.This architectural pattern was used in the design and implementation for transmitting events among loosely, coupled software components and services.The IGTM event-driven system consists of event emitters/agents, event consumers/sinks, and event channels.These channels, or collection of channels, are also referred to as the message bus.The emitters have the responsibility to detect, gather, and transfer events.Event emitters are unaware of a consumer of events, when a consumer does exist the event emitters do not know how the event is used or further processed.Sinks have the responsibility of applying a reaction as soon as an event is presented.These emitter/consumers components are the Dashboard, Controller, Big Data Analytics, AOC, and the SOSS Adapter.The Apache ActiveMQ broker fulfills the event channel role.Event channels are conduits in which events are transmitted from event emitters to event consumers.The knowledge of the correct distribution of events is exclusively present within the event channel.The physical implementation of event channels can be based on traditional components such as message-oriented middleware or point-to-point communication.The selection of ActiveMQ as the messaging broker conduit was based on its ease of configuration, its support in other third party API and its ability to handle various types of payload data.One such data payload type is JSON.In the purposes of IGTM the ActiveMQ message bus use was to simulate the NEMS/SWIM message bus (also ActiveMQ).
+Mongo Server or MongoDBNoSQL databases have emerged in recent years to provide the performance, scalability, and flexibility required of modern applications.This new wave of databases is much better suited for Big Data applications and agile software development practices than its relational counterparts.Mongo was selected because it is one of the leaders in the NoSQL arena and that it couples with other application frameworks.The framework selected for rapid application development was SpringFramework.NoSQL databases offer many benefits, including:• Flexible Data Model.Unlike relational databases, NoSQL databases easily store and combine any type of data, both structured and unstructured as JSON.• Elastic Scalability.MongoNoSQL databases scale out on low cost, commodity hardware, allowing for almost unlimited growth.• High Performance.NoSQL databases are built for great performance, measured in terms of both throughput and latency.These advantages account for the growing popularity of NoSQL databases, and specifically MongoDB.MongoDB stands apart from its peers with its Nexus Architecture that incorporates the strengths of relational databases along with the innovations of NoSQL.MongoDB is the only NoSQL options, which offer an expressive query language, strong consistency, and secondary indexes.IGTM chose MongoDB for this reason and that it mates easily to changes the data model and to not have to tinker with the data layer code.Mongo does all the work for you.Mongo also makes it easily possible to work with many of the Big Data Analytic tools such as Tableau, JasperSoft, OpenRefine, Knime, NodeXL, Import.io and others because of the simplicity of the Mongo Query language.
+USE CASESTwo use cases, i.e., early arrival with gate conflict and gate recovery of a late arrival, of the IGTM prototype were demonstrated.Figure 6 shows a CDM display of gate conflict due to an early arrival.The IGTM Controller would identify available gates, which could be available for the estimated arrival time and required turnaround time if applicable, and display on the Gate Availability display.This will allow the user to identify and select an available option, which may require coordination among stakeholders, with minimum time lost at the tarmac, cost, and ground crew resources.The second use case was to demonstrate a speed-up gate turnaround in order to meet the Scheduled pushback time due to a late arrival.The IGTM Controller would identify the most critical path among all critical paths and allow the user to adjust the time performance within the available resources or methods based on AOC data or BAI data.The user used the Gate-Turnaround Management display as shown in Figure 7 to speed up the Passenger Deplane rate, Passenger Boarding rate, and increase the number of cabin service (or cleaning) crew, typical methods to recovering lost time at the gate.In this case, the Estimated turnaround time was reduced from original 52 minutes to 43 minutes.
+CONCLUDING REMARKSThe IGTM prototype demonstrated the concept and benefit of technologies that provide a stream of real-time analytics combined with historic archived data that bound the uncertainties in a gate turnaround NAS operational space.NAS stakeholders can share the flight information, resources, and time management tools through a common messaging network service to coordinately improve the NAS performance under the nominal and off-nominal conditions.The prototype also offers a modular design to incorporate additional Big Data and Analytics products to support future ATM research.2. AOC App, to simulate the Airline Operational Control Center 3. BAI App, to simulate the Big Data/Analytics interface and provide the analytical data.4. Surface Operations Simulator and Scheduler (SOSS), to simulate live traffic data.
+FigureFigure 1.The IGTM concept
+Figure 3 .3Figure 3. Data flow of the IGTM events
+Figure4Figure4.The IGTM prototype's Control Panel
+FigureFigure 6.A Gate Availability display for gate conflict mitigation
+)be)modified)by)users On)the)critical)path If)the)entry)is)greater)than)the)maximum)assigned)valued)as)described)in)Appendix)D,)set)the)value)to)the)maximum)and) set)the)background)to)RED
+Table 1 . Critical paths during the gate turnaround events1Critical PathsSequence of Events (from left to right)1Deplaning of passengersCabin servicesBoarding of passengers2Baggage unloadingBaggage loading3Catering services4Deplaning of passengersFuelingBoarding of passengers5Maintenance
+		
+		
+
+			
+ACKNOWLEDGEMENTSThis work was funded by a NASA Big Data, Data Analytics, and Net-enabled ATM and Airspace Operations Project under Contract NNA15AB05C.Authors wishes to thank Deepak Kulkarni and Yao Wang of NASA Ames Research Center for their technical guidance and advice.Authors wish to thank industrial team, which include Joe Burns of XCELAR Inc., Steve Koczo and Arlen Breiholz of Rockwell-Collins, Henry Smith and Warren Qualley of Harris Corp., Randall Ho of the IBM Software Group Federal, Bruce Sawhill of the NextGen Aero Sciences, and Ben DeCosta of DeCosta Consulting LLC, for the concept development.Authors also wish to thank John Walker and Darrell Wooten of the SAIC software development group for their support in developing the IGTM prototype.
+			
+
+
+			
+He has been working on a wide range of flight simulations, including fixed-wing and rotorcraft, and NextGen projects as well as unmanned aircraft systems.He has a Master Degree in Aeronautics and Astronautics from Stanford University, a Master Degree in Mechanical Engineering from Oregon State University, and a Bachelor Degree of Science in Industrial Engineering from Chung Yuan University. is the technical lead for NASA's FutureFlight Central facility and the software manager of the Airspace Traffic Generator system.Carla currently works in the Aerospace Simulation Research and Development Branch at Ames Research Center and has been supporting air traffic management simulation research for over 18 years, covering the en route, TRACON, and airport domains.Her prior research experiences include high-fidelity rotorcraft research simulations.She has a B.S. degree in Mechanical Engineering from California State University at Chico.Chachad has 36 years of engineering experience with real-time flight simulations, and in structural analysis and design.Mr. Chachad has an MBA from San Jose State University, an MS in Civil Engineering from the University of Rhode Island, and a BS in Technology for Civil Engineering from the Indian Institute of Technology.
+			
+
+			
+
+				
+
+
+	
+		Cloud Computing for Air Traffic Management - Cost/Benefit Analysis
+		
+			LilingRen
+		
+		
+			BenjaminBeckmann
+		
+		
+			ThomasCitriniti
+		
+		
+			MauricioCastillo-Effen
+		
+		10.2514/6.2014-2582
+	
+	
+		14th AIAA Aviation Technology, Integration, and Operations Conference
+		
+			American Institute of Aeronautics and Astronautics
+			June 2014
+		
+	
+	Cloud Computing for Air Traffic Management -Cost/Benefit Analysis
+	Ren, L; Beckmann, B; Citriniti., T., and Castillo-Effen, M: "Cloud Computing for Air Traffic Management - Cost/Benefit Analysis" (16-20 June 2014, 14th AIAA Aviation Technology, Integration, and Operations Conference)
+
+
+
+	
+		
+			M;Ebbers
+		
+		
+			AAbdel-Gayed
+		
+		Addressing Data Volume, Velocity, and Variety with IBM InfoSphere Streams V3.0
+		
+			March 2013
+		
+	
+	Ebbers, M; Abdel-Gayed, A; and et al.: "Addressing Data Volume, Velocity, and Variety with IBM InfoSphere Streams V3.0," (March 2013)
+
+
+
+	
+		System Wide Information Management (SWIM): Program overview and status update
+		
+			JimRobb
+		
+		10.1109/icnsurv.2014.6820078
+	
+	
+		2014 Integrated Communications, Navigation and Surveillance Conference (ICNS) Conference Proceedings
+		
+			IEEE
+			August 2015
+		
+	
+	Air Transportation Information Exchange Conference
+	Robb, j.: "System Wide Information Management (SWIM) Program Overview and Status Update," Air Transportation Information Exchange Conference (August 2015)
+
+
+
+	
+		Airport Gate Scheduling for Passengers, Aircraft, and Operations
+		
+			SangHyunKim
+		
+		
+			EricFeron
+		
+		
+			John-PaulClarke
+		
+		
+			AudeMarzuoli
+		
+		
+			DanielDelahaye
+		
+		10.2514/1.d0079
+	
+	
+		Journal of Air Transportation
+		Journal of Air Transportation
+		2380-9450
+		
+			25
+			4
+			
+			2013. April 17, 2013
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	Kim, S.H., et al.: "Airport Gate Scheduling for Passengers, Aircraft, and Operations, "Tenth USA/Europe Air Traf- fic Management Research and Development Seminar, ATM2013 (April 17, 2013).
+
+
+
+	
+		Airport Characterization for the Adaptation of Surface Congestion Management Approaches
+		
+			MSandberg
+		
+	
+	
+		Air Traffic Management Research and Development Seminar
+		
+			2013. April 17, 2013
+		
+	
+	ATM
+	Sandberg, M., et al.: "Airport Characterization for the Adaptation of Surface Congestion Management Approaches," Air Traffic Management Research and Development Seminar, ATM2013 (April 17, 2013)
+
+
+
+	
+		Architecture and capabilities of a data warehouse for ATM research
+		
+			MichelleEshow
+		
+		
+			MaxLui
+		
+		
+			ShubhaRanjan
+		
+		10.1109/dasc.2014.6979560
+	
+	
+		2014 IEEE/AIAA 33rd Digital Avionics Systems Conference (DASC)
+		
+			IEEE
+			October 2014
+		
+	
+	Eshow, M. and Lui, M.: "Architecture and Capability of Data Warehouse for ATM Research," 33 rd Digital Avionics Systems Conference (DASC) (October 2014)
+
+
+
+	
+		Hugh Waddington and Howard White: Farmer field schools—from agricultural extension to adult education
+		
+			ElskeVan De Fliert
+		
+		10.1007/s12571-014-0378-9
+		
+	
+	
+		Food Security
+		Food Sec.
+		1876-4517
+		1876-4525
+		
+			6
+			5
+			
+			May 2014. February 2015
+			Springer Science and Business Media LLC
+		
+	
+	Airbus A319 Aircraft Characteristics for Airport and Maintenance Planning. Retrieved May 2014 from http://www.airbus.com/fileadmin/media_gallery/files/tech_data/AC/Airbus-AC_A319_May2014.pdf Boeing B752-200/300 Airplane Characteristics for Airport Planning. Retrieved February 2015 from http://www.boeing.com/assets/pdf/commercial/airports/acaps/757_23.pdf
+
+
+				
+			
+		
+	

file145.txt

@@ -0,0 +1,385 @@
+
+	
+	
+		
+I. Introductionate Turnaround Performance Management, i.e., aircraft arrival at the gate, off-loading, servicing, reloading of passengers, baggage and cargo, door-closing, and pushback (referred to as Gate Turnaround), plays a key role in the National Airspace System (NAS) gate-to-gate performance by receiving aircraft when they reach their destination airport, and delivering aircraft into the NAS upon pushback from the gate and subsequent takeoff.The time the aircraft spends at the gate in preparation to meet the planned departure time is influenced by many factors; and some having considerable uncertainties.Principal factors affecting gate turnaround include: weather, early or late aircraft arrivals, time spent disembarking/boarding passengers, unloading/reloading cargo, aircraft logistics/maintenance services, ground handling, traffic density on the ramp, availability of movement areas for taxiin and taxi-out, deicing, and departure queue management for takeoff.Missing the scheduled pushback time can produce a delayed departure that can cause sufficient schedule deviation to potentially cause a schedule disturbance throughout the flight.Large delays can ripple into an airline operator's schedule for other aircraft in their fleet.In contrast to its importance, the gate turnaround process is not managed well in today's operations and does not effectively make use of technologies that provide enhanced data and modeling of gate operations.Gate Turnaround requires multiple participants, who may not be under a single jurisdiction.Gate turnaround, i.e., door closing time, falls under individual airlines' responsibility, and is managed by multiple organizations within the airline.Gate pushback, on the other hand, is under the responsibility of either a Ramp or Ground controller.There is no coordination on pushback times between the airlines and ground control.Gate arrival and departure times are not fully integrated with surface and terminal area NAS automation and frequently take place in an ad-hoc, first-comefirst-served manner.To address these issues, NASA's Spot and Runway Departure Advisory (SARDA) 1 has been developing tower controller advisory tools to improve the flow of surface traffic.Additional work has been done to applying machine learning techniques in taxi-out time prediction 2 to improve the takeoff time performance.FAA has also identified Traffic Flow Data Manager (TFDM) 3 to provide better predictive and collaborative decision-support tools to the stakeholders such that more informed tactical decisions can be made to improve the surface traffic under uncertainties.The Big Data/Analytics technologies 4 could offer additional benefits by providing predictive models extracted from historical data according to specific set of uncertainty parameters, which are potentially well suited for the complicated gate turnaround environment.The Integrated Gate Turnaround Management (IGTM) concept was developed under the NASA "Big Data Analytics, and Net-enabled ATM and Airspace Operations Project" to identify air traffic management (ATM) functions that can benefit from networked, net-enabled, and/or cloud-based architecture.The project team assembled to tackle this challenge included subject matter experts in ATM, airline operations, airport operations, and in big data management architectures.The IGTM concept is therefore focused on the NAS performance in an integrated service space at an airport terminal or terminals by leveraging following technologies to improve the traffic throughput performance at the gate in meeting the on time performance:1) Data analytics technologies with multiple historical databases and live data to establish bounds of uncertainties of dependent parameters associated with NAS performance, 4,5,6 2) A distributed data network shared by stakeholders, 7 3) Collaborative decision tools, 8,9 for stakeholders to optimize the arrival and departure performance at an airport through en route, terminal, and the gate.
+II. Concept of Operations (CONOPS) for the IGTMThe IGTM concept is developed based on the methodology shown in Figure 1.By analyzing historical data and identifying the dependency of uncertainties in NAS performance parameters or patterns, the adaptive analytics will develop performance models as functions of independent parameters, such as: weather, traffic density, time of the day, day of the week or year, type of airplane, and origin and/or destination of the flight.With these models, IGTM can develop descriptive, predictive, and prescriptive information for a given flight's service performance based on the current states of the NAS, either nominal or off-nominal, and deliver scheduled, estimated (or predicted), and current (or actual) status of all flights within the IGTM's operational space at a given airport to all stakeholders via a user interface.These stakeholders include Air Traffic Control (ATC) ground controllers, ramp controllers, airport operators, and airline personnel, which includes dispatchers, flight and cabin crews, gate agents, and those responsible for fueling, catering, baggage/cargo handling, maintenance, and aircraft parking and pushback.The adaptive analytics analysis will continuously check the patterns developed from the historical data and compare with the live data, then make adjustments if significant deviation in the trend is developed from the process.IGTM will also provide collaborative decision tools by applying the performance models to allow stakeholders to enhance NAS performance collaboratively through efficient surface movement, gate turnaround, and pushback under the nominal and off-nominal operational conditions.To maximize the IGTM's benefits, integration with the en route and terminal automation tools, such as Traffic Flow Management System (TFMS), 10 Terminal Flight Data Manager (TFDM), 3 and Center TRACON Automation System (CTAS) 11 as well as surface sequencing optimization tool such as the NASA developed Spot and Runway Departure Advisor (SARDA) 1 is expected.
+A. Predict System Status with UncertaintiesThe IGTM provides a unified picture of the schedule status of each aircraft and all of the processes and support functions that are required for the flights to depart on schedule.This tracking begins while the aircraft is en route with an estimation of its ability to meet its scheduled time from the top of descent.At the same time, the system tracks the schedule performance of each element of the airport facilities that are required for on-time gate turn around and takeoff of the aircraft.These factors include:• Geography of the airport surface The IGTM will use live data and historical performance models from the Big Data/Analytics, Figure 1, under similar conditions to forecast the times of aircraft touchdown, arrival at the in-bound spot, arrival at the gate, gate pushback, arrival at the outbound spot, and takeoff.It also provides a detailed snapshot of expected schedule performance of each operation involved in gate turn around.In order to accomplish its mission, the IGTM uses a combination of information including detailed geography of the airport surface, characteristics of various aircraft models, actual and planned passenger, baggage, and fuel loads, assignments of gates, aircraft crews, and ground facilities along with schedule and current performance of with respect to the schedule.In addition, the IGTM collects performance data over time building a model of the average performance of various entities.This historical information allows the IGTM to more accurately estimate schedule performance of each flight as a function of the aircraft, ambient conditions, and required services.Each airline user has a detailed operational view of the schedules affecting each of its aircraft.Ramp controllers, service providers, and other aircraft operators see a summary view that protects proprietary information but provides summary schedule information that enables each stakeholder to plan for smooth operations.
+B. Network CommunicationsIGTM requires data content and/or message exchanges among NAS and non-NAS operational services as well as collaborative decision communications among the stakeholders.Figure 2 presents notional system architecture for sharing IGTM information among NAS and non-NAS stakeholders.Communications under the System-Wide Information Management (SWIM) 7 and data messaging exchange will be an ideal application to support the IGTM communication functions.This diagram depicts the programs-of-record that produce/publish information for stakeholder consumption, historical data sources, stakeholders, and interface for IGTM value-added services (i.e.user-interface).It should be noted that it is assumed that Enterprise Service Buses (ESBs)/Interfaces are in place to consume and/or produce information for a particular stakeholder domain (Airline, Airport, etc.).Consumed content is securely accessed, filtered per stakeholder requirements, and distributed within that stakeholder domain as determined by stakeholder operations.Again, the method of access will be determined by whether the information source or stakeholder is considered internal to NAS operations or external.
+C. Collaborative Decision ToolsThe IGTM will provide collaborative decision tools to allow stakeholders to manage nominal and off-nominal events with inherent uncertainties, such as late or early arrival, adverse weather conditions (e.g., limited visibility and thunderstorm, runway condition due to rain, snow, and/or ice), congestion at the surface, and flight/cabin/ground crew availability.A notional decision tree is shown in Figure 3.The collaborative decision tools will allow stakeholders to collectively examine available options and resources in order to deal with operational issues, such as mitigating gate conflicts upon arrival, achieving an optimal pushback time upon departure, and optimizing the departure queuing sequence by utilizing descriptive, predictive, and prescriptive information from the Big Data/Analytics.In short, the tool will enable stakeholders to better meet the scheduled or planned time of arrival and departure, or, if need be, to delay or cancel a flight.
+III. The IGTM System's RequirementsThe IGTM must first evaluate current gate-turnaround operations and identify the primary independent parameters and their uncertainties that have impact on the NAS performance according to a given ATC ConOps on the surface and airlines' operational procedures at the airport terminal.The IGTM then identifies information resources that can be leveraged from Big Data and/or Net-enabled ATM data, and utilize analytics to develop performance models based on statistically significant patterns.With these performance models, descriptive, predictive, and prescriptive information about the current state of the NAS can be delivered to stakeholders.Thereby improving predictability to addressing uncertainties in a coordinated decision making process leading to improved NAS performance.As shown in Figure 1, the Big Data/Analytics hosts the big data warehouse, reads in live ATM data, executes analytics analysis (adaptively) according to scenarios, and generate and deliver statistical performance data to the IGTM module.The Big Data/Analytics software needs to process large volumes of live data, which include flight tracks, ground tracks, schedule changes, weather, and runway configuration changes, and deliver the results with minimum latency.Industry has shown a data processing performance of 12 million messages per second with results returned in 120 msec. 6It is yet to be determined if this level of data processing performance is sufficient to support the mass live and historical data analysis in the IGTM application.Inputs to the Big Data/Analytics: 1) Historical data, which includes flight plans, performance data (delays at arrival), and weather data 2) Performance parameter query from the IGTM Outputs of the Big Data/Analytics 3) Statistical performance of queried performance parameters 4) Live data sources, which include traffic, flight plans, and weather
+A. Systems Engineering ApproachA systems engineering approach was developed to identify independent parameters and functional requirements according to the IGTM system architecture.Figures 456show the operational research approach in identifying the events and activities required during the approach, at the gate, and upon the departure at an airport respectively.Table 1 shows a further breakdown to independent parameters associated with representative events to be managed within the IGTM operational space.These events and independent parameters will define requirements for: Big Data, Analytics, collaborative decision tools, and network communications to support the IGTM concept for the gate operations.
+B. Data SourcesThe IGTM products and decision tools rely on data fusion from historical data and live data as shown in Figure 1.Representative historical data sources available from NASA's ATM data warehouse 12 for ATC-NAS services are listed in Table 2. Data for non-NAS services specifically related to airplane turnaround performance at the gate, such as: passenger deplaning rate, passenger boarding rate, cabin service time, catering service time, fueling time, baggage/cargo unloading rate, baggage/cargo loading rate, maintenance time per type of services, flight/cabin crew availability rate, and ground crew availability rate, will be needed from airlines.Additional non-NAS services data will be needed from airport operators.These include data for runway configurations, runway construction, snow plowing rate, and anti-icing facilities and processing rate.
+IV. The IGTM PrototypeAn IGTM prototype 13 was developed to evaluate the modeling and simulation system architecture of the IGTM concept and demonstrate use cases for mitigating off-nominal conditions.Figure 7 shows a fast-time IGTM simulation system architecture using NASA developed Surface Operations Simulator and Scheduler (SOSS) 14 to simulate live arrival and departure traffic as well as surface traffic to and from the gates at the Dallas Fort Worth International Airport.Use cases were developed using the collaborative decision making tools to mitigate the uncertainties associated with early and late arrival of gate operations in meeting the on time pushback performance.Under this system architecture, the prototype is suitable to support future research and development topics.1) Evaluate data processing and throughput performance requirements in applying Big Data/Analytics according to volume and type of the data in the IGTM application space.2) Identify issues in ATM applications associated with developing modeling patterns from historical data.3) Investigate and evaluate different adaptive analytics methods.4) Integrate with surface automation tools, such as SARDA, to streamline the on time performance optimization including the gate operations.The'IGTM' System'5) Develop data sharing protocol among stakeholders to ensure integrity, security, and safe use of the data.6) Develop collaborative decision making tools for mitigating uncertainties to improve the on time performance.
+V. Concluding RemarksAn IGTM concept was developed to leverage the potential descriptive, predictive, and prescriptive capabilities as well as adaptively adjust the performance models derived from the Big Data/Analytics to improve the surface and gate management at the airport.The concept also leverages using a distributed communication network to exchange ATC NAS and non-ATC NAS service data to promote the NAS performance optimization across jurisdiction of traffic flow control authorities.Collaborative decision tools are also a key component of the concept to promote total system performance optimization via a local solution through coordination among stakeholders.Finally, a prototype was developed to evaluate the system architecture as well as systems' requirements to support such a concept.Figure 1 .1Figure 1.An Integrated Gate Turnaround Management methodology
+Figure 2 .Figure 323Figure 2. A notational IGTM system architecture
+Figure 4 .4Figure 4. Flow of events for arrival flights
+Figure 5 .5Figure 5. Flow of events for turnaround flights at the gate
+Figure 6 .6Figure 6.Flow of events for departing flights
+Figure 7 .7Figure 7. IGTM prototype's system architecture
+• Projected roll-out and taxi time • Traffic congestion at the in-bound spot • Availability of the gate (i.e., is the previous flight leaving on schedule?)• Availability of the baggage/cargo unloading crew • Availability of the baggage/cargo loading crew and the baggage/cargo itself• Catering service availability • Fueling service availability • Availability of technicians and materials for any anticipated maintenance • Deicing service availability • Snow plowing service availability • Projected out-bound taxi time
+Table 1 . Independent parameters for Big Data analytics lwc1Delay at ramp=f(snow accumulation rate or rain rate)Dealy at rwy exit=f(aircraft type, lwc, snow accumulation rate or rain rate)Taxi speed=f(RVR, lwc, snow accumulation rate or rain rate, separation distance)Taxi-in time=f(taxi speed, separation distance, rwy exit, gate assignment)GateTime to unload passenger=f(aircraft type, # passengers)Unload baggage delay=f(thunderstorm, lightning , snow accumulation)Time to unload_baggage=f(aircraft type, # passengers, # of special need passengers, # of crew, equipment)Load baggage delay=f(thunderstorm, lightning, snow accumulation)Time to load baggage=f(aircraft type, # passengers, # of crew, equipment)Feuling delay=f(thunderstorm, lightning, snow accumulation)Time of fueling=f(aircraft type, fuel load)Time of cleaning service=f(aircraft type, # of cleaning crew)Time of catering=f(# passenger)Time of maintenance=f(aircraft type, year, type of issues, facilities)Time to board passenger=f(aircraft type, # passengers, # of special need passenger, # gate agent)Time to get clearance=f(throughput)Time to get to deicing facility=f(taxi_speed, distance to deicing pad)Taxi speed=f(RVR, separation distance)Time of deicing=f(lwc, snow accumulation rate, aircraft type, type of fluid)Departure Taxi out time=f(taxi speed, separation distance, departure rwy, gate assignment)Taxi speed=f(RVR, lwc, snow accumulation rate or rain rate, separation distance)Separation distance=f(RVR, lwc, snow accumulation rate or rain rate)Take Off Rate=f(lwc, snow accumulation rate or rain rate): liquid water content RVR: restricted visual range rwy: runway Phase Analytics Functions Arrival Top of descent to touchdown time=f(aircraft type, equipage, separation distance, wind speed, wind direction) Separation distance=f(RVR, equipage) Roll out time=f(aircraft type, lwc, snow or rain accumulation rate)Landing rate=f(aircraft type, lwc, snow accumulation rate or rain rate)
+Table 2 . Potential historical data sources2
+			Downloaded by NASA AMES RESEARCH CENTER on August 18, 2016 | http://arc.aiaa.org| DOI: 10.2514/6.2016-3909
+			A"NoSQL"" database" Simulated"live"traffic"data" Downloaded by NASA AMES RESEARCH CENTER on August 18, 2016 | http://arc.aiaa.org| DOI: 10.2514/6.2016-3909
+		
+		
+
+			
+AcknowledgmentsThis work was funded by a NASA Big Data, Data Analytics, and Net-enabled ATM and Airspace Operations Project under Contract NNA15AB05C.Authors wish to thank Parimal Kopardekar, Deepak Kulkarni and Yao Wang of NASA Ames Research Center for their technical guidance and advice.Authors wish to thank the industry SME team, which includes Joe Burns of XCELAR Inc., Steve Koczo and Arlen Breiholz of Rockwell-Collins, Henry Smith and Warren Qualley of Harris Corp., Randall Ho of the IBM Software Group Federal, Bruce Sawhill of the NextGen Aero Sciences, and Ben DeCosta of DeCosta Consulting LLC, for the concept development.Authors also wish to thank Carla Ingram, John Walker and Darrell Wooten of the SAIC software development group and Doug Ahlquist of Metis Technology Solutions for their support in developing the IGTM prototype.
+			
+
+			
+
+				
+
+
+	
+		Development and Findings from the Spot and Runway Departure Advisor (SARDA) Human-in-the-Loop (HITL) Simulation Experiment
+		
+			THoang
+		
+		NASA TM-2014-218383
+	
+	
+		NASA
+		
+			November 2014
+		
+	
+	Hoang, T. el al.: "Development and Findings from the Spot and Runway Departure Advisor (SARDA) Human-in-the-Loop (HITL) Simulation Experiment," NASA, NASA TM-2014-218383, November 2014.
+
+
+
+	
+		Taxi-Out Time Prediction for Departures at Charlotte Airport Using Machine Learning Techniques
+		
+			HanbongLee
+		
+		
+			WaqarMalik
+		
+		
+			YoonCJung
+		
+		10.2514/6.2016-3910
+	
+	
+		16th AIAA Aviation Technology, Integration, and Operations Conference
+		Washington D.C.
+		
+			American Institute of Aeronautics and Astronautics
+			June 2016
+		
+	
+	Lee, H., Malik, W., and Jung, Y.: "Taxi-Out Time Predicition for Departures at Charlotte Airport Using Maching Learning Techniques," AIAA, 2016 Aviation Technology, Integration, and Operations Conference, Washington D.C., June 2016.
+
+
+
+	
+		
+			MHuffman
+		
+	
+	
+		Terminal Flight Data Manager (TFDM)
+		
+			April 24, 2014
+		
+	
+	Huffman, M.: "Terminal Flight Data Manager (TFDM)," FAA Terminal Program Industry Forum, April 24, 2014.
+
+
+
+	
+		
+			SAyhan
+		
+		Predictive Analytics with Aviation Big Data" Boeing Research & Technology, IEEE, Navigation and Surveillance Conference
+		
+			April 2013
+		
+	
+	Ayhan, S, el al.: "Predictive Analytics with Aviation Big Data" Boeing Research & Technology, IEEE, Navigation and Surveillance Conference, April 2013.
+
+
+
+	
+		Cloud Computing for Air Traffic Management - Cost/Benefit Analysis
+		
+			LilingRen
+		
+		
+			BenjaminBeckmann
+		
+		
+			ThomasCitriniti
+		
+		
+			MauricioCastillo-Effen
+		
+		10.2514/6.2014-2582
+	
+	
+		14th AIAA Aviation Technology, Integration, and Operations Conference
+		
+			American Institute of Aeronautics and Astronautics
+			20 June 2014
+		
+	
+	Cloud Computing for Air Traffic Management -Cost/Benefit Analysis
+	Ren, L, Beckmann, B, Citriniti, T., and Castillo-Effen, M: "Cloud Computing for Air Traffic Management -Cost/Benefit Analysis," 16-20 June 2014, 14th AIAA Aviation Technology, Integration, and Operations Conference.
+
+
+
+	
+		Addressing Data Volume, Velocity, and Variety with IBM InfoSphere Streams V3.0," IBM, Redbooks
+		
+			MEbbers
+		
+		
+			AAbdel-Gayed
+		
+		
+			March 2013
+		
+	
+	Ebbers, M, Abdel-Gayed, A, and et al.: "Addressing Data Volume, Velocity, and Variety with IBM InfoSphere Streams V3.0," IBM, Redbooks, March 2013.
+
+
+
+	
+		System Wide Information Management (SWIM): Program overview and status update
+		
+			JimRobb
+		
+		10.1109/icnsurv.2014.6820078
+	
+	
+		2014 Integrated Communications, Navigation and Surveillance Conference (ICNS) Conference Proceedings
+		
+			IEEE
+			August 2015
+		
+	
+	Robb, J.: "System Wide Information Management (SWIM) Program Overview and Status Update," Air Transportation Information Exchange Conference, August 2015.
+
+
+
+	
+		Airport Gate Scheduling for Passengers, Aircraft, and Operations
+		
+			SangHyunKim
+		
+		
+			EricFeron
+		
+		
+			John-PaulClarke
+		
+		
+			AudeMarzuoli
+		
+		
+			DanielDelahaye
+		
+		10.2514/1.d0079
+	
+	
+		Journal of Air Transportation
+		Journal of Air Transportation
+		2380-9450
+		
+			25
+			4
+			
+			2013. April 17, 2013
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	Kim, S.H., et al.: "Airport Gate Scheduling for Passengers, Aircraft, and Operations, "Tenth USA/Europe Air Traffic Management Research and Development Seminar, ATM2013, April 17, 2013.
+
+
+
+	
+		Airport Characterization for the Adaptation of Surface Congestion Management Approaches
+		
+			MSandberg
+		
+	
+	
+		Air Traffic Management Research and Development Seminar
+		
+			2013. April 17, 2013
+		
+	
+	ATM
+	Sandberg, M., et al.: "Airport Characterization for the Adaptation of Surface Congestion Management Approaches," Air Traffic Management Research and Development Seminar, ATM2013, April 17, 2013.
+
+
+
+	
+		Weather Forecasting Accuracy for FAA Traffic Flow Management
+		
+			MNovak
+		
+		
+			JShea
+		
+		10.17226/10637
+	
+	
+		Traffic Flow Management System (TFMS)
+		
+			National Academies Press
+			April 23, 2014
+		
+	
+	Novak, M, and Shea, J.: "Traffic Flow Management System (TFMS)," FAA Industry Forum, April 23, 2014.
+
+
+
+	
+		Challenges of air traffic management research - Analysis, simulation, and field test
+		
+			DallasDenery
+		
+		
+			HeinzErzberger
+		
+		
+			ThomasDavis
+		
+		
+			StevenGreen
+		
+		
+			BMcnally
+		
+		
+			DallasDenery
+		
+		
+			HeinzErzberger
+		
+		
+			ThomasDavis
+		
+		
+			StevenGreen
+		
+		
+			BMcnally
+		
+		10.2514/6.1997-3832
+		AIAA-1997-3832
+	
+	
+		Guidance, Navigation, and Control Conference
+		
+			American Institute of Aeronautics and Astronautics
+			1997
+		
+	
+	AIAA Guidance, Navigation, and Control Conference
+	Denery, D., Erzberger, H., Davis, T., Green, S., and McNally, D.: "Challenges of Air Traffic Management Research: Analysis, Simulation, and Field Test," AIAA Guidance, Navigation, and Control Conference, AIAA-1997-3832, 1997.
+
+
+
+	
+		Architecture and capabilities of a data warehouse for ATM research
+		
+			MichelleEshow
+		
+		
+			MaxLui
+		
+		
+			ShubhaRanjan
+		
+		10.1109/dasc.2014.6979560
+	
+	
+		2014 IEEE/AIAA 33rd Digital Avionics Systems Conference (DASC)
+		
+			IEEE
+			October 2014
+		
+	
+	Eshow, M., Lui, M., and Ranjan, S.: "Architecture and Capability of Data Warehouse for ATM Research," 33 rd Digital Avionics Systems Conference (DASC), October 2014.
+
+
+
+	
+		An Integrated Gate Turnaround Management Concept Leveraging Big Data/Analytics for NAS Performance Improvements
+		
+			WilliamWChung
+		
+		
+			CIngram
+		
+		
+			DAhlquist
+		
+		
+			GChachad
+		
+		
+			SMonheim
+		
+		10.2514/6.2016-3909
+	
+	
+		16th AIAA Aviation Technology, Integration, and Operations Conference
+		Virginia Beach, VA
+		
+			American Institute of Aeronautics and Astronautics
+			2016. April 2016
+		
+	
+	Chung, W., Ingram, C., Ahlquist, D., Chachad, G., and Monheim, S: "Modeling and Simulation of an Integrated Gate Turnaround Management Concept," 2016 MODSIM, Virginia Beach, VA, April 2016.
+
+
+
+	
+		Validation of Simulations of Airport Surface Traffic with the Surface Operations Simulator and Scheduler
+		
+			RobertDWindhorst
+		
+		
+			JustinVMontoya
+		
+		
+			ZhifanZhu
+		
+		
+			SergeiGridnev
+		
+		
+			KatyGriffin
+		
+		
+			AdityaSaraf
+		
+		
+			SteveStroiney
+		
+		10.2514/6.2013-4207
+		AIAA-2013-4207
+	
+	
+		2013 Aviation Technology, Integration, and Operations Conference
+		Los Angles
+		
+			American Institute of Aeronautics and Astronautics
+			2013. August 2013
+		
+	
+	AIAA
+	Windhorst, R., et al.: "Validation of Simulations of Airport Surface Traffic with the Surface Operations Simulator and Scheduler," AIAA, 2013 Aviation Technology, Integration, and Operations Conference, AIAA-2013-4207, Los Angles, August 2013.
+
+
+				
+			
+		
+	

file146.txt

@@ -0,0 +1,174 @@
+
+	
+	
+		
+between the visual and motion cues, the results also suggest that visual delay compensation had little or no effect on pilots '  within the physical displacement constraints, i.e. the angular and translational limits.Therefore, the washout filters must be tuned to deliver consistent onset accelerations that complement the cues perceived by the pilot from other simulated devices.To establish a direct correlation between simulation fidelity and handling qualities, Reference 1 suggests a criteria based on washout gains and phase characteristics as a measurement of motion cueing fidelity.Reference 2 follows the same frequency response approach and develops a 30 degree phase distortion criteria to compare perceived simulation cues for handling qualities evaluations.These were also the guidelines applied in developing the motion configurations for this experiment.Reference 3 suggests that many motion cue errors are introduced in flight simulation due to physical constraints of motion platforms.Of all the motion cues perceived by the pilot, there is a fundamental element that is directly dependent on the kinetically cross-coupled motion system dynamic characteristics.This is a result of rigid body induced linear accelerations due to angular motion.Both Reference 4 and 5 indicate that translational accelerations sensed by the pilot are from the vestibular system and tactile mechanisms in the body.Due to the nature of human organ characteristics, lower frequency motion perception is sensed by the vestibular system and higher frequency motion perception is sensed by pressure from the pilot tactile mechanisms.Therefore, when the pilot station is not at the rotational center of the motion platform, an element of translational accelerations, i.e. induced linear accelerations, will be sensed by pilots due to angular motion.Induced linear accelerations are generally compensated in motion commands by assuming that the cross-coupled motion axis responses are the same.However, if the dynamic characteristics of two cross-coupled motion axes are not the same, or caused by different motion washout filter characteristics, discrepancies will be presented to the pilot and have an impact on the simulation fidelity.The objective of this experiment was to study the effect of phase differences between two kinetically cross-coupled motion axes and to determine if a requirement that defines acceptable phase discrepancy between the cross-coupled motion axes is necessary for ground based flight simulators.
+Description of the ExperimentFor a motion simulator where the pilot center of gravity is not at the rotational center of the motion platform, the specific force vector that is sensed by the pilot is governed by equation 1.a ps = a mp -The accelerations produced by the motion system at the pilot station, a m p, is defined by equation 2.»mp = a rc + r + to m X r + co m X (co m X r)+ 2 co m X r (2)The position vector of the pilot, r, is fixed relative to the rotational center, i.e. r and r arc zero.By assuming the rotational rate of the simulator cockpit, oo m , is relatively small, equation 2 can be simplified to a mp *rc X r(3)The second term at the right hand side of equation 3 is the induced linear acceleration due to rotational motion, and the effect of this term is generally compensated in the motion commands such that this motion-platform-dependent term is not presented to the pilot.Similar reason also applies to the second term in equation 1 in compensating for the gravity component as a function of cab attitude.Therefore, the simulator translational and angular motion commands a rc_cmd and to s_cmd we defined as:a rc _cmd = W t (s) -apiiot -W a (s) -cb p i lot X r + W a (s).T mc .g (4) ct>m_cmd = w a( s ) '(5)But the actual simulator responses from translational and angular motion commands are determined by the individual motion axis dynamic characteristics, given by, Tt(s)-a rc _cmd T a (s)-a> m cm( i (6) (7)Therefore, the perceived motion cues in kinetically crosscoupled axes are dependent on the dynamic characteristics of both the washout filters and the motion hardware.If the overall dynamic characteristics, such as the phase characteristics of the lateral and roll motion responses, are not the same, then pilots will be subjected to erroneous linear acceleration cueing.
+Math Model DescriptionA mathematical model in stability derivative form was developed to represent the dynamic characteristics of a rate command helicopter^ that is fully decoupled.The equations of motion are defined by equation 8 and 9. u w q .6..00 .[Slat]Ur J (9)
+Motion SystemThe VMS, as shown in Fig. 2, is a six degree-of-freedom motion platform that permits large excursions in the vertical and lateral axes.The vertical motion axis is driven by eight mechanically coupled 150-horsepower direct-current servomotors as outlined in Ref. 9. The lateral axis is driven by four 40-horsepower direct-current servomotors.Roll, pitch, yaw, and longitudinal are driven by four independent hydraulic systems with 2400 psi hydraulic pressure.The motion system's roll and lateral dynamic characteristics were tuned to three configurations for this experiment to study the effect of the phase difference between the two cross-coupled motion axes.The lateral accelerations due to yaw motion were not present due to the fact that the pilot longitudinal e.g.position was near to the gimbals rotational center for this experiment.Three motion configurations were developed by using the visual system's 60 msec time delay as a reference.The VMS visual and motion system responses were fitted in a form defined by equation 10 which consists a linear transfer function, H(s), and a time delay, T.P(s) = H(s) (10)The characteristics of the visual system and the roll and lateral motion systems, P(s), of the three motion configurations and their equivalent time delays are shown in
+Motion Washout FiltersThe VMS motion drive logic is shown in Fig. 6.Washout filters are applied to translational and rotational pilot station accelerations after being transferred to the inertial frame to keep the simulator within the physical travel limits.Turn coordination and induced acceleration compensation keep the cross-coupled motion commands in accordance to pilot position states relative to the rotational center.A low pass filter is used to tilt the cab in supplementing linear motion cueing at low frequency.For the experiment, two motion washout configurations as shown in Fig. 7, were developed for the hover task to investigate the phase difference effect on pilots' handling qualities.The high fidelity configuration was developed to keep the phase of roll and lateral washout filters the same, i.e. <t>(W a (s)) = <)>(Wt(s)), and to keep both angular and translational motion cueing within the high fidelity region according to Ref. 1.The mixed fidelity configuration represented a case investigated in Ref. 10.For the sidestep task, the motion washout filters were configured as shown in Fig. 8. Again, the washout filter frequencies for roll and lateral axes were chosen to have the same phase characteristics.The dynamic characteristics of the rotorcraft and perceived visual and motion cueing under each motion washout and motion dynamic configuration (with the visual time delay of 60 msec) are shown in Fig. 9 to 11.The acceptable fidelity range for the high fidelity washout configuration, based on the 30 degree phase distortion criteria from Ref. 2, is summarized in Table 3 for all three motion configurations.The acceptable fidelity range is defined as the frequency spectrum where the phase difference between perceived visual cueing and motion cueing is less than 30 degrees.The acceptable fidelity range for the same group of motion configurations but with a visual compensation of 60 msec are shown in the same table to present the effect of the improved pilot perceived model response.As shown with the visual compensation, the pilot perceived an improved roll model response from a bandwidth of 4.5 rad/sec to 10 rad/sec as defined by the math model.However, phase improvement in the visual cueing alone would also increase the discrepancy between perceived visual and motion responses.As a result, a more restricted lateral-directional acceptable fidelity range over the frequency spectrum was developed.
+TasksTwo low speed tasks, Hover and Sidestep, were developed following the guidelines from ADS-33D under the no wind condition.Portions of the task procedures were modified to match the procedures developed in Ref. 2.For the Hover task, the pilot was positioned at an angle with respect to the designated hover point, outlined in Fig.12. The helicopter was initialized at 15 ft altitude.The pilot was asked to translate to a hover position over the desired hover point, with a ground speed of 6 kts, while maintaining the altitude.The desired hover point was defined by a hover target with a sight to indicate lateral position and height cues and a color-coded wall at a 45 degree angle to define longitudinal position cues.The transition to the hover point was to be achieved in one smooth maneuver, i.e. a smooth acceleration command followed by a smooth deceleration command.Creeping up to the final position was not allowed.The time for the pilot to stabilize at the desired hover point, from initiation of deceleration control input, was 15 seconds.Once in a stabilized hover, the pilot was asked to maintain hover position for 30 seconds.Rotorcraft deviations were measured from the desired hover point to determine pilots performance with respect to specified performance criteria, as given in Table 4.For the Sidestep task, starting from a stabilized hover with the longitudinal axis of the rotorcraft oriented 90 degrees to the runway, as shown in Fig. 13, the pilot was asked to initiate a rapid and aggressive lateral translation, with a bank angle of at least 20 degrees, holding altitude constant with power.When the rotorcraft achieved a lateral velocity within 5 knots of the maximum allowable lateral airspeed, 30 knots, the pilot immediately initiated an aggressive deceleration to hover at constant altitude.The peak bank angle during deceleration was kept to at least 20 degrees, and occurred just before the rotorcraft came to a stop.Longitudinal and vertical position deviations were measured against the desired performance criteria, as shown in Table 4.The visual data base was developed to provide visual cues for each task.Pylons and walls were color-coded such that the pilot could easily identify desirable and adequate performance envelopes.At the end of each task, the pilot was asked to give a handling qualities rating (HQR) based on the Cooper-Harper scale of Ref. 11.A modified sidestep task was developed during the experiment to better reveal the significance of phase characteristics of the model-to-motion response.A closed loop task was added at the end of the sidestep maneuver by asking the pilot to hover before a designated pylon, with the same desirable performance criteria defined as before.Due to time limitations, only one pilot examined the modified sidestep task, and no pilot HQR was taken.
+ResultsThe effects of kinetically cross-coupled motion dynamics were analyzed by studying HQRs and comments.Pilot control stick response and task performance data were also evaluated.The summary of the results are as follows:Hover with high fidelity washout configuration Pilot HQRs for three motion configurations are shown in Fig. 14.In comparing the first two motion configurations, i.e. matched cueing response (MCI) versus delayed lateral motion (MC2) two pilots, A and B, noted coordinated rolllateral motion cueing which allowed them make accurate lateral inputs and pay more attention to longitudinal position control under the matched cueing configuration.Pilot B rated the matched cueing configuration better than the delayed lateral motion configuration.Pilot A felt that the matched cueing case (MCI) provided more solid motion cueing relative to the visual response, which reduced his physical and mental workload from that of the lagged lateral case.The increase in physical workload is strongly supported by the representative pilot lateral stick power spectral density (PSD) plot, given in Fig. 15, and the time trace of the pilot stick motion during the position-holding part of the task, Fig. 16.The power spectral density is the normalized energy distribution across the frequency spectrum.These data clearly showed that pilot workload associated with the lateral controller was reduced significantly across the frequency spectrum in the matched visual and motion cueing configuration.However, according to pilot A, the noted improvements in roll-lateral motion cueing response did not outweigh the required workload to hold the longitudinal position.Pilot C felt that both configurations required moderate pilot compensation to meet satisfactory performance criteria.He also felt that the delayed lateral motion had a slight advantage in pilot workload over the well matched case.For the delayed lateral motion configuration, jerkiness was among the common comments shared by all pilots.The third motion configuration, motion lagged visual, was rated by two pilots, A and B. Pilot A rated this configuration worse than the matched cueing case and pilot B rated these two configurations with the same rating.Since the phase characteristics of the roll and lateral motion axes were the same in both configurations, the difference in pilot ratings could only result from the pilots' cueing preference, i.e. between the visual cueing and the motion cueing.Pilot A noted that some motion cues were lagging while pilot B noted that visual and motion cues were in harmony.A summary of pilot lateral control mean-square-value (cp^) and pilot cutoff frequency (ffl c ) from PSDs developed by using CIFER, Ref. 12, is shown in Table 5.The pilot cutoff frequency approach is developed in Ref. 13 to compare pilot response characteristics under both flight and simulation conditions.By assuming a first order pilot response model, pilot cutoff frequency is defined as the frequency at half the power point of the total power spectral density of the given pilot control input, i.e. (p c ^ / q>t 0 tal = 0.5.The mean-square-value of the control with respect to the frequency spectra from 0 to oof, tpcof > is equal to the total area under the PSD plot and is defined by equation 11, where 655 contains the control power content as a function of frequency.Table 5 shows that under the matched cueing case, the total energy of the lateral control stick input consistently stays low among pilots in comparison with the other two mismatched conditions, which show comparable pilot cutoff frequencies.Standard deviations of longitudinal and lateral position holding errors are given in Table 6.This table shows that pilots were able to maintain about the same level of performance regardless the test configurations, i.e. the change of motion parameters appeared to only affect the workload.The longitudinal position cues were provided by the colorcoded wall on the side window when in the stabilized hover position.Nonetheless, it did not provide an adequate range cueing sensitivity.This visual cueing deficiency combined with poorly coordinated pitch and surge dynamic characteristics with respect to visual cueing, Fig. 17, kept pilots' workload high in keeping longitudinal position within the satisfactory performance criteria, and made it more difficult in achieving Level I handling qualities performance.
+Hover with mixed fidelity washout configurationPilot HQRs are shown in Fig. 14.The mixed fidelity motion configuration had a deviation in washout frequency between roll and lateral, 0.1 and 0.6 rad/sec respectively versus 0.3 for both axis in the high fidelity motion washout configuration.The washout gain on the lateral axis was also reduced from 0.9 to 0.4 in the mixed fidelity washout configuration.Roll washout gain was kept the same as the high fidelity washout case.The perceived roll and lateral motion cueing discrepancies as shown in Fig. 18 to 20, are much more significant at the low frequency range than in the high fidelity washout configuration.For pilot B and C, who evaluated these tasks, both felt that the matched case had much better coordinated motion cueing than the other two cases.The pilot comments were very similar to those in the high fidelity motion configuration.The workload for the matched configuration again showed reduced lateral control energy by both pilots, as given in Fig. 21.A summary of pilot cutoff frequency is shown in Table 5.It is noted that from the PSD data, and pilot comments, that there is no significant difference between the high fidelity and mixed fidelity motion configurations.The large phase discrepancy between roll and lateral motion at low frequency did not have a significant effect on pilot workload, or on performance.The phase discrepancy effect in high frequency, however, had a definite effect on pilot workload.Pilot B's HQR was consistent with the result from Ref. 10.Pilot A evaluated all three motion configurations in mixed fidelity configuration.However, his data was contaminated with an incorrect washout filter setup.Therefore, no conclusion can be drawn to confirm the consistency between the experiments.
+SidestepPilot HQRs for the sidestep task are shown in Fig. 22.There is no clear trend to indicate the effect of cross-coupled motion dynamic response.The results for this task were hampered by a lack of range cues when the pilot proceeded to a hover stop.The lack of longitudinal position information, lightly damped pitch motion characteristics, and visualmotion phase discrepancies again led to an appreciable amount of pilot effort in stabilizing the helicopter within desirable performance criteria.For the modified closed-loop sidestep task, only one pilot data point was taken to evaluate two motion configurations, i.e. the matched cueing and delayed lateral motion cases, without taking any HQR.The time traces of the control stick and position error from deceleration to a stabilized hover are shown in Fig. 23.The power spectrum of the lateral stick is shown in Fig. 24.The power spectral density of lateral stick and the pilot cutoff frequency are shown in Table 5.The PSD did not show any significant differences between the two motion configurations.However, pilot A commented that overall control felt solid without any overshoot tendency in the matched cueing configuration.Desirable performance was easily achieved.With lagged lateral motion, however, it was harder to stabilize, and there was a tendency to overshoot.This is shown in the position error time trace, given in Fig. 23.The motion in the latter configuration "felt jerky and artificial".It also required at least moderate pilot compensation to achieve desired performance, which would be a Level 2 handling qualities rating.
+Visual DelayHQRs from pilot A and B with visual delay compensation turned on and off are shown in Fig. 25.From both pilots' HQR on two washout configurations and three motion dynamic configurations, there is no significant difference in their ratings with and without the visual delay.This result suggests that the improved model bandwidth response by removing the visual delay from the system was offset by the phase discrepancy between visual and motion cueing.Cueing discrepancies over the acceptable frequency range (Table 3) requires the pilot to mentally cross check the overall sensed model response, which meant increased pilot workload.The 30 degree phase distortion criteria provides a credible rationale for such a result.
+ConclusionsA piloted motion based handling qualities flight simulation experiment was conducted to evaluate the significance of kinetically cross-coupled motion dynamic characteristics.Roll and lateral motion dynamic characteristics were perturbed for both precision hover and sidestep tasks.Visual delay and visual compensation were also evaluated under the same test conditions.From pilot workload data, the phase characteristics of crosscoupled roll and lateral motion cueing has a significant effect on overall handling qualities of given tasks.Therefore, a requirement on cross-coupled motion axes phase characteristics with respect to visual response is strongly recommended to ensure the fidelity of flight simulation.The data from this experiment suggest that the roll dynamic response from motion cueing should at least match the visual response.The phase lag in lateral motion response with respect to the roll motion response should not be larger than 40 msec.Further investigations are required to define the specific phase criteria associated with the cross-coupled motion dynamic characteristics.Visual delay compensation theoretically improves the simulation visual cueing responses, which should lead to better control bandwidth responses as well.Under the given test conditions, no noticeable pilot HQR or task performance improvement was found.That leads to the conclusion that the model response improvement made by visual cueing alone must be lost in the discrepancy between visual cueing and motion cueing.However, without the visual delay compensation, the vehicle's response characteristics is effectively reduced due to the inherited time delay in the digital flight simulation.
+0.0Table 1.Damping characteristics and control sensitivityX u (I/sec) /-w (I/sec) M q (1/sec) Lp (I/sec) Y v (I/sec) N r (I/sec) -0.7 -4.3-10.5 -0.12 -2.0(ft/sec^/in) (rad/sec 2 /in) (rad/sec 2 /in) (rad/sec 2 /in) -9.873 O45 L8 004< -> £• 4 o o rt •} ^ V) cd «5 2 Pilot a A 0 B AC • • data from ref. 10 ,-a. ..-o & -A 'qn o -^2* -A _ _ _ s^ _ _ ^ _ _ _ o -cr A X • i i i i i i 1 MC1MC2MC3 MC1MC2MC3High fidelity Mixed fidelityTable 2 .2The equivalent time delay is defined as a pure time delay that matches the phase response of P(s) between . 1 to 10 rad/sec.The frequency responses of these three motion configurations, i.e. acceleration output versus acceleration input, are shown in Fig.3, 4, and 5.The first motion configuration, MCI, the matched visual and roll and lateral motion cueing, was developed such that both roll and lateral motion dynamic phase responses matched the visual phase response.The second motion configuration, MC2, delayed lateral motion, was developed to keep the roll axis phase response in phase with the visual system, but to delay the lateral axis phase response by 40 msec.The third motion configuration, MC3, delayed roll and lateral motion, was designed to keep the phase response of both roll and lateral motion axes 40 msec behind the visual response.The first configuration represents the best phase match of both visual response and roll-lateral motion response as perceived by the pilot.Dynamic response for each configuration was tuned to have a satisfactory phase response up to 10 rad/sec.
+Figure 3 .Figure 4 .Figure 6 .Figure 11 .34611Figure 3. Matched visual, and roll and lateral motion configuration, MCI
+FigureFigure 16 .Figure 17 .Figure 19 .Figure 23 .Figure 25 .1617192325Figure 14.Pilot HQRs for hover task
+handling qualities ratings under the given test conditions.command,rad/sec 2 roll motion acceleration response, rad/sec 2 helicopter model pitch rate, body axis, rad/sec helicopter pitch angular acceleration, rad/sec 2 pitch acceleration motion command, rad/sec 2 pitch motion acceleration response, rad/sec 2 helicopter model yaw rate, body axis, rad/sec helicopter yaw angular acceleration, rad/sec2P(s)a linear representation of visual and motionresponse with time delayPcmdp mqqposition vector of the pilot station with respectto simulator RC, ftrelative velocity vector of the pilot station withrespect to simulator RC, ft/secNOMENCLATURErelative pilot station linear acceleration withT a (s) T m L mcrespect to simulator RC, ft/sec 2 transfer function of angular motion axis direction cosine matrix from inertial to body axes of the simulator, n.d. direction cosine matrix from inertial to simulator body axes attitude excluding the component used for low frequency linear specific force, n.d.8 C $lat ^lon 8 r <) > 'Pcof 2pilot collective stick input, in. Pil ot lateral stick input, in. pilot longitudinal stick input, in. rudder pedal input, in. roll attitude, rad mean-square-value over the specified frequency spectrum, n.d.T t (s) utransfer function of translational motion axis helicopter model translational velocity, x-body axis, ft/sec helicopter model translational velocity, y-body axis, ft/sec9 T ro m d) mpitch attitude, rad fitted time delay for visual and motion response, sec simulator angular rate vector, rad/sec, simulator angular acceleration vector, rad/sec 2W a (s) Wj(s) x e X u y e y c g Y vhelicopter model translational velocity, z-body axis, ft/sec transfer function of angular washout filter transfer function of translational washout filter longitudinal position error for hover and sidestep tasks, ft longitudinal damping coefficient, I/sec lateral position error for hover task, ftcbpilot ci) m _ cm( j simulator angular acceleration command vector, helicopter angular acceleration vector, rad/sec 2 rad/sec 2 a mp linear accelerations generated by the motion simulator at the pilot station, ft/sec 2 3p S total linear accelerations sensed by the pilot, ft/sec 2 a pilol helicopter pilot station acceleration vector, ft/sec 2a rcsimulator rotational center (RC) accelerationa rc_cmdvector, ft/sec 2 simulator rotational center (RC) accelerationIntroductioncommand vector, ft/sec 2ggravitational vector, ft/sec 2Motion simulators are widely used in handling qualitiesH(s)fitted linear transfer function of visual andresearch and flight training. These applications depend onmotion response without the time delayonset accelerations produced by the motion platform in combination with cues presented to the pilot from visual displays, control force feel, audio effects, and instrumentation displays. The fidelity of the onset accelerations is subject to the modeled aircraft dynamic characteristics, motion system's dynamic characteristics,LSlat Lp M5lon Mq N5rr °U control power, rad/sec 2 /in. roll damping coefficient, 1/scc pitch control power, rad/sec 2 /in. pitch damping coefficient, I/sec yaw control power, rad/sec 2 /in.motion control algorithms, and displacement constraints.N ryaw damping coefficient, I/secphelicopter model roll rate, body axis, rad/secphelicopter roll angular acceleration, rad/sec 2r °U acceleration motion For ground based motion simulators, this presents quite a challenge, because the displacement constraints dominate the motion fidelity issue.Washout filters are generally used in motion control logic to generate initial onset accelerations
+Table 2 .2Fitted VMS visual and roll-lateral motion response model, and equivalent time delayFitted model response, P(s)Equivalent timedelay, msecMotionVisuale -0.060s60configuration1. Well matched visual andRoll77 -9 s+80P -0.05?,s65motionLateral2.39(152.4)(s 2 +12s+94)_ n ni<;68(s 2 +21s+225)(s 2 +16.2s+164.5) &2. Delayed lateral motionRoll77 -9 s+80" P -0.052s65Lateral152.4.nnis108s 2 + 16.2 s + 164. 5 C3. Delayed roll-lateral motionRoll19 ' 75 s+20P -0.072s107Lateral1 CO A ij^A-001s108s 2 +16.2s+164.5 C
+Table 3 .3Acceptable simulation fidelity range for high fidelity washout filter configuration, rad/secWith visual delayWith visual compensationMotionRoll axisLateral axisRoll axisLateral axisconfigurationMin MaxMin MaxMin MaxMin Max1. matched visual and0.8 >60 0.817 0.750.75roll-lateral motion2. delayed lateral motion0.8>600.87.80.7590.7553. delayed roll and lateral0.88.50.87.80.7550.755motion
+Table 4 .4Task performance criteriaPositionAltitudeHeadingTime toTaskTolerance (ft) D ATolerance (ft) D ATolerance (deg) D AComplete (sec) D AHover±3±8±2 ± 4± 5±10<15<30Sidestep±20±50±10 ±15±10±15Table 5. Pilot lateral stick power spectrum density and pilot cut-off frequencyfor hover taskHigh fidelity washoutMixed fidelity washoutModified sidestepPilot Motion configuration -? ______________________(rad/sec) Q*(Poof 2«c (rad/sec)(Pcof 20)c (rad/sec)1. matched visual0.0042.10.481.7and roll-lateralmotion2. delayed lateral0.0131.70.881.4motion3. delayed roll and0.0152.4lateral motion1 . matched visual0.0172.10.0052.2and roll-lateralmotionB2. delayed lateral0.0552.40.0592.1motion3. delayed roll and0.0651.80.0371.9lateral motion1 . matched visual0.042.50.062.7and roll-lateralmotion2. delayed lateral0.0512.40.0373.3motion3. delayed roll and0.0554.1lateral motion
+Table 6 .6Hover perfonnance data with high fidelity washout configurationLongitudinal position error (ft)Lateral position error(ft)Pilot Motion configuration Average CTMax Min Average aMax Min1. matched visual-0.21 0.64 1.26 -1.08 0.10.44 1.27 -0.68and roll-lateralmotionA 2. delayed lateral0.80.89 2.53 -1.12 0.24 0.41 1.26 -0.79motion3. delayed roll and0.90.88 2.53 -1.4 0.32 0.47 1.26 -0.80_____lateral motion_______________________________________1. matched visual0.03 1.34 2.04 -3.13 -0.04 0.71 1.0 -2.2and roll-lateralmotionB 2. delayed lateral0.04 1.39 3.1 -2.42 -0.3 0.67 1.35 -1.32motion3. delayed roll and0.93 1.55 3.1 -3.1 0.810.67 2.4 -0.54______lateral motion________________________________________1. matched visual07T1L02 2.67 -1.72 ^035OA60.66 -1.57and roll-lateralmotionC 2. delayed lateral0.17 0.98 1.26 -2.47 -0.03 0.57 1.15 -1.51motion3. delayed roll andlateral motion
+		
+		
+			
+
+				
+
+
+	
+		The Determination of Some Requirements for a Helicopter Flight Research Simulation Facility
+		
+			JBSinacori
+		
+		STI-TR-1097-1
+		
+			September 1977
+		
+	
+	Contractor Report
+	Sinacori, J. B..: The Determination of Some Requirements for a Helicopter Flight Research Simulation Facility. Contractor Report STI-TR-1097-1, September 1977.
+
+
+
+	
+		Simulation evaluation of the effects of time delay and motion on rotorcraft handling qualities
+		
+			DavidMitchell
+		
+		
+			RogerHoh
+		
+		
+			AdolphAtencio, Jr.
+		
+		
+			DavidKey
+		
+		10.2514/6.1991-2892
+	
+	
+		18th Atmospheric Flight Mechanics Conference
+		
+			American Institute of Aeronautics and Astronautics
+			Jan. 1993
+		
+	
+	Mitchell, D. G.; and Hart, D. C.: Effects of Simulator Motion and Visual Characteristics on Rotorcraft Handling Qualities Evaluations. American Helicopter Society Conference on Piloting Vertical Flight Aircraft, Jan. 1993.
+
+
+
+	
+		Motion washout filter tuning - Rules and requirements (expert systems flight simulators)
+		
+			PeterGrant
+		
+		
+			LloydReid
+		
+		10.2514/6.1995-3408
+		AFHRL-TR-72-54
+	
+	
+		Flight Simulation Technologies Conference
+		
+			American Institute of Aeronautics and Astronautics
+			August 1995. May 1995. June 1973
+		
+	
+	Proceedings of the AGARD Flight Vehicle Integration Panel Symposium on Flight Simulation
+	Grant, P. R.; and Reid, Lloyd D: Motion Washout Filter Tuning: Rules and Requirements. AIAA Flight Simulation Technologies Conference, August 1995. ^Schroeder, J. A.; and Johnson, W. W.: Yaw Motion Cues in Helicopter Simulation. Proceedings of the AGARD Flight Vehicle Integration Panel Symposium on Flight Simulation, May 1995. ^Gum, D. R.: Modeling of the Human Force and Motion-Sensing Mechanisms. Air Force Human Resources Laboratory, AFHRL-TR-72-54, June 1973.
+
+
+
+	
+		A Piloted Simulation of Helicopter Air Combat to Investigate Effects of Variations in Selected Performance and Control Response Characteristics
+		
+			MSLewis
+		
+		
+			MHMansur
+		
+		
+			RT NChen
+		
+	
+	
+		^Danek, G. L.: Vertical Motion Simulator Familiarization Guide. NASA TM-103923
+		
+			1987. July 1994. Jan. 1988. May 1993
+		
+	
+	NASA TM 100084
+	Lewis, M. S..; Mansur, M. H..; Chen, R. T. N.: A Piloted Simulation of Helicopter Air Combat to Investigate Effects of Variations in Selected Performance and Control Response Characteristics. NASA TM-89438, 1987. 'Aeronautical Design Standard, Handling Qualities Requirements for Military Rotorcraft. ADS-33D, July 1994. ^McFarland, R. E.: Transport Delay Compensation for Computer-Generated Imagery Systems.. NASA TM 100084, Jan. 1988. ^Danek, G. L.: Vertical Motion Simulator Familiarization Guide. NASA TM-103923, May 1993. 10
+
+
+
+	
+		A Simulation Investigation of Motion Cueing and Visual Time Delay Effects on Two Helicopter Tasks
+		
+			DCHart
+		
+		
+			DGMitchell
+		
+		
+			GEHarper
+		
+		
+			RPJr
+		
+		NASA TN D-5153
+	
+	
+		The Use of Pilot Rating in the Evaluation of Aircraft Handling Qualities
+		
+			April 1996. Apr. 1969. 12
+		
+	
+	NASA TM 110385
+	Hart, D. C.; and Mitchell, D. G.:A Simulation Investigation of Motion Cueing and Visual Time Delay Effects on Two Helicopter Tasks. NASA TM 110385, April 1996. ^Cooper, G. E.; and Harper, R. P., Jr.: The Use of Pilot Rating in the Evaluation of Aircraft Handling Qualities. NASA TN D-5153, Apr. 1969. 12
+
+
+
+	
+		Frequency-Response Method for Rotorcraft System Identification: Flight Applications to BO 105 Coupled Rotor/Fuselage Dynamics
+		
+			MarkBTischler
+		
+		
+			MavisGCauffman
+		
+		10.4050/jahs.37.3
+	
+	
+		Journal of the American Helicopter Society
+		j am helicopter soc
+		2161-6027
+		
+			37
+			3
+			
+			July 1992
+			American Helicopter Society
+		
+	
+	Tischler, M. B.; and Cauffman, M. G.: Frequency Response Method for Rotorcraft System Identification: Flight Applications to the BO-105 Coupled Rotor/Fuselage Dynamics. Journal of American Helicopter Society, Vol. 37, No. 3, pp. 3-17, July 1992.
+
+
+				
+			
+		
+	

file147.txt

@@ -0,0 +1,424 @@
+
+	
+	
+		
+IntroductionLow cost alternatives to traditional motion platforms have been sought to provide motion cues in ground-based flight simulators to meet mission objectives.One method that has been shown to be effective is the dynamic seat, which provides high-frequency/low-amplitude motions at the pilot station.Subjectively, high frequency vibration cues provide familiar cockpit oscillations due to structure, rotor dynamics, and airspeed for a helicopter flight simulation.Objectively, the limited onset cues may aid the pilot to develop similar control strategies in meeting mission requirements.Previous studies 1 ' 2 have shown that there are benefits in using limited-travel vibration devices in helicopter simulations, especially as a training device.White 1 found there was a significant difference in collective activity in a bob-up task using an idealized helicopter simulation with and without a g-seat.The g-seat had two independent actuators in heave degree-of-freedom (DOF) and was mounted on a three DOF motion platform, i.e., heave, pitch, and roll.The cockpit had a field-of-view (FOV) of 48 degrees in azimuth and 36 degrees in elevation.White also reports that pilots were more consistent in maintaining a linear relationship between collective activity and time to impact in a hurdle task with the g-seat.Pilot comments in this study gave preference to the use of the g-seat.Greig 2 investigated the effectiveness of a multi-axis dynamic seat in the simulation of a Lynx helicopter on the Large Motion System (LMS) at UK's Defence Research Agency Advanced Flight Simulator (AFS).The dynamic seat had 5 independent hydraulic actuators to produce three DOF motion in heave, surge, and sway.The LMS has five DOF, i.e., heave, sway, roll, pitch, and yaw, and a FOV of +/-63 degrees in azimuth and 24 degrees in elevation.The study found that subjective pilot ratings and comments favor the use of a dynamic seat in the five tasks evaluated, i.e., sidestep, quick hop, lateral jinking, spot turn, and NoE course.The Joint Shipboard Helicopter Integration Process (JSHEP), a Navy program sponsored by the Office of the Secretary of Defense, was initiated to investigate the minimum groundbased simulation requirements to develop the launch and recovery operational envelope.Among many JSHIP investigation objectives, a multi-axis dynamic seat, Figure 1, that was similar to Greig's investigation was one of the simulation cueing devices evaluated.For this purpose, a UH-60 Black Hawk motion-based flight simulation experiment was developed at NASA Ames Research Center's Vertical Motion Simulator (VMS), Figure 2, using six ADS-33D 3 maneuvers.The JSHIP simulator cockpit has a FOV of 220 degrees in azimuth and 70 degrees in elevation.Four different motion cueing levels were chosen to investigate the effects of the dynamic seat.The effectiveness of the dynamic seat was then determined by comparing pilots' workload, the perceived vehicle performance, and task performance in six selected maneuvers.
+Experiment Description
+Math ModelA high fidelity mathematical model of the UH-60A Black Hawk known as Gen Hel 4 was used in the investigation.The real-time simulation had a frame rate of 100 Hz.In hover and low speed, the Black Hawk was configured to have an augmented angular rate command system, and the collective controlled vertical acceleration.The angular rate frequency responses at hover generated by a handling qualities analysis program, CIFER 35 , are shown in Figure 3, and the heave control response is shown in Figure 4.
+Motion CueingFour levels of motion cueing were developed to investigate the effects of the dynamic seat.They are: I.The 3-DOF dynamic seat: Uses all three DOF of the dynamic seat, i.e., heave, sway, and surge.The dynamic seat provided high frequency heave and lateral vibrations, onset cues for heave, sway, and surge, and sustained sway and surge motion cues.II.Hexapod-like travel: The VMS was driven by adaptive motion drive algorithms developed for a hexapod motion system 6 ' 7 with six 60-inch stroke actuators.III.Hexapod-like travel plus dynamic seat with only heave mode: The VMS was driven the same way as Level II.The dynamic seat was activated in heave DOF only as a seat shaker to provide the vertical vibration cues.IV.Large motion travel plus 2-DOF dynamic seat: Full VMS travel was utilized to achieve the best possible motion fidelity.VMS was driven by the standard classical motion drive algorithms.The dynamic seat was activated in two DOF, i.e., heave and sway, to supplement the large motion travel with high frequency vibration cues.The dynamic seat commands, which provided sustained surge and sway components, were disabled.Level I motion represents a low-cost option in providing motion cues.Level II represents a motion cueing fidelity that is common to the training community.With the addition of a seat shaker feature, any difference between Level II and IE could be attributed directly to the effect of high frequency heave vibration.Level IV represents the best possible ground-based motion cueing fidelity by using the full translational travel envelope of the VMS.Displacement, rate, and acceleration limits of the VMS and a hexapod-like system are shown in Table 1.The smallamplitude frequency responses of the VMS are plotted against the FAA Advisory Circular 120-63 8 motion specifications as shown in Figure 5.The motion fidelity according to Ref. 9 for all six DOF is shown in Figure 6.Another important motion fidelity factor, the lateral translational motion relative to simulator roll motion, to maintain the proper specific force direction, is low for the hexapod-like case (Level n and III), and is high for the large motion case (Level IV), according to Ref. 10.
+Motion Cueing -Dynamic SeatA multi-axis dynamic seat 11 provided by the Army Apache Training Command was integrated in one of the VMS's inter-changeable cabs.The dynamic seat has four independent actuators to provide three DOF of motion, i.e., heave, sway surge, and.The performance of each actuator is shown in Table 2.The small-amplitude frequency responses of the four actuators are shown in Figure 7.The high frequency heave vibration cues were generated by the seat pan and driven directly according to four per rev of the UH-60 rotor rpm, i.e., at 17 Hz.According to pilot comments, one per rev high frequency lateral vibration cues were added to the back pad to mimic the UH-60 cockpit vibration characteristics during flight.The magnitude of heave vibrations was adjusted based on the Bob-Up/Bob-Down flight test data.The dynamic seat's gains and frequency content were adjusted to match the power-spectral density of the vertical acceleration sensor response taken from the flight test as shown in Figure 8.The onset cues in heave due to pilot control inputs and/or flight conditions have four components, which are translational lift, collective, normal acceleration, and airspeed.The translational lift provides the vibrations due to the change in inflow orientation between the forward and aft portions of the rotor disk in the speed range between 20 and 30 knots.Sustained sway acceleration cues were developed by moving the back pad laterally as a function of pilot-station lateral accelerations.Onset lateral acceleration cues were generated by feeding roll angular acceleration and the high frequency component of lateral acceleration to drive the back pad in lateral motion.Sustained deceleration was generated by moving the back pad forward and the seat pan downward synchronously.Sustained acceleration was developed by moving the back pad aft and the seat pan upward together.Onset longitudinal acceleration cues were generated by feeding pitch angular acceleration and the high frequency component of longitudinal acceleration to drive the back pad fore and aft.
+Visual CueingThe cockpit, as shown in Figure 9, with a wide field-of-view (FOV) display system, producing 220 degrees in azimuth and 70 degrees in elevation, was specially designed and developed for the JSHIP experiment.The primary image generation system is a five-channel E&S ESIG 4530 system operating at 60 Hz with a transport delay measured at 60 msec.The projection system used a projector-mirror design with five BARCO projectors.A high resolution LHA visual model, LHA-5 USS Peleliu, was used for all test maneuvers.The model consists of 3000 textured polygons and employs 4 levels-of-detail.An E&S 3-Dimensional (3D) sea wave model provided additional wave dynamics relative to wave heights and period.
+Aural CueingThe simulator cab had a stereo sound system with six speakers and one sub-woofer around the pilot to provide high quality aural cues that included main rotor, tail rotor, engine, American Institute of Aeronautics and Astronautics transmission, air, and landing gear as functions of collective control and flight conditions.Sound cues were evaluated by UH-60 pilots and were found to be representative of the UH-60 in test tasks evaluated.
+Task DescriptionSix maneuvers modified from ADS-33D for shipboard operations were evaluated in the investigation.They were Acceleration/Deceleration, Bob-up/Bob-down, Hover, Pirouette, Sidestep, and Vertical Landing.Descriptions of maneuvers and performance criteria are presented in Ref. 12. Four experienced Army test pilots participated in this evaluation.An additional test was done fixed-base with the dynamic seat on and off using a modified Bob-Up/Bob-Down maneuver to evaluate the effectiveness of the dynamic seat independent of platform motion.Instead making a Bob-Down maneuver immediately after a brief stabilization at the top, pilots were instructed to maintain stabilization for at least 10 seconds before initiating a Bob-Down.Three UH-60 pilots (two NASA and one Army) participated in this test.
+Results
+Subjective EvaluationsHandling Qualities Rating (HQR) 13 results for the six ADS-33D maneuvers are shown in Figure 10.Results from the 3-DOF dynamic seat, Level I, compare well with the large motion plus 2-DOF dynamic seat, Level IV, except Acceleration/Deceleration and Sidestep, where maneuvers in surge and sway DOF are more dominant.Heave vibration cues do improve the HQR for most of the maneuvers when comparing Level III motion with Level II motion.HQR results for the fixed-base Bob-Up/Bob-Down task with the dynamic seat on and off are shown in Figure 11.A Motion Fidelity Scale 9 (MFS), as shown in Table 3, was used to subjectively determine consistency between perceived visual cues and motion cues.MFS results with the seat on and off are also shown in Figure 11.
+Objective Performance DataObjective performance data were analyzed for two test maneuvers, i.e., Bob-Up/Bob-Down, and Vertical Landing.Both maneuvers emphasized the vertical DOF, which was relevant to VMS large motion and the dynamic seat's primary motion cueing characteristic, i.e., heave.In the Bob-Up/Bob-Down task, the simulated Black Hawk's altitude offset at the lower hover position was analyzed to investigate the pilot's altitude stabilization performance after the bob-down.Maximum descent speed was also analyzed to investigate the pilot's vertical speed control relative to the bob-down task.Both results are shown in Table 4.In the Vertical Landing task, the pilot's landing spot offset in longitudinal and lateral directions were analyzed as well as the maximum descent speed.Results are shown in Table 5.In the fixed-base Bob-Up/Bob-Down test, the simulated Black Hawk's altitude offset at the lower hover position and the maximum descent speed with and without the use of the dynamic seat are shown in Table 6.Power spectral density (PSD) of the collective and pilot's cut-off frequency were analyzed to characterize the pilot's inner-loop response that was related to work load and the task.The PSD directly reflects pilot control magnitude in the frequency domain.The cut-off frequency is defined as a measure of the pilot's control activity bandwidth.When the aircraft's bandwidth exceeds the task bandwidth, the pilot cut-off frequency approaches the pilot crossover frequency and gives a good approximation of the task bandwidth. 14he purpose of using these measurements was to investigate the motion cueing effects in pilot control strategy and aggressiveness.Studies have shown that improved motion fidelity has led to increases in pilot's gain and crossover frequency. 15' 16 Consequently, higher pilot gain leads to lower control PSD.Average Root-Mean-Square (RMS) of the collective PSD and average pilot cut-off frequencies for four different levels of motion cueing conditions are shown in Table 4 for the Bob-Up/Bob-Down maneuver and in Table 5 for the Vertical Landing.Average RMS of the collective PSD, and pilot's cut-off frequency of the fixed-base Bob-Up/Bob-Down test are shown in Table 6.
+DiscussionSubjective Data -HQR As shown in Figure 10, according to the average HQRs, Level IV motion shows the best match with the flight test data among all six ADS-33 maneuvers.Level ffl also shows good results when compared with the flight test data.The differences between Level EH and IV are minimal.Overall, pilots gave good marks to Level IV on motion cueing fidelity, citing that there was no negative cueing and that the realism was good.Level I motion shows a good match in mean HQR with the flight test data in Hover and Vertical Landing tasks.In another vertical DOF task, Bob-Up/Bob-Down, the dynamic seat also fares well relative to the flight data with a mean HQR difference of 0.25 (A L _ I/Flight =0.25).Level I has the worst mean HQR in Acceleration/Deceleration (A L .IAnight =0.85) and Sidestep (A^^g^ =0.5) tasks, which may be attributed to the lack of motion travel in those two DOF.Level I also has the largest standard deviation in Pirouette (a L .!=1.29), Sidestep (0^=1.0), and Vertical Landing (ar =0.63).The widespread ratings suggest there is an inconsistency in pilots' determination in their workload and vehicle performance relative to the task.Some pilots commented that using the back pad to provide sustained sway cues was unnatural because only the upper body moved.Level II motion shows a poor match in mean HQR relative to the flight test data (A L .II/FHght >0.5) for Acceleration/Deceleration (A L _ H/FIight =0.65), Hover (A L ., I/Fliglu =0.75), and Sidestep (A L _ Il/FUglu =0.8).Level II has the largest standard deviation in Acceleration/Deceleration (a L _ n =0.96), and Bob-Up/Bob-Down (a L .n =1.15).Level III improves the mean HQR relative to flight test data in Acceleration/Deceleration (A L _ III/LII =0.25), Hover (A L .III/LII =0.75), Sidestep (A L .III/L .n =0.68), and Vertical Landing (A L _ m/L-ii =0.25) tasks.Level HI matches very well with the flight test's mean HQR in Bob-Up/Bob-Down (A L .III/Flight =0.17), Hover (A L .IiyFlight =0), Sidestep (A L _ III/Flight =0.12), and Vertical Landing (A L _ in/FUght =0.25).The results suggest that there is a benefit of having the high frequency heave vibration in a motion platform.Level IV, the large motion travel and the 2 DOF dynamic seat, matches well with the mean HQR from the flight test in Acceleration/Deceleration (A L _ IV/Flight =0.2), Bob-Up/Bob-Down (A L .IV/Flight =0), hover (A L _ IV/Flight =0.29), and Vertical Landing (A L .rv/Flight =0.14).
+Subjective Data -Fixed-BaseFrom Figure 11, with the dynamic seat on, the mean HQR of the Bob-Up/Bob-Down task improves by 0.5 relative to the seat-off condition.The standard deviation of the mean HQR with the seat on (o>seat-on=0-71) is also smaller than with the seat off (a Seat _on= 1.325).Both results indicate an improvement in pilots' workload and their determination of the vehicle performance when the dynamic seat was on.Motion Fidelity Scale results in Figure 11 show that pilots were less objectionable to the cueing differences between the flight response perceived from visual and the motion cues when the dynamic seat was on.All three pilots found the onset cues were helpful and recommended the use of the seat for the Bob-Up/Bob-Down task.Two of the pilots recommended the use of the vibration cues.Objective Data -Bob-Up/Bob-Down From Table 4, the average altitude stabilization error at the lower hover position after a bob-down for all four motion cueing levels are very similar and are well within the satisfactory performance criterion, i.e., +/-3 ft, for the task.Level IV motion has the smallest standard deviation (a L _ IV =0.33 ft), but differences are relatively small.There is little difference in average maximum descent speed among the four motion cueing levels.Level I motion and Level II motion, however, have larger standard deviations, i.e., 2.82 ft/sec and 3.17 ft/sec respectively, which indicates pilots were not as consistent in their vertical speed control.The mean standard deviations for the other two motion cueing conditions are 0.86 ft/sec for Level III andl.08 ft/sec for Level IV.There is very little difference in average collective RMS and pilot cut-off frequency in this task.With platform motion on, i.e., Level II, III, and IV, the data show a trend with lower collective RMS and higher pilot cut-off frequency as the motion cueing fidelity increases from Level II to IV.This trend is consistent with the concept that pilot's gain and crossover frequency increases as the motion cueing fidelity improves.The increased pilot gain subsequently leads to lower control RMS.The 3-DOF dynamic seat, Level I, however, has the lowest collective RMS and a pilot cut-off frequency higher than the two hexapod motion cueing conditions which contradicts the trend.One possible explanation could be found in the pilot comments where all pilots explicitly indicated that they relied more on visual cues such as the superstructure to judge the translational rate when platform motion was absent.
+Objective Data -Vertical LandingFrom Table 5, landing spot offsets in longitudinal and lateral directions for all four motion-cueing levels are similar.No obvious trends could be found.Only Level IV motion had an average longitudinal offset that was within the satisfactory performance criterion, i.e., +/-1 ft.There is an obvious trend in the average maximum descent speed, where the maximum descent speed decreases as the motion fidelity increases from Level I through Level IV.This result is consistent with the finding from a PIO study 17 and shows pilots are more conscious of the descent speed as the motion fidelity improves.The difference in average collective RMS and pilot cut-off frequency was relatively small among the four motion cueing levels.The large motion travel plus 2-DOF dynamic seat, Level IV, had the least average collective RMS and the pilot cut-off frequency suggests pilots might be easing off the collective due to pronounced vertical speed cues.The small standard deviations under the Level IV motion, i.e., 0.02 inch for collective RMS, 0.01 rad/sec in pilot cut-off frequency, and 0.77 ft/sec in the maximum descent speed, suggest pilots were more consistent in controlling the vertical speed in Level IV than in the other three levels.Objective Data -Fixed-Base From Table 6, the altitude error when stabilizing after the bob-down for the Bob-Up/Bob-Down task is improved when the dynamic seat is on, 1.12 ft vs. 1.52 ft when the dynamic seat is off.There is little difference in the other three objective measurements, which suggests the dynamic seat helps in improving realism of the Bob-Up/Bob-Down task and the task performance, but not pilots' perception of the vertical speed and their control activities.Frequency (rad/sec) Seat pan (heave) displacement response -------Fore-and-aft (surge) displacement response ----------Lateral (sway) displacement response ------Bucket (not used) displacement responsec)2001 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.
+Figure 1 .Figure 2 .Figure 3 .Figure 5 .Figure 6 .12356Figure 1. 3 degree-of-freedom (heave, surge, and sway) dynamic seat
+Figure 7 . 2 Figure 9 .729Figure 7. Dynamic seat actuator frequency response
+Table 1 .1VMS and Hexapod-Like operational limitsAxisDisplacementVelocityAccelerationVMSHexapod-LikeRoll±18±18±40±115Pitch±18±18±40±115Yaw±24±24±40±115Longitudinal±4±4±4±10Lateral±20±4±8±16Vertical±303.3 up/ 2.5 down±16±24All numbers, units ft, deg, sec
+Table 2 .2System limits of the Dynamic SeatSeat-Pan (heave)Back-Pad (sway)Back-Pad (Surge)Bucket (heave)Displacement± 0.59 inch± 0.59 inch± 0.59 inch± 0.59 inchVelocity± 2.4 in/sec± 2.4 in/sec±0.8 in/sec± 2.4 in/secAcceleration± 39.4 in/sec 2± 39.4 in/sec 2± 39.4 in/sec 2± 39.4 in/sec 2
+Table 3 .3Motion fidelity scaleDescriptionScoreHigh FidelityMotion sensations are not noticeably different1from those of visual flightMedium FidelityMotion sensations are noticeably different from2those of visual flight, but not objectionableLow FidelityMotion sensations are noticeably different from those3of visual flight and objectionableTable 4. Objective data for Bob-Up/Bob-Down taskBob-Up/Bob-Down3-DOF dynamic seat (Level I)Hexapod like only (Level II)Hexapod like + seat shakerLarge motion + 2-DOF dynamic(Level III)seat (Level IV)Altitude errorAverage1.451.361.11.41(lower hover1 standard0.570.460.640.33position),., ftdeviationMaximumAverage-10.83-11.47-11.87-11.68descent speed,1 standard2.823.170.861.08ft/secdeviationRoot-Mean-Average0.3720.4450.4350.415Square,1 standard0.100.120.080.12Collective,deviationinchesPilot cut-offAverage1.341.281.261.37frequency, rad/secdeviation 1 standard0.290.260.170.31American Institute of Aeronautics and Astronauticsc)2001 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.
+Table 5 .5Objective data for the Vertical Landing taskVertical Landing3-DOF dynamicHexapod likeHexapod like +Large motion +seat (Level I)only (Level II)seat shaker2-DOF dynamic(Level III)seat (Level IV)Landing spotAverage1.271.11.350.54offset,1 standard1.191.040.540.34longitudinal, ftdeviationLanding spotAverage1.051.181.191.29offset, lateral, ft1 standard0.740.80.731.33deviationMaximumAverage-4.77-4.55-3.72-2.87descent speed, ft/secdeviation 1 standard2.432.331.580.77Root-Mean-Average0.70.750.620.63Square,1 standard0.20.110.220.02Collective,deviationinchesPilot cut-offAverage0.930.9250.910.83frequency,1 standard0.160.150.080.01rad/secdeviation
+Table 6 .6Objective data for a Bob-Up/Bob-Down task in fixed-baseBob-Up/Bob-DownDynamic SeatDynamic Seat(Fixed-Base)OnOffAltitude errorAverage1.121.52(lower hover1 standard0.470.32position), ftdeviationMaximumAverage-13.30-13.69descent speed,1 standard3.323.19ft/secdeviationRoot-Mean-Average0.620.58Square,1 standard0.20.23Collective,deviationinchesPilot cut-offAverage1.221.20frequency,1 standard0.180.19rad/secdeviationAmerican Institute of Aeronautics and Astronautics
+		
+		
+			
+ConclusionsThere are benefits to use the dynamic seat in ground-based flight simulations.However, dynamic seat alone may not be adequate to meet certain mission requirements.Addition of high frequency heave vibrations to the hexapodlike system has positive effects both subjectively and objectively.Large motion travel with the 2-DOF dynamic seat has the closest representation of the flight.			
+			
+
+				
+
+
+	
+		G-seat heave motion cueing for improved handling in helicopter simulators
+		
+			ADWhite
+		
+		10.2514/6.1989-3337
+		AIAA-89- 3337-CP
+	
+	
+		Flight Simulation Technologies Conference and Exhibit
+		
+			American Institute of Aeronautics and Astronautics
+			1989
+		
+	
+	White, A.D.: "G-Seat Heave Motion Cueing for Improved Handling in Helicopter Simulators," AIAA-89- 3337-CP, 1989.
+
+
+
+	
+		Advanced thermal barrier coating system development. Technical progress report, September 1, 1996--November 30, 1996
+		
+			IGreig
+		
+		10.2172/560766
+	
+	
+		Defence Research Agency, United Kingdom, I/ITSEC 1996
+		Orlando, FL
+		
+			Office of Scientific and Technical Information (OSTI)
+			November, 1996
+		
+	
+	Greig, I.: "Evaluation of a Multi-Axis Dynamic Cueing Seat for Use in Helicopter Training Devices," Defence Research Agency, United Kingdom, I/ITSEC 1996, Orlando, FL, November, 1996.
+
+
+
+	
+		Appraisal of Rotorcraft Handling Qualities Requirements for Lateral-Directional Dynamics
+		10.2514/6.2021-0592.vid
+		
+			July 1994
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	Aeronautical Design Standard, Handling Qualities Requirements for Military Rotorcraft, ADS-33D, July 1994.
+
+
+
+	
+		UH-60A Black Hawk Engineering Simulation Program: Vol. I -Mathematical Model, NASA CR-166309
+		
+			JJHewlett
+		
+		
+			December 1981
+		
+	
+	Hewlett, J.J.: UH-60A Black Hawk Engineering Simulation Program: Vol. I -Mathematical Model, NASA CR-166309, December 1981.
+
+
+
+	
+		Frequency‐Response Method for Rotorcraft System Identification: Flight Applications to BO 105 Coupled Rotor/Fuselage Dynamics
+		
+			MarkBTischler
+		
+		
+			MavisGCauffman
+		
+		10.4050/jahs.37.3.3
+	
+	
+		Journal of the American Helicopter Society
+		j am helicopter soc
+		2161-6027
+		
+			37
+			3
+			
+			July 1992
+			American Helicopter Society
+		
+	
+	Tischler, M. B., Cauffman, M.G.: "Frequency-Response Method for Rotorcraft System Identification: Flight Applications to BO-105 Coupled Rotor/Fuselage Dynamics," Journal of the American Helicopter Society, Vol 37, No 3, pgs 3-17, July 1992.
+
+
+
+	
+		Coordinated Adaptive Washout for Motion Simulators
+		
+			RussellVParrish
+		
+		
+			JamesEDieudonne
+		
+		
+			RolandLBowles
+		
+		
+			DennisJMartin
+		
+		10.2514/3.59800
+	
+	
+		Journal of Aircraft
+		Journal of Aircraft
+		0021-8669
+		1533-3868
+		
+			12
+			1
+			
+			Jan., 1975
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	Parrish, R.V., Dieudonne, J.E., Bowles, R.L., and Martin, Jr., D.J., "Coordinated Adaptive Washout for Motion Simulators," Journal of Aircraft, Vol. 12, No. 1, Jan., 1975, pp. 44-50.
+
+
+
+	
+		Newton-Raphson Method
+		
+			JEDieudonne
+		
+		
+			RVParrish
+		
+		
+			REBardusch
+		
+		10.1007/springerreference_2034
+	
+	
+		NASA
+		
+			7067
+			1972
+			Springer-Verlag
+		
+	
+	Dieudonne, J.E.; Parrish, R.V.; and Bardusch, R.E.: "An Actuator Extension Transformation for a Motion Simulator and an Inverse Transformation Applying Newton-Raphson's Method", NASA TN D-7067, 1972.
+
+
+
+	
+		Spatial frequency and platform motion effects on helicopter altitude control
+		
+			JefferySchroeder
+		
+		
+			WilliamChung
+		
+		
+			RonaldHess
+		
+		10.2514/6.1999-4113
+	
+	
+		Modeling and Simulation Technologies Conference and Exhibit
+		
+			American Institute of Aeronautics and Astronautics
+			July 1999
+		
+	
+	NASA/TP-1999-208766
+	Schroeder, J.A.: "Helicopter Flight Simulation Motion Platform Requirements," NASA/TP-1999-208766, July 1999.
+
+
+
+	
+		Motion fidelity criteria for roll-lateral translational tasks
+		
+			JulieMikula
+		
+		
+			DucTran
+		
+		
+			WilliamChung
+		
+		10.2514/6.1999-4329
+		AIAA 99-4329
+	
+	
+		Modeling and Simulation Technologies Conference and Exhibit
+		Portland, Oregon
+		
+			American Institute of Aeronautics and Astronautics
+			August, 1999
+		
+	
+	Mikula, J.; Chung, W.W.; and Tran, D.: "Motion Fidelity Criteria for Roll-Lateral Translational Tasks," AIAA Modeling and Simulation Technologies Conference, Portland, Oregon, AIAA 99-4329, August, 1999.
+
+
+
+	
+		Development of a Multi-Axis Active Seat Mount System for Helicopter Aircrew Whole-Body Vibration Mitigation
+		
+			PCorlyon
+		
+		
+			THumphrey
+		
+		10.2514/6.2021-1835.vid
+	
+	
+		I/ITSEC 1999
+		Oriando, FL
+		
+			American Institute of Aeronautics and Astronautics (AIAA)
+			November, 1999
+		
+	
+	Corlyon, P. and Humphrey, T.: "Force and Vibration Cueing with a Multi-Axis Dynamic Seat," I/ITSEC 1999, Oriando, FL, November, 1999.
+
+
+
+	
+		The use of ADS-33D useable cue environment techniques for defining minimum visual fidelity requirements
+		
+			MichaelRoscoe
+		
+		
+			GeryVandervliet
+		
+		
+			ColinWilkinson
+		
+		10.2514/6.2001-4063
+		AIAA 2001-4063
+	
+	
+		AIAA Modeling and Simulation Technologies Conference and Exhibit
+		Montreal, Quebec, Canada
+		
+			American Institute of Aeronautics and Astronautics
+			August 2001
+		
+	
+	AIAA Modeling and Simulation Technologies Conference
+	Roscoe, M.F.; Wilkinson, C.H.; and VanderVliet, G.M.: "The Use of ADS-33D Useabie Cue Environment Techniques for Defining Minimum Visual Fidelity Requirements," AIAA Modeling and Simulation Technologies Conference, Montreal, Quebec, Canada, AIAA 2001-4063, August 2001.
+
+
+
+	
+		The Use of Pilot Rating in the Evaluation of Aircraft Handling Qualities
+		
+			GECooper
+		
+		
+			RPHarper
+		
+		
+			Jr
+		
+		NASA TN D-5153
+		
+			April 1969
+		
+	
+	Cooper, G. E., and Harper, R. P., Jr.: "The Use of Pilot Rating in the Evaluation of Aircraft Handling Qualities," NASA TN D-5153, April 1969.
+
+
+
+	
+		Investigation of the Effects of Bandwidth and Time Delay on Helicopter Roll‐Axis Handling Qualities
+		
+			ChrisLBlanken
+		
+		
+			Heinz‐ju¨rgenPausder
+		
+		10.4050/jahs.39.3.24
+	
+	
+		Journal of the American Helicopter Society
+		j am helicopter soc
+		2161-6027
+		
+			39
+			3
+			
+			July 1994
+			American Helicopter Society
+		
+	
+	Blanken, C.L. and Pausder H.-J.: "Investigation of the Effects of Bandwidth and Time Delay on Helicopter Roll- Axis Handling Qualities," Journal of the American Helicopter Society, July 1994, Vol. 39 No. 3, p24-33.
+
+
+
+	
+		Experiments and a Model for Pilot Dynamics with Visual and Motion Inputs
+		
+			RLStapleford
+		
+		
+			RAPeters
+		
+		
+			Alex
+		
+		
+			FR
+		
+		NASA CR-1325
+		
+			1969
+		
+	
+	Stapleford, R.L.; Peters, R.A.; and Alex, F.R.: "Experiments and a Model for Pilot Dynamics with Visual and Motion Inputs, " NASA CR-1325, 1969.
+
+
+
+	
+		Roll Tracking Effects of G-Vector Tilt and Various Types of Motion Washout
+		
+			HRJex
+		
+		
+			REMagdaleno
+		
+		
+			AMJunker
+		
+	
+	
+		NASA CP-2060
+		
+			November 1978
+			
+		
+	
+	Jex, H.R.; Magdaleno, R.E.; and Junker, A.M.: "Roll Tracking Effects of G-Vector Tilt and Various Types of Motion Washout," NASA CP-2060, November 1978, pp. 463-502.
+
+
+
+	
+		Simulator Platform Motion Effects on Pilot-Induced Oscillation Prediction
+		
+			JefferyASchroeder
+		
+		
+			WilliamW YChung
+		
+		10.2514/2.4578
+	
+	
+		Journal of Guidance, Control, and Dynamics
+		Journal of Guidance, Control, and Dynamics
+		0731-5090
+		1533-3884
+		
+			23
+			3
+			
+			May-June 2000
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	Schroeder, J.A.; and Chung, W.: "Simulator Platform Motion Effects on Pilot-Induced Oscillation Prediction," Journal of Guidance, Control, and dynamics, May-June 2000, Vol. 23, No. 3, p438-444.
+
+
+
+	
+		Subject and Author Indexes of Technical Papers Published in the AIAA Journals, Progress in Astronautics and Aeronautics, and Astronautics & Aeronautics in 1974
+		10.2514/3.49613
+	
+	
+		AIAA Journal
+		AIAA Journal
+		0001-1452
+		1533-385X
+		
+			12
+			12
+			
+			
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.
+
+
+				
+			
+		
+	

file148.txt

@@ -0,0 +1,376 @@
+
+	
+	
+		
+I. Nomenclature
+II. IntroductionWhile flying simulated vehicles, pilots adapt to different stimuli provided in a simulator, e.g., out-the-window visual, audio, motion, and hand-controller force feedback, depending on the task or maneuver.Timely motion feedback through the motion platform, as well as feedback from the force-feel system, can provide lead compensation in closed-loop control tasks and improve handling qualities and task performance [1][2][3][4][5].Force-feel characteristics, such as the breakout, dead-band, damping, force gradient, and inertia of the controller all play an important role in the handling qualities of a (simulated) rotorcraft.Previous research focused on cyclic inceptor force-feel characteristics for improved handling qualities for both passive and active controllers [6][7][8].Different from previous investigations, this study investigated haptic cues that are missing in fixed-base flight simulators that could contribute to tactile feedback pilots would have experienced otherwise in real flight.Specifically, this study focused on the inertial forces and moments a cyclic inceptor experiences due to the aircraft's motion that are either missing completely in fixed-base flight simulators, or being attenuated due to the application of washout filters in motion-based simulators.In an experiment conducted in the Vertical Motion Simulator (VMS) at NASA Ames Research Center, the effects on task performance, control behavior, and handling qualities ratings were investigated when restoring these reaction forces in fixed-base flight simulators.The paper is structured as follows: Section III mathematically derives the control force compensation missing in fixed-base simulation.Section IV provides an overview of the experiment setup, including verification of the dynamic models, and the experimental hypotheses.The results are presented in Section V and discussed in Section VI.Finally, conclusions are provided in Section VII.
+III. Control Force CompensationThe inertial force from the dynamics of the simulated rotorcraft's lateral acceleration, a y , and roll angular acceleration, p b , Fig. 1a, are translated through the center control stick to the pilot as shown in Fig. 1b for the lateral and roll degrees-of-freedom.The equations of motion of the lateral stick response due to the simulated rotorcraft's lateral and roll angular accelerations are described by Eqs. 1 and 2 as a function of control force-feel system damping, ζ y , and force gradient, k y .I xx φ a y = -m c a y r c -ζ y φ a y -k y φ a y(1)I xx φ pbd = -I xx p b -ζ y φ pbd -k y φ pbd(2)The lateral stick displacements, due to lateral accelerations, a y , and angular accelerations, p b , are defined by Eqs. 3 and 4, respectively.Φ a y = - m c r c I xx a y / s 2 + ζ y I xx s + k y I xx(3)Φ pbd = -p b / s 2 + ζ y I xx s + k y I xx(4)The total compensation for the lateral control, Φ c , or what is missing in a fixed-based simulation is the sum of these two dynamic components as shown in Eq. (5).Φ c can be added to the control trim position to move the lateral stick in addition to pilot control inputs to simulate the stick force response due to the inertial force and moment from the rotorcraft's dynamics.Φ c = Φ a y + Φ pbd(5)
+IV. Experiment SetupTo investigate if control force compensation affects pilot control behavior and performance, a two degrees-of-freedom (DOF) lateral side-step task was developed in the VMS with the aim to compare results from no-motion conditions with and without control force compensation.The results were further compared with data from a near one-to-one motion condition, and a medium-fidelity motion condition representing motion found in typical hexapod motion simulators.
+A. Rotorcraft DynamicsA simulated linear 2-DOF rotorcraft model was adopted from previous work investigating the effects of roll-lateral motion on pilot performance [3].The simulated rotorcraft model is provided in Eqs.6 and 7.φ = -4.5 φ + 1.7δ lat (6) v = g sin φ(7)A second rotorcraft model, a full nonlinear UH-60 model [9] with a heavier force gradient of the lateral stick, was also used to investigate if there is a different effect from control force compensation depending on the controlled dynamics.For the UH-60, the Stability Augmentation System (SAS) was turned off purposely with an intent to force pilots to maintain close-loop stability.A comparison of the agility of the two rotorcraft models is shown in Fig. 2.
+B. Control Force-Feel SystemTo determine the mass and moment of inertia of the control stick in Eqs.1-5, a pull test was conducted by setting the damping of the stick to zero and releasing the stick at its maximum lateral travel limit.The force gradient of the linear model was set to 0.6 lbf/in as in a previous test [3].The resulting undamped natural frequency of the stick, ω n , was measured at 1.194 Hz (or 7.5 rad/s).Using Eq. ( 1), with zero external force and zero damping based on a 25 inches rotational reference center for the particular control loader system, the moment of the inertia of the lateral stick was calculated from Eqs. 8 to be 2577 lbm-in about the lateral stick's rotational center.The effective damping ratio of the stick would come out to be 0.85.With these properties, the stick's lateral characteristics would be acceptable according to Ref. [8].A check of the pull-release response of the measured moment inertia vs. the lateral stick response from the simulator is shown in Fig. 3.ω n = k y /I xx or I xx = k y /ω 2 n (8)The product of mass and the rotational center to the center of c.g. of the stick, m c r c in Eq. (3), was estimated through an experimental process since it was difficult to disassemble the roll rotational assembly from the control loader system.The VMS was configured with near one-for-one motion by setting the second-order high-pass washout filter frequencies for both the roll and lateral degree of freedom (DOF) to 0.001 rad/s with a unity gain.A doublet of lateral acceleration was commanded to drive the motion system.By comparing the stick displacement response and the compensation model, m c r c , is estimated to be 52 lb min 2 .With this estimate, the resulted lateral stick displacement compensation shows a good match to the simulator response as shown in Fig. 4. The compensation due to a roll angular acceleration also shows a good match as shown in the same figure.The lateral stick's control-force feel for the UH-60 had a breakout force of 1 lbf, a damping ratio of 0.722, and a force gradient of 1 lbf/in, which resulted an undamped natural frequency of 9.7 rad/s.This would also be acceptable according to Ref. [8].
+C. TaskA 20-foot side-step task was developed to investigate the effect of the control force compensation to the stick due to the lateral accelerations and roll angular accelerations in ground-based flight simulators.Altitude hold, heading hold, as well as forward position hold were assumed to be active to limit the DOFs to roll and lateral only as defined by the model described in Eqs. 1 and 2.All test participants were briefed on the test procedures and task performance criteria prior to taking the test.Test participants were instructed to initiate a smooth lateral input toward the station-keeping position 20 feet to the right in 5 seconds (or time-to-target) for desired performance, and 7 seconds for adequate performance.Once the simulated rotorcraft was visually within the desired station-keeping region, test participants were directed to maintain the station-keeping position in a light disturbance for 10 seconds.Desired, adequate, and not-adequate performance criteria for the station-keeping phase of the task were given to the test participants as shown in Fig. 5.
+D. ConditionsTo investigate the effect of inertial stick force compensation and its interaction with aircraft dynamics, the experiment had two independent variables: motion fidelity with four levels (high-fidelity, medium-fidelity, low-fidelity, and low-fidelity with compensation) and rotorcraft model with two levels (linear and UH-60 dynamics).The experiment had a full-factorial design resulting in eight experimental conditions.The motion configurations were plotted against the modified Sinacori motion fidelity criteria [3] are shown in Fig. 6.The motion gain for the high-fidelity configuration was reduced from unity to 0.8 to alleviate excessive lateral accelerations from the UH-60 model.The test matrix is shown in Table 1.
+E. Participants and ProceduresSeven pilots participated in the experiment.All had extensive rotorcraft flying experience.Every pilot received an extensive briefing and safety-walk-around before the start of the experiment.The original experiment was divided into two sessions.Each session tested the experimental conditions for one rotorcraft model.Training was provided to all test participants at the beginning of each session.Six simulated runs of each test configurations were given to test participants in random order.However, due to an error in the implementation of the control force compensation, the low-fidelity conditions were repeated by the seven test participants several weeks after the first trial of testing.During the second trial of testing, the low-fidelity conditions were presented randomly in two sessions as well.The two motion conditions were not repeated.
+F. ApparatusThe VMS, Fig. 7a, with its large motion envelope provides the realistic cueing environment necessary for performing handling qualities studies, has an operational lateral travel of 30 feet.The simulator was positioned to the left-side of the
+G. Dependent MeasuresThe experiment considered two subjective handling qualities ratings (HQR) [10] as dependent measures: the HQR rating during the transnational phase of the task (HQR tr ans ) and the HQR rating during the station keeping phase (HQR sk ).These ratings were collected in separate runs for conditions 1, 3, and 4 only at the end of the experiment.The HQR were collected only for the linear model.The following eight objective performance variables were considered as dependent measures: the time-to-target, t g ; the root mean square (RMS) of pilots' control inputs, RMS u ; the bandwidth of the control inputs, ω c f [11]; the station keeping score, SK; the RMS of the lateral position and velocity during the station-keeping phase, RMS ye and RMS v ; and the RMS of the roll angle and rate during the station-keeping phase, RMS φ and RMS p respectively.The time-to-target, t g , was the time between two button presses on the center stick.Pilots pressed the event button the first time when they were ready to start transitioning to the hover target.They pressed the button a second time when they felt they were within the desired hover bound and would likely stay in the desired area.Both ω c f and RMS u were calculated for the entire run, i.e., for both the translation and the station-keeping phases.The station-keeping score, SK, ranged from 1 to 3. A score of 1 was given to desired performance achieved during the station-keep phase of the task, 2 was given to adequate performance, and 3 was given to not-adequate performance (Fig. 5).The remaining objective measures (RMS ye , RMS v , RMS φ , and RMS p ) were all calculated for the station-keeping phase only.The lateral position error, RMS ye , was calculated relative to the center of the hover-target board (Fig. 5).
+H. HypothesesThe following hypotheses were formulated based on the characteristics of the control force compensation as derived in Section III and the controlled dynamics presented in Section IV.A: H1: Since the control force compensation in the fixed-base condition is simulating the inertial force and moment feedback due to the motion of the simulated aircraft, it was expected that pilots performing the task with compensation would control more similarly to performing the task under medium-fidelity or high-fidelity motion compared to performing the task without compensation.This would be visible in ω c f and RMS u .H2: As the control force compensation provides lead information similar to motion feedback, albeit less efficient, it was expected that the compensation would result in improved task performance compared to the fixed-base condition without compensation, i.e., a shorter time-to-target t g ; smaller RMS of the lateral position error RMS ye , roll angle RMS φ , lateral velocity RMS v , and roll rate RMS p during the station-keeping task; and a better station-keeping score SK.H3: Since the higher force gradient of the UH-60 model would result in a smaller magnitude of force compensation at the stick, it was hypothesized that pilots controlling the UH-60 model would benefit less from the control force compensation as compared to the linear model, i.e., the performance improvement would be less for the UH-60 model compared to the linear model and control behavior would be more similar to the condition without compensation.H4: It was expected that both high-fidelity and medium-fidelity motion would still provide larger improvements in pilot performance compared to both low-fidelity conditions, as motion provides faster lead information as compared to haptic feedback.Therefore, performance in the low-fidelity condition with compensation was expected to lie between performance in the medium-fidelity and low-fidelity-without-compensation conditions.
+V. ResultsIn this section, the main results of the experiment and the associated data analysis are presented.The seven continuous dependent measures under consideration are: t g , RMS u , ω c f , RMS ye , RMS v , RMS φ , RMS p .The three ordinary variables considered are the performance score SK and the HQR rating for the translation and station keeping phases, respectively.Table 2 and Table 3 present the means and standard deviations of the dependent measures for each condition with the exception of the HQR ratings, which deviated strongly from normality.Table 4 provides the means of the data collapsed over the rotorcraft models for the same dependent measures.
+A. Ordinal Variables
+HQR RatingsThe participants were asked to subjectively evaluate the handling qualities of the linear model using the Cooper Harper rating.They were asked to rate the translation and station keeping phase of the task separately (HQR tr ans and HQR sk , respectively).The assigned scores ranged from 1 to 10, with 1 indicating the best handling characteristics and 10 the worst [10].The pilot responses are shown in Figs.8a and8b.The Jonckheere trend test [12] was used to test the hypothesis that high-fidelity motion would receive the lowest scores, followed by the fixed-base condition with compensation, followed by the fixed base condition without compensation, in this order.The results of the test show a non-significant trend: JT = 57, p = 0.1304.Furthermore, a generalized linear model based on Generalized Estimating Equations (GEE based on logistic regression) was fit to the data, using   the Gaussian family and as cluster variable the participant ID.The planned contrasts for the model compared the high fidelity case against the low-fidelity ones and the low-fidelity cases with and without compensation against each other.For the variable HQR tr ans , the first contrast shows that there is no significant difference between high-fidelity and (combined) low-fidelity conditions.The same holds for the difference between the low fidelity with compensation and without.For the variable HQR sk , the first contrast shows that there is no significant difference between high-fidelity and (combined) low-fidelity conditions.The same hold for the difference between the low-fidelity with compensation and without.The full test results are shown in Table 5.The means and standard errors of the aggregated data are summarized in Table 6.
+Station-Keeping Score, SKThe station-keeping score, or SK, ranged from 1 to 3. A score of 1 indicates desired performance during the station-keep phase of the task, 2 indicates adequate performance, and 3 indicates not-adequate performance as shown in Fig. 5.The data, aggregated over motion-fidelity and aircraft-model variables, are shown in histograms in Fig. 9. Since the data are ordinal and highly non-normal a generalized linear model based on Generalized Estimating Equations (GEE based on logistic regression) was fit to the data, using the Poisson family with logarithmic link function and as cluster variable the participant id.The summary of the statistical results derived from the model are shown in Table 7.The rotorcraft model suggestively affected the score; b = -0.077,χ 2 = 3.308, p = 0.069.The mean SK rating for the UH-60 model (M = 1.25) was lower than for the linear model (M = 1.35).
+B. Continuous VariablesFor the continuous variables linear regression models were used for hypothesis testing, having as independent within-subject variables the simulation fidelity (Fidelity) and aircraft model (Model).First, a repeated-measures Analysis of variance (ANOVA) was conducted on the continuous variables.Unfortunately, most fidelity conditions violated the assumption of homogeneity of variances of the residuals.For this reason, a linear mixed-effect model was fit for all the continuous variables since it does not assume homogeneity of variances of the residuals and can account for the residual dependency by using the pilot ID as a random factor [13]. Adding pilot ID as the random factor significantly improved every model that is discussed further.To compare specific conditions without correcting for multiple comparisons, orthogonal constants where used [14].The orthogonal contrasts considered were the same for all the conditions: 1) High fidelity vs. others compares the mean of the high-fidelity against the aggregate mean of the all the other conditions.2) Medium fidelity vs. low compares the mean of the medium-fidelity condition against the aggregate mean of both the low-fidelity conditions.3) Low fidelity with vs. without compensation compares the mean of the low-fidelity condition with and without compensation.The analysis of variance test results are summarized in Table 8.The overall effect sizes are tabulated in Table 9 to    Table 14, and represented graphically in Figs. 10 and11.
+Time-to-Target, t gThe error-bar plot of the time-to-target for each of the conditions can be seen in Fig. 10a.The summary of the linear mixed effect model is provided in Table 9.The different aircraft models did not have a significant effect on the time-to-target.The time-to-target was significantly lower between the medium-fidelity (M = 4.38) and the low-fidelity conditions (M = 4.6 s); b = -0.115(SE = 0.048), t(29.3)= -2.401,p = 0.023.
+Cut-Off Frequency, ω c fThe error-bar plot Of the control input cut-off frequency for each of the conditions can be seen in Fig. 10b.The summary of the linear mixed-effect model is shown in Table 10.There was a significant interaction effect between the rotorcraft model type and the high-fidelity vs. other contrast; b = -0.137(SE = 0.044), t(18) = -3.146,p = 0.006.As can be seen in Fig. 10b at low-and medium-fidelity conditions, the cut-off frequency for the UH-60 was systematically higher than the one for the linear model while for the high fidelity case this difference disappeared.The cut-off frequency was significantly higher with high-fidelity compared to the other conditions; b = 0.417 (SE = 0.045), t(28, 22) = 9.336, p < 0.001.Moreover, the cut-off frequency for the low-fidelity condition with compensation (M = 2.71 rad/s) was significantly higher than the one for the low fidelity condition without compensation (M = 2.46 rad/s); b = 0.231 (SE = 0.109), t(18) = -3.146,p = 0.044.
+RMS Control Input, RMS uThe error-bar plot of time-to-target for each of the conditions is depicted in Fig. 10c.The summary of the linear mixed effect model is shown in Table 11.There was a significant interaction between the aircraft model and the mediumvs.low-fidelity conditions; b = -0.057(SE = 0.023), t(18) = -2.466,p = 0.024.Even in presence of an interaction, the overall mean of the RMS u for the linear model (M = 0.529) was significantly higher than the one for the UH-60 (M = 0.436); b = -0.093(SE = 0.034), t(6) = -2.744,p = 0.034.The RMS u for the high-fidelity case (M = 0.37) is lower than the one for the medium (M = 0.44) and low fidelity cases (M = 0.56), although the effect is not significant, p = 0.055.
+RMS Lateral Position Error, RMS yeThe error-bar plot of the RMS lateral position error for each of the conditions can be seen in Fig. 11a.The summary of the linear mixed effect model is provided in Table 12.The RMS ye for the linear model (M = 1.66) was significantly higher than the RMS ye for the UH-60 model (M = 1.33); b = -0.311(SE = 0.038), t(5.94) = -3.2,p = 0.019.Furthermore, the statistical analysis reveals that the mean RMS ye of low fidelity with compensation (M = 1.45) was significantly lower than the one of the same fidelity but without compensation (M = 1.7); b = -0.0252(SE = 0.094), t(35.62)= -2.691,p = 0.011.Indeed, for the UH-60 the means for the low fidelity case with (M = 1.429) and without compensation (M = 1.434) were quite similar.On the other hand, for the linear case the means for the low fidelity case with (M = 1.465) and without compensation (M = 1.970) showed a difference.
+B. SummaryThe rotorcraft model introduced significant differences in the dependent measures.These differences were probably caused by the fact that the linear model had a responsive first-order rate response, while the UH-60 model had a somewhat unsteady rate command response by purposely turning off the SAS to force pilots to stay in the loop to expose the effect of the controller compensation.This, however, might unexpectedly have led to pilots staying low-gain during the station-keeping phase.In addition, the higher force gradient of the control inceptor in the UH-60 resulted in smaller force compensations relative to the linear model with a lower force gradient.With the combination of these two factors, the benefit of having the lead provided by the force compensation was found to be significantly reduced with the UH-60 model as observed by comparing the control input cut-off frequency and lateral position error between the low-fidelity conditions with and without the force compensation.The contrast comparing the high-fidelity condition with the rest of the fidelity conditions, and its interactions showed significant differences in the following variables: ω c f , RMS u , RMS v , RMS φ .Pilots had a higher control input cut-off frequency for the motion conditions.This might be a result of the enhanced lead information motion provides as shown by previous studies [1][2][3][4][5].In this case, the simulation cueing feedback from the visual and motion were consistent with little to no phase error between them.Even though not significant, the lower RMS u for motion conditions compared to the conditions without motion could suggest a lower control activity with motion.Furthermore, motion resulted in lower values for RMS v and RMS φ , indicating that pilots could more easily stabilize the aircraft with motion.The medium-fidelity motion condition introduced significant differences compared to both low-fidelity conditions for t g and RMS u .The time-to-target t g is lower for the medium-fidelity motion condition, even though t g for the high-fidelity condition is not significantly different from the rest.Since the time-to-target was measuring the time of a 20-feet sidestep to the hover target, which involved taking out the lateral velocity before stabilizing and hovering, the medium-fidelity motion was able to provide the comfort for pilots to generate a larger bank, or lateral acceleration, to get to the hover target and with the right amount of damping needed to stabilize the simulator via the motion feedback.The RMS control input RMS u was also different for the medium-fidelity motion condition compared to the other conditions due to an interaction with the aircraft model: for the UH-60, RMS u was significantly lower than in the low-fidelity cases, while for the linear model it was comparable to the low-fidelity conditions.This could have been caused by the UH-60's unsteady rate command system allowing motion feedback to provide the damping in the pilot's inner-loop control behavior.The comparison between the low-fidelity condition with and without compensation found significant differences for the following variables: ω c f , RMS ye , RMS v , RMS φ .Even though the interaction effect is not significant (p = 0.07), the compensation algorithm seems to mostly increase the cutoff frequency for the linear model leaving ω c f almost unchanged across the two conditions for the UH-60 model.A significant interaction effect was found for RMS ye and RMS v : when pilots controlled the linear model, the RMS of the lateral deviation and velocity were significantly lower for the condition with compensation.On the other hand, for conditions with the UH-60 model, RMS ye and RMS v were almost unaffected across the two conditions.This is likely due to the fact that the UH-60 had an unsteady rate command response as shown in Fig. 2, resulting in pilots remaining low-gain during the station keeping phase for both low-fidelity conditions.In addition, the UH-60 had a higher force-gradient, which reduced the magnitude of the force compensation relative to the linear model.Overall, the RMS of the force compensation from all pilots was 0.155 lb f for the linear model, and 0.107 lb f for the UH-60.Another interaction effect was found for RMS φ : the RMS of the roll angle was significantly lower for the condition with compensation and the linear model.On the contrary, RMS φ was almost unaffected between the two conditions for the UH-60.This interaction was most likely caused by the same factors as the interaction for RMSE ye and RMS v .It is apparent that the UH-60 model's response and the force gradient of the lateral stick affected pilots' approach to the task.In hindsight, a different approach to the test matrix, e.g., by using the linear model only, but varying the force gradient, might have provided more insight into the relation of pilot control behavior and task performance with motion force feedback.
+VII. ConclusionsTo investigate if control force compensation affects pilot control behavior and performance, pilots performed a two degrees-of-freedom lateral side-step task in the VMS under four different motion configurations (high-fidelity, medium-fidelity, low-fidelity, and low-fidelity with compensation) and with two simulated rotorcraft models (linear and UH-60 dynamics).By comparing pilots' control behavior and task performance between conditions, several conclusions could be drawn.The inertial control force compensation introduced significant differences in some of the dependentφ, φ, φ = Roll attitude, rate, and acceleration of simulated rotorcraft φ a y , φ a y , φ a y = Roll attitude, rate, and acceleration of the lateral stick due to inertia effect of lateral acceleration Φ a y = Lateral stick displacement compensation due to lateral acceleration Φ c = Combined lateral stick displacement compensation due to inertial force and torque φ pbd , φ pbd , φ pbd = Roll attitude, rate, and acceleration of the lateral stick due to inertia effect of roll angular acceleration Φ pbd = Lateral stick displacement compensation due to roll angular acceleration ω c f = Cut-off frequency of pilot's lateral stick input ω n = Natural frequency of the lateral stick ζ y = Damping ratio of the lateral stick a y = Lateral acceleration at the lateral stick's pivot point of the simulated rotorcraft in body frame g = Gravitational acceleration constant I xx = Moment of inertia about the pivot point of the lateral stick about the roll rotational axis k y = Force gradient of the lateral stick l c = Length of the lateral stick m c = Mass of the lateral stick p or p b = Roll angular rate of the simulated rotorcraft p b = Roll angular acceleration of the simulated rotorcraft r c = Distance between the pivot point of the stick and the c.g. of the lateral stick RMS = Root mean square SK = Station-keeping score sk = Station-keeping phase of the task t g = Time-to-target in the translation phase of the task trans = Translational phase of the task u = Lateral stick input v = Lateral velocity of the simulated rotorcraft v = Lateral acceleration of the simulated rotorcraft ye = Lateral position error
+( a )Fig. 1a1Fig. 1 Inertial forces acted on the stick due to simulated rotorcraft's dynamics.
+Fig. 2 Fig. 323Fig. 2 Roll rate responses from both models from a lateral stick doublet.
+Fig. 44Fig. 4 Verification of the control force compensation for the lateral stick displacements.
+Fig. 5 Fig. 656Fig. 5 Task performance criteria via a hover-target board in the out-the-window visual scene.
+Fig. 77Fig. 7 The Vertical Motion Simulator at NASA Ames Research Center.
+( a )Fig. 8 HQRa8Fig. 8 HQR rating given for the translation and station keeping phase of the task.
+Fig. 9 Station9Fig. 9 Station Keeping scores.
+Fig. 1010Fig. 10 Time to target and control input depended measures.
+RMS roll attitude.(d) RMS roll rate.
+Fig. 1111Fig. 11 Lateral and roll performance depended measures.
+Table 11Experimental conditions.Condition Rotorcraft ModelMotion Fidelity1high2 3linear modelmedium low4low + compensation5high6 7UH-60medium low8low + compensation
+Table 2 Mean of the dependent measures.2RMS ye RMS v RMS φ RMS pModel Fidelity ω c f Linear High t g SK RMS u 4.779 1.333 0.401 4.0721.5441.4013.1380.104UH-60 High4.476 1.2140.334 3.9991.1691.2404.6320.102Linear Medium4.369 1.4050.547 2.4501.6921.6894.6030.140UH-60 Medium4.392 1.2860.332 3.0721.3941.3755.1100.113Linear Low+Comp 4.729 1.2620.577 2.6141.4651.5635.0500.184UH-60 Low+Comp 4.462 1.2380.547 2.8091.4291.4896.1930.296Linear Low4.698 1.4290.591 2.1531.9701.9595.9810.195UH-60 Low4.502 1.2860.532 2.7581.4341.4766.1850.195
+Table 3 Standard deviation of the dependent measures.3RMS ye RMS v RMS φ RMS pModel Fidelity ω c f Linear High t g SK RMS u 0.736 0.473 0.104 0.6980.5870.4850.8550.027UH-60 High0.617 0.4110.131 0.6250.4150.3780.9080.042Linear Medium0.500 0.5380.479 0.6190.8500.6431.9400.064UH-60 Medium0.445 0.4530.150 0.7500.6530.5651.4780.053Linear Low+Comp 0.567 0.4920.403 0.5130.5920.8313.7600.148UH-60 Low+Comp 0.556 0.4800.492 0.6090.9040.7592.8610.619Linear Low0.984 0.5420.416 0.4351.0180.9644.0950.143UH-60 Low0.513 0.4530.488 0.6780.6860.8603.4950.181
+Table 4 Mean aggregate value for fidelity.4Fidelity t High 4.63 1.270.37 4.041.361.323.890.10Medium4.38 1.350.44 2.761.541.534.860.13Low+Comp 4.60 1.250.56 2.711.451.535.620.24Low4.60 1.360.56 2.461.701.726.080.19g SK RMS u ω c f RMS ye RMS v RMS φ RMS p
+Table 5 HQR statistical test results.5HQR tr ansHQR skEstimate Std.err Waldp Estimate Std.err WaldpHigh Fidelity vs. Low0.0480.135 0.124 0.725-0.0240.198 0.014 0.904Low Fidelity Without vs. With Comp0.2860.256 1.244 0.2650.2140.273 0.618 0.432
+Table 6 Summary of aggregated HQR means and standard errors.6HQR tr ansHQR skFidelityMean Std.err Mean Std.errHigh2.143 0.3402.571 0.571Low2.285 0.3592.857 0.459Low+Comp. 1.714 0.4202.428 0.369
+Table 7 Summary of statistical results for the station keeping score.7Estimate Std.err Waldp= significant (p ≤ 0.05)
+Table 8 Type III Analysis of Variance Table with Satterthwaite's method for the Multilevel Models8MeasureModelFidelityModel × FidelitydfFpdfFpdfFpt g1,6 1.9056 0.2167 3, 18 1.29210.3075 3, 18 1.4894 0.2511RMS u1,67.528 0.0336 3, 182.0150.148 3, 182.1640.128ω c f1, 66.318 0.0457 3, 1827.59 < 0.001 3, 185.0130.011RMS ye1,610.23 0.0186 3,182.2860.113 3,181.3520.289RMS v1, 611.90.014 3, 181.6360.216 3, 182.4150.100RMS φ1, 6 14.3650.009 3, 183.3880.041 3, 184.4890.016RMS p1,60.5930.471 3, 184.326 0.01836 3, 181.4690.256= significant (p ≤ 0.05)
+Table 9 Summary of the mixed effect model fitted for the time-to-target.9Estimate Std. Errordf t valuep
+Table 10 Summary of the mixed effect model fitted to the cut-off frequency.10Estimate Std. Errordf t valuep
+Table 11 Summary of the mixed effect model fitted to the RMS of the control input.11Estimate Std. Errordf t valuep
+Table 12 Summary of the mixed effect model fitted for the RMS of the lateral position error.12Estimate Std. Errordf t valuep
+		
+		
+
+			
+measures, mainly for the linear model.The control input cutoff frequency was higher, the station keeping score was better, and the RMS of the lateral error was lower with compensation when no motion was present.The RMS of the lateral velocity was marginally significantly lower.Therefore, control force compensation allowed for pilot control behavior and performance more similar to that under high-or medium-fidelity motion to some extent only.Considering all performance variables, we conclude that the control force compensation did not increase overall task performance considering both rotorcraft models at the same time.For the UH-60, the unsteady roll rate command response might have affected pilots' approach to the task and led to a low-gain control technique.In addition, the higher force gradient in the lateral stick for the UH-60 resulted in less inertial force compensation.As a result, the control force compensation only had a minimal effect on pilots' control behavior and task performance for the UH-60 model.This suggests that the control force compensation has limited benefits for controllers that have higher stiffness.Finally, high-fidelity and medium-fidelity motion did not always provide significant improvements in pilot performance compared to the low-fidelity conditions with and without compensation.
+			
+
+			
+RMS Lateral Velocity, RMS vThe error-bar plot of the RMS lateral velocity for each of the conditions can be seen in Fig. 11b.The summary of the linear mixed effect model is shown in Table 13.RMS v for the UH-60 (M = 1.39) was significantly lower than the RMS v of the linear model (M = 1.65); b = -0.258(SE = 0.075), t(6) = -3.450,p = 0.014.Furthermore, the RMS v was significantly lower for the high-fidelity condition compared to the rest of the conditions; b = -0.084(SE = 0.04), t(25.198)= -2.087,p = 0.047.There was a significant interaction between the model type and the low fidelity conditions; b = 0.205 (SE = 0.082), t(18) = 2.500, p = 0.022.The biggest effect of the force compensation algorithm is observed for the linear model.
+RMS Roll Attitude, RMS φThe error-bar plot of the RMS roll attitude for each of the conditions can be seen in Fig. 11c.The summary of the linear mixed effect model is provided in Table 14.There was a significant interaction between the model type in the contrast of the high-fidelity condition against the other conditions; b = 0.219 (SE = 0.0), t(20.53)= 2.738, p = 0.014.Furthermore the model significantly interacted between the low-fidelity condition with (for the linear model M = 5.05 and for the UH-60 M = 6.19)) and without compensation (for the linear model M = 5.98 and for the UH-60 M = 6.185); b = 0.470 (SE = 0.196), t(18) = 2.394, p = 0.028.
+VI. DiscussionThe purpose of this study was to investigate if control force compensation affects pilot control behavior and performance.Therefore, the main goal was comparing the low-fidelity condition and without compensation to assess the benefit, if any, of the force feedback algorithm.The high-and medium-fidelity conditions were used as baselines for the low-fidelity conditions.Due to an error in the implementation of the force compensation, the low-fidelity conditions were repeated by the seven pilots in a second experiment session several weeks after the first session.The low-fidelity conditions without compensation (conditions 3 and 7 in Table 1) were identical between the two sessions and were used to verify that pilots performed similarly between the two sessions and assured that session was not a significant confounding variable.Comparing all dependent measures for these conditions between the two sessions revealed that pilots performed the same overall, minimizing the chances of session having a significant effect on the results.
+A. HypothesesThe hypotheses provided in Section IV.H were accepted/rejected as follows: H1: The cutoff frequency ω c f was significantly higher in the low-fidelity condition with compensation compared to without compensation for both model types and closer to that in the medium-fidelity condition.However, this trend was not detected in the RMS of the control signal RMS u , which was similar for all low-fidelity conditions.The null hypothesis can not be fully rejected and it must be concluded that control behavior with force compensation is only more similar to that with high-or medium-fidelity motion in some respects.H2: The compensation did not lead to an improvement in all task performance measures.Only RMS ye were significantly lower with compensation.RMS v was marginally significantly lower.We fail to reject the null hypothesis and we conclude that the control force compensation did not increase overall task performance considering both rotorcraft models at the same time.H3: Taking into account all the RMS performance parameters in Table 2, there were little differences between the compensation and no compensation data for the UH-60.Results did not reveal a clear benefit of the force compensation with the UH-60 in low-fidelity conditions.No differences were noted in the lateral position error, lateral velocity, and roll attitude.Results showed that the RMS control input for the UH-60 was significantly lower than that of the linear model, indicating the unsteady roll rate response led to pilots staying in low-gain during the station-keeping phase.In addition, the higher force gradient for the UH-60 resulted in less control compensation relative to the linear model.This resulted in many significant interaction effects between the model and the force compensation, suggesting the null hypothesis can be rejected.H4: Only high-fidelity motion had a much higher control input cut-off frequency than all other test conditions.Pilots performing the task under high-fidelity motion also achieved a significantly lower roll attitude and lateral velocity, which suggested pilots were able to use the motion feedback to quickly damp out the roll attitude that commanded the lateral accelerations.The RMS of control input was significantly lower in the medium-fidelity condition than in both low-fidelity conditions.The null hypothesis can not be fully rejected based on these results, i.e., high-fidelity and medium-fidelity motion do not always provide larger improvements in pilot performance compared to both low-fidelity conditions.			
+			
+
+				
+
+
+	
+		The Determination of Some Requirements for a Helicopter Research Simulation Facility
+		
+			JBSinacori
+		
+		
+			Sep. 1977
+			Systems Technology, Inc
+		
+	
+	Tech. Rep. NASA CR-152066
+	Sinacori, J. B., "The Determination of Some Requirements for a Helicopter Research Simulation Facility," Tech. Rep. NASA CR-152066, Systems Technology, Inc., Sep. 1977.
+
+
+
+	
+		Visual and Motion Cueing in Helicopter Simulation
+		
+			RSBray
+		
+		NASA TM-86818
+		
+			Sep. 1985
+			Ames Research Center, Moffett Field (CA
+		
+	
+	Technical Memorandum
+	Bray, R. S., "Visual and Motion Cueing in Helicopter Simulation," Technical Memorandum NASA TM-86818, Ames Research Center, Moffett Field (CA), Sep. 1985.
+
+
+
+	
+		Spatial frequency and platform motion effects on helicopter altitude control
+		
+			JefferySchroeder
+		
+		
+			WilliamChung
+		
+		
+			RonaldHess
+		
+		10.2514/6.1999-4113
+		NASA/TP-1999-208766
+	
+	
+		Modeling and Simulation Technologies Conference and Exhibit
+		
+			American Institute of Aeronautics and Astronautics
+			Jul. 1999
+		
+	
+	Tech. Rep.
+	Schroeder, J. A., "Helicopter Flight Simulation Motion Platform Requirements," Tech. Rep. NASA/TP-1999-208766, NASA, Jul. 1999.
+
+
+
+	
+		Experiments and a Model for Pilot Dynamics with Visual and Motion Inputs
+		
+			RLStapleford
+		
+		
+			RAPeters
+		
+		
+			Alex
+		
+		
+			FR
+		
+		NASA CR-1325
+	
+	
+		NASA
+		
+			1969
+		
+	
+	Tech. Rep.
+	Stapleford, R. L., Peters, R. A., and Alex, F. R., "Experiments and a Model for Pilot Dynamics with Visual and Motion Inputs," Tech. Rep. NASA CR-1325, NASA, 1969.
+
+
+
+	
+		Roll Tracking Effects of G-vector Tilt and Various Types of Motion Washout
+		
+			HRJex
+		
+		
+			REMagdaleno
+		
+	
+	
+		Fourteenth Annual Conference on Manual Control
+		University of Southern California, Los Angeles (CA
+		
+			1978
+			
+		
+	
+	Jex, H. R., and Magdaleno, R. E., "Roll Tracking Effects of G-vector Tilt and Various Types of Motion Washout," Fourteenth Annual Conference on Manual Control, University of Southern California, Los Angeles (CA), 1978, pp. 463-502.
+
+
+
+	
+		Effects of Stick Dynamics on Helicopter Flying Qualities
+		
+			DCWatson
+		
+		
+			JASchroeder
+		
+	
+	
+		AIAA Guidance, Navigation and Control Conference
+		
+			1990
+		
+	
+	Watson, D. C., and Schroeder, J. A., "Effects of Stick Dynamics on Helicopter Flying Qualities," AIAA Guidance, Navigation and Control Conference, 1990.
+
+
+
+	
+		Feel Systems and Flying Qualities
+		
+			DavidMitchell
+		
+		
+			BimalAponso
+		
+		
+			DavidKlyde
+		
+		10.2514/6.1995-3425
+	
+	
+		20th Atmospheric Flight Mechanics Conference
+		
+			American Institute of Aeronautics and Astronautics
+			1995
+		
+	
+	Mitchell, D. G., Aponso, B. L., and Klyde, D. H., "Feel Systems and Flying Qualities," AIAA Atmospheric Flight Mechanics Conference, 1995.
+
+
+
+	
+		
+			JALusardi
+		
+		
+			CLBlanken
+		
+		
+			CROtt
+		
+		
+			CAMalpica
+		
+		
+			WVon Grünhagen
+		
+		Flight Evaluation of Active Inceptor Force-Feel Characteristics and Handling Qualities
+		
+			American Helicopter Society 68 th Annual Forum
+			2012
+		
+	
+	Lusardi, J. A., Blanken, C. L., Ott, C. R., Malpica, C. A., and von Grünhagen, W., "In Flight Evaluation of Active Inceptor Force-Feel Characteristics and Handling Qualities," American Helicopter Society 68 th Annual Forum, 2012.
+
+
+
+	
+		Validation of a Real-Time Engineering Simulation of the UH-60A Helicopter
+		
+			MGBallin
+		
+		
+			Feb. 1987
+			Ames Research Center, Moffett Field (CA
+		
+	
+	Technical Memorandum NASA TM-88360
+	Ballin, M. G., "Validation of a Real-Time Engineering Simulation of the UH-60A Helicopter," Technical Memorandum NASA TM-88360, Ames Research Center, Moffett Field (CA), Feb. 1987.
+
+
+
+	
+		The Use of Pilot Rating in the Evaluation of Aircraft Handling Qualities
+		
+			GECooper
+		
+		
+			RPHarper
+		
+		
+			Jr
+		
+		NASA TN D-5153
+	
+	
+		NASA Technical Note
+		
+			1969
+		
+		
+			NASA
+		
+	
+	Cooper, G. E., and Harper, R. P., Jr., "The Use of Pilot Rating in the Evaluation of Aircraft Handling Qualities," NASA Technical Note NASA TN D-5153, NASA, 1969.
+
+
+
+	
+		Investigation of the Effects of Bandwidth and Time Delay on Helicopter Roll-Axis Handling Qualities
+		
+			H.-JPausder
+		
+		
+			CLBlanken
+		
+		
+			1992
+		
+	
+	18 th European Rotorcraft Forum
+	Pausder, H.-J., and Blanken, C. L., "Investigation of the Effects of Bandwidth and Time Delay on Helicopter Roll-Axis Handling Qualities," 18 th European Rotorcraft Forum, 1992.
+
+
+
+	
+		A DISTRIBUTION-FREE <i>k</i>-SAMPLE TEST AGAINST ORDERED ALTERNATIVES
+		
+			ARJonckheere
+		
+		10.1093/biomet/41.1-2.133
+		
+	
+	
+		Biometrika
+		Biometrika
+		0006-3444
+		1464-3510
+		
+			41
+			1-2
+			
+			1954
+			Oxford University Press (OUP)
+		
+	
+	Jonckheere, A. R., "A Distribution-Free k-Sample Test Against Ordered Alternatives," Biometrika, Vol. 41, No. 1/2, 1954, pp. 133-145. URL http://www.jstor.org/stable/2333011.
+
+
+
+	
+		Random-Effects Models for Longitudinal Data
+		
+			NanMLaird
+		
+		
+			JamesHWare
+		
+		10.2307/2529876
+		
+	
+	
+		Biometrics
+		Biometrics
+		0006-341X
+		
+			38
+			4
+			963
+			1982
+			JSTOR
+		
+	
+	Laird, N. M., and Ware, J. H., "Random-Effects Models for Longitudinal Data," Biometrics, Vol. 38, No. 4, 1982, pp. 963-974. URL http://www.jstor.org/stable/2529876.
+
+
+
+	
+		
+		
+			AField
+		
+		
+			JMiles
+		
+		
+			ZField
+		
+	
+	
+		Discovering Statistics Using R
+		
+			2012
+			Sage
+		
+	
+	Field, A., Miles, J., and Field, Z., Discovering Statistics Using R, Sage, 2012.
+
+
+				
+			
+		
+	

file149.txt

@@ -0,0 +1,644 @@
+
+	
+	
+		
+I. Introduction
+Aflights to runways at an airport is a critical function that influences all aspects of airport operations and performance.These assignments, made by air traffic controllers, indicate the runway on which a flight must land or take off from.In this paper, we describe our research developing data-driven machine learning (ML) models to predict runway assignments for arriving and departing flights for various airports in the U.S.While the precise roles and responsibilities vary for arriving and departing flights, air traffic controllers in general have tremendous expertise in making these runway assignments.They apply a heuristic based on a multitude of factors including at least direction of flight, airport configuration (i.e., which runways are available for arrivals and departures), configuration of nearby airports (e.g., in the same metroplex region), aircraft size and engine type, noise and other environmental regulations, and traffic volumes.They also consider requests from flight crews for specific runways due to preference or operational necessity.Furthermore, the rules applying to each of these factors may evolve over time as new procedures or routes are developed.This heuristic is clearly complex and learned by controllers over years of site-specific training and experience.Some prior research such as Isaacson et al. [1] has attempted to use observational studies on Subject Matter Experts (SMEs) to capture the decision heuristics that controllers employ to assign runways.Likewise, the development of a similar knowledge base has taken place on the ATD-2 project [2] (of which this research is part), and against which the results of the ML-based approach described here will be compared.The ATD-2 expert-driven approach consists of decision trees.For example, the decision tree may have a path such that all flights with a jet engine, headed for the BLECO fix, while the airport is operating in "South Normal" configuration, will be assigned to depart from runway 18L.The challenge with expert-driven approaches like these is that they are expensive and difficult to construct, given the requirement for access to personnel with specialized and timely knowledge, and the highly-nuanced differences in the heuristics that each individual might apply.Our approach in this research is to leverage the vast amounts of data on airspace and airport operations, in conjunction with modern ML techniques, to predict these runway assignments with performance nearly equivalent to that which is achievable through the expert-driven approach.Whereas research has been conducted to apply machine learning to various prediction problems relating to aircraft and airport operations (e.g., predicting airport configurations, landing times, taxi out times), relatively little attention has been paid to predicting runway assignments.However, two recent works do cover the problem in an international context.In [3], the author developed a predictive model for runway assignments using support vector machines (SVM) for Amsterdam Schiphol Airport.Likewise, in [4], the authors developed a neural network based ML approach to predict the runway to be used by arrivals into Tokyo International Airport.Our work builds on this prior research by employing a more generic framework that can be easily applied to multiple airports, by focusing on deployment to a real-time system, and by exploring different ML approaches and metrics to understand model performance.Rather than focusing on the problem of predicting runway assignments, considerable previous research has addressed on the problem of optimizing runway assignments (and sequence of runway use) to achieve various objectives.For example, Berge et al [5] jointly optimize the runway assignment and sequencing decisions.Likewise, see Lohr et al [6] for work that jointly optimizes airport configurations and runway assignments.An example with a longer time scale is [7], in which the authors planned a longer-term runway assignment strategy to achieve various delay, environmental, and safety related objectives.This distinction between prescription (i.e., optimization) and prediction is critical in justifying the approach that we have developed.Although previous research on optimizing runway assignments and utilization has demonstrated potential benefits (e.g., greater runway throughput, shorter taxi distances), these approaches in general make significant assumptions about the ease with which such changes could be implemented.Our approach assumes that air traffic controllers will continue to use their considerable expertise and available automation tools, and we seek to capture (through machine learning) the runway assignment heuristics that exist within this structure.This ML model will enable us to make detailed forecasts of runway utilization several hours in advance of the actual operation.This distinction is essential to understand the broader context of the research as part of an ML-powered shadow system to help evaluate the its ability to match the performance of the the expert-driven approaches currently used in the ATD-2 system.In the remainder of this paper, we describe our modeling approach, the data used for our study, results and discussion, and finally conclusions and directions for continued investigation.
+II. Modeling ApproachIn this section, we describe the approach developed to train ML models that imitate the decision heuristics used by air traffic controllers outlined in the previous section.The modeling approach developed here encompasses several elements, each of which is described in greater detail in the following subsections:• Requirements: Several requirements drove our decision-making process in developing these ML models.• Target: The target value is the runway identifier actually used by the flight in the historical data.• Features: Based on literature review, brainstorming, and consultation with available SMEs, identify, explore, and compute relevant input features.This section also describes imputation and encoding strategies used for each feature.• Building the Dataset: Several steps are described in this section that are used to translate from a list of features to a rectangular dataset that can be used by one of the machine learning algorithms we have evaluated.• Machine learning algorithms: This section describes applying machine learning algorithms to the full prepared training dataset to create models.
+A. Modeling RequirementsImportant requirements informed the development of these models.First, it was of crucial importance that the models developed as part of this research be suitable for deployment in a system processing live data.Thus, they must use features that can be readily computed at runtime.Further, they must have a full suite of imputers to handle inevitably missing data, or alternatively, must have a well-defined filtering approach to ensure that non-imputed features are never passed to the model as nulls.Finally, their query time must be reasonable to support processing large batches of flights at regular intervals.This real-time support was necessary because these runway prediction models are part of a suite of ML models to be assembled into a shadow system to evaluate against the performance of the legacy ATD-2 systems.Second, an additional objective in training these models is that the process used is highly data-driven, in that it does not require the maintenance of significant adaptation data (i.e., site-specific geometric or procedural information).Achieving this objective allows the models to be generated and updated rapidly to cover dozens of airports across the U.S., provided that standard data formats are employed.Finally, the models needed to be trained (and served in real-time) using data from the ATD-2 Fuser [8], a data processing and fusion system that is designed to handle many different data feeds and formats simultaneously, generating a standard relational format.
+B. TargetThe target value for these models is the actual runway on which each flight operated.For this effort, we derived these values from surveillance data (Airport Surface Detection Equipment, Model X [ASDE-X] or Traffic Flow Management System [TFMS]) and runway adaptation from the National Flight Data Center (NFDC) data [9] to ensure consistency and reliability of these crucial data elements.When surveillance located a flight within a runway polygon and with physics consistent with aircraft landing, it was straightforward to determine that the flight operated on that runway at a certain time.When this is not the case, which is relatively rare at many large airports, we used the airborne surveillance data from TFMS to infer which runway was used and at what time.This was accomplished with an approach developed by Robert Kille [10], under funding by NASA.
+C. FeaturesThe set of features used in training the arrival and departure runway assignment models to date is relatively simple, with considerable overlap between the two models.Table 1 summarizes which features are included in each model.Descriptions of each feature follow the table.As with any ML modeling workflow, this is an area of iterative exploration, and we believe that additional features may improve the predictive power of each model.
+Aircraft Engine ClassThis categorical value (handled via one-hot encoding) indicates whether a flight has a jet, turboprop, or propeller engine.Discussions with SMEs and insight from previous ATD-2 work indicated that this was an important discriminator in runway assignments.For example, some runways are reserved exclusively for propeller-driven aircraft, due to the significant differences in their takeoff and landing performance.Missing values for this feature are imputed, by using the most frequently observed value in the training dataset.For most training datasets, this will be jet engine, as jets comprise the vast majority of operations at airports for which this model is relevant.
+Wake Turbulence CategoryThis categorical value (handled via one-hot encoding) indicates the impact of the wake vortex induced by the flight.This roughly correlates with the size of the aircraft, and implies the separation required between operations on the same runway.As with engine class, the wake vortex category was identified during discussions with SMEs and in previous ATD-2 work.For example, at some airports, the largest aircraft may be restricted to using certain runways due to available length for takeoffs or geometric constraints.For all analysis, FAA RECAT wake vortex categories [11] are used.Missing values for this feature are imputed, by using the most frequently observed value in the training dataset to replace missing values.Typically there is only one dominant weight class, making this imputation insignificant.
+Planned FixAs part of filing a flight plan or through communication with air traffic controllers, a flight operator provides some indication about their path through the terminal area and which fix they plan to use to transition into / out of the terminal area.The fix name is generally available through the data feeds used to build training datasets, and is handled in the model through one-hot encoding.Observations with missing values are not used in training models.According to discussions with SMEs, for both arrivals and departures, this planned fix, in conjunction with the current airport configuration, provides significant information about which runway will be used.
+Flight Plan FiledThis boolean indicates whether a flight plan has been filed by the flight operator.The flight plan provides the planned fix, as described in the previous section.However, before the flight plan is actually filed, the FAA automation systems will provide a 'default' fix value for most flights.Missing values are replaced with a false value.By including this indicator, the model is able to differentiate between these filed and default values, and learn the circumstances under which each provides valuable information.
+Airport ConfigurationAirport configuration lists which runways are being used for arrivals and for departures at a specific instant in time.Runways may be in both lists; in this case, they are known as dual-use runways.The distinct combinations of these lists form the set of configurations available for the runway assignment models.Neither in reality, nor in the runway assignment models, do the airport configurations totally constrain which runways may be assigned, however they strongly influence this.This value could be provided by either a data feed that provides live updates (e.g., Digital Automatic Terminal Information Service (D-ATIS)), or a predictive model for airport configurations (e.g., [12]).For model training, the D-ATIS data indicating the actual configuration at the time of the operation was used.This ensures that the model is not biased by errors from another model.In future work, this assumption should be revisited.Regardless of the source, a custom encoder (in place of one-hotting) is used to translate the airport configuration to values usable by the model.The custom encoder, for an airport with runways, creates 2 columns, one for each combination of either 'arrival' or 'departure' and runway name.We hypothesize that this encoding strategy should help the model learn from similar configurations (e.g., add / remove a single runway) in a way that considering the name alone would not.Missing values are extremely unlikely for this data source, as we assume that the previous configuration continues until a new one is explicitly provided by the data feed.However, should a missing value be encountered (e.g., at the beginning of a time period), those observations are not used to train models.
+Time since Airport Configuration ChangedAs there may also be a relationship between runway assignments and the amount of time that the airport has been operating in the current airport configuration, we include this duration in seconds as a feature in the model.The logic for this feature is that the relationship between configuration and runway assignment policies may be more flexible when a configuration is newly in place, e.g., as a result of aircraft already lined up for specific runways.
+Time until Estimated OperationAnother feature that may influence the runway assignment policy is the time expected until a flight operates on the runway.For example, there may be less certainty about the policy to be applied when this lookahead is very long.This lookahead is computed as the difference between the time at which the prediction is being made, and an estimated operation time.For departures, this estimated operation time is the Earliest Off Block Time (EOBT) value provided by the airline (if unavailable, other estimates from FAA systems).For arrivals, this value is one of several landing time estimates provided by FAA systems.Which of these estimates to use at each instant is a research question unto itself, and an approach for selecting this 'best' landing time is described in [13].
+TBFM-assigned RunwayOne additional feature included in the arrival runway model is the runway assignment generated by the FAA's Time Based Flow Management (TBFM) system [14].One of the many features of this decision support system is the capability to predict runway assignments for arriving flights.However, the accuracy of these predictions varies significantly between TBFM systems used in different terminal areas.Under the proper conditions, this TBFM-assigned runway may be sufficiently accurate to match the performance of an ML model, as it is the product of an expensive and lengthy process of codifying the controllers' runway assignment heuristics into adaptation data.In other situations, this value may simply indicate a flow direction for the airport, without a specific runway identifier.In any case, the value of training runway assignment models for each airport individually allows the ML model to learn the value of this data as a feature and use it accordingly.These data are provided as categories, encoded in the model as one-hot features.Missing values for this feature are imputed by filling with a constant value of UNKN, creating a new category.
+D. Building the DatasetSeveral steps are required to build a rectangular dataset that an ML algorithm can use to train a model.These steps are primarily mechanical, but are described in the interest of promoting reproducible research.It is critical to recognize that the various features listed in the previous section are available at different instants in time, and are updated at different rates.Through the use of the ATD-2 Fuser, described in the requirements section, the state of each flight is readily available by carrying forward values from previous messages.However, even with this approach, data are only available at the instants at which messages were received from various automation systems.To create a more uniform dataset from which to sample (e.g., to avoid bias induced by certain kinds of flights producing more messages), this approach of carrying values forward was extended further by creating a uniformly-spaced sequence of lookahead (i.e., time until estimated operation) values against which the raw dataset was joined.An example is shown by converting the data in Table 2 to that in Table 3.This is clearly a large dataset.Assuming a medium-sized airport with 500 flights/day, a four-hour lookahead period with a one-minute update rate, and six months of training data, there are 21.6 million rows of data available.For some algorithms and computational setups, this volume may introduce difficulties, so lower sampling rates may be necessary than are traditionally employed.Once a 'uniform' dataset of this nature has been constructed, some rows may be filtered out for having unsuitable data.First, we check that the target values (i.e., runway names) for each flight fall into the set of known runway identifiers for the airport being studied.After exploratory data analysis, we identified several features that seemed unwise to impute.As a result, these features were identified as core for training a model and making predictions.Thus, any observation with a missing value for a core feature was not used in model training.Because these rules are logged with the trained model itself, they are applied during real-time operations.Any flight in the real-time environment that fails any of these rules (which would not have been used to train the model in the first place) is assigned a default runway.Based on our analysis and discussion with SMEs, the following features are considered core, and so any rows with a missing value are discarded:• Planned fix: could be missing if FAA automation systems malfunction, or data is lost• Airport configuration: could be missing if FAA automation systems malfunction, or data is lost • Time since airport configuration changed: follows mechanically with airport configuration • Time until estimated operation: could be missing if all relevant FAA systems are not providing valid data
+E. Machine Learning AlgorithmsSeveral ML algorithms have been evaluated thus far in training models for the runway assignment problem.As formulated, this problem is a multinomial classification problem, for which many algorithms currently exist.As will be highlighted in the results, we have thus far used the classic logistic regression [15] available through scikit-learn [16], and the more recently developed and very popular XGBoost [17].Cross-validation and hyper-parameter tuning for each approach has been evaluated, but results are not included in this paper.Initial analysis of those processes indicated limited improvement in performance metrics as compared to the default parameters available in their implementations.
+III. Results and DiscussionUsing the features and algorithms described in previous sections, with the Kedro framework [18], we have developed a modeling process that can be easily replicated for any airport for which data are available in the suitable format.A similar series of pipelines for data query, data engineering, and model training / evaluation have been developed for both the arrival and departure runway prediction problems.There is considerable overlap, and code re-use, between the two pipelines.In this section, we first present an overview of the data used in this study, and then a description of various performance metrics of the models trained for various U.S. airports.
+A. DataAppropriate data are essential to conducting research using ML.For this work, we leveraged the Fuser technology previously developed for the ATD-2 project.To generate the datasets used in this research, the Fuser was configured to consume data from the following FAA data feeds: TFM Flight [19], TBFM [14], and STDDS ASDE-X / SMES [20].The details of each of these data feeds is beyond the scope of this paper, but they provide comprehensive detail about the planned and actual operation of each flight in the U.S. from gate to gate.The results presented here are based on models trained and evaluated on data from April 25, 2020 through December 31, 2020.Because of the large size of this dataset, as illustrated in the previous section, just 5% of the dataset is used for training, and 5% for evaluation.These values are approximate because the 5% value for training actually represents the fraction of individual flights, rather than rows.If we were to sample only rows without considering the panel nature of the data, then rows for the same flight might end up in both the training and evaluation samples.After removing the training sample, exactly 5% of the remaining rows are selected as the evaluation dataset.Note that both datasets contain a variety of lookahead values, and both are sampled randomly.It is essential to acknowledge the role of the pandemic in the data used for this research.Flight operations were reduced by an enormous amount at the start of the pandemic, and this is reflected in the data.Were these models simply trained for the purposes of writing this paper, then this drastic change in the data would present few problems, as we would simply use pre-pandemic (e.g., 2018-2019) data to represent normal levels of air traffic.However, this step change in the data creates challenging problems deploying a model trained on pre-pandemic data, and achieving comparable performance using live data.Deploying trained ML models to operate as part of a shadow system for the ATD-2 project was a key requirement of this research, as described earlier.As a result, many of the results presented in this section rely on models trained using data from during the pandemic.This was done to maximize the likelihood that when the models were deployed (still during reduced traffic levels) the conditions used in training the models would be similar to those present.
+B. Arrival Runway Assignment ModelsArrival runway assignment prediction models have been trained for a variety of airports.An initial summary of model accuracy is shown in Table 4.These results reflect the performance of the model trained using XGBoost on the evaluation sample, using the features listed in previous sections.Note that these accuracy metrics include observations that are sampled from a variety of lookahead values.This initial summary of model performance demonstrates that the models are performing at a reasonable accuracy level, given the relative complexity of each airport (e.g., KDFW has more runways used for arrivals than KEWR, so we should expect more uncertainty).However, these results should also be evaluated in a variety of other dimensions, as outlined below.
+Performance of Different AlgorithmsModels were trained using Logistic Regression for some airports.The performance of these models is compared against the XGBoost models in Table 5.The same training and evaluation samples were used for each algorithm.From these results, it is clear that the XGBoost models perform significantly better than those trained using Logistic Regression.Although there may be interesting explanations and insight related to the relative performance of these algorithms, to help achieve the objectives of this research, we are satisfied to identify the superior performance of the XGBoost models and use them going forward.
+Comparison to Expert-Driven approachAs described in the introduction, the legacy ATD-2 system used at several airports has decision trees for predicting runway assignments developed through data analysis and interviews with SMEs.In Table 6, we compare the accuracy of the XGBoost models with that of the expert-driven models.In this comparison, the accuracy metrics for the XGBoost models presented earlier are repeated, but the metrics for the legacy ATD-2 systems are sampled at the fix crossing event, where the accuracy should be highest.There are two interesting trends in this comparison between the ML model results and the expert-driven models.First, the performance of the expert-driven model at KCLT exceeds that of the ML model, reflecting how well-tuned the legacy ATD-2 system is at that facility.In contrast, for the Dallas-area airports, the ML model is able to achieve superior performance.The important difference between KCLT and KDFW is that, during the pandemic, KCLT mostly continued operating in the same fashion (albeit with reduced traffic) while at KDFW, several operational changes were implemented (e.g., arrival runway closed for maintenance).The decision trees in the legacy system were not updated to reflect these operational changes.This highlights the advantage of using an ML approach that is early to update relative to SME informed decision trees.
+Confusion MatricesThe ways in which each model might make incorrect predictions is also of interest.One way to evaluate these incorrect predictions is through the use of a confusion matrix.In Figure 1, the relatively simply confusion matrix for KDAL is shown.Rather than show counts (which would be large numbers) the fraction of the evaluation dataset in each cell is shown as a percentage.Warmer colors (e.g., yellow) indicate a larger fraction of the dataset, while cooler colors (e.g., blue) indicate a smaller fraction.It is clear that the bulk of the observations fall on the diagonal, in which the model correctly predicts the arrival runway.However, an interesting effect for KDAL is that there is a significant amount of incorrect predictions on a parallel runway (e.g., predict 13L, land 13R).In some sense, these incorrect predictions are not as bad as those for which the "flow" direction of the airport is incorrect (e.g., 13 vs 31).
+Fig. 1 KDAL Confusion MatrixIn Figure 2, the confusion matrix for KCLT is shown.This matrix exhibits a sort of block diagonal structure, with blocks for the two directions of the primary group of runways (i.e., 18 and 36).In line with the high accuracy for KCLT, the cells with the highest fraction of observations are along the true diagonal.
+Fig. 2 KCLT Confusion MatrixWe have generalized this important notion about an incorrect prediction to a parallel runway being less bad than an incorrect prediction to a non-parallel runway.Table 7 shows the fraction of rows from the evaluation dataset for which the prediction was incorrect, but for which the numeric portion of the runway identifier (i.e., the direction) matched the true runway used.Only airports with some parallel runways are included in these data.From these data, it is clear that some airports have a more flexible utilization strategy for parallel runways than other airports (e.g., KDFW vs KEWR).In future work, we hope to improve these incorrect predictions to parallel runways by identifying features that may indicate such a balancing strategy is in use, and leverage the "less bad" nature of this incorrect prediction in the model training itself.
+Evolution of Predictions over TimeIn previous analysis sections, results have been evaluated together across all lookahead times.In some ways, this is quite a fair evaluation strategy, because much of the data used in making these predictions is static (excepting the TBFM-assigned runway).However, it is also important to acknowledge and evaluate the dynamic nature of these models.To that end, Figures 3 and4 depict the accuracy of models for KCLT and KJFK over time, as flights approach and land at the airport.These two airports make an interesting contrast, as the accuracy for KJFK is relatively constant, while the accuracy for KCLT is steadily increasing.This likely reflects differences in the flexibility of each airport to make changes to arrival runway assignment planning as flights progress.8.These results reflect the performance of the model trained using XGBoost on the evaluation sample, using the features listed in previous sections.Note that these accuracy metrics include observations that are sampled from a variety of lookahead values.The accuracy levels are slightly higher for these departure runway models than for the arrival runway models, indicating that the decision heuristic employed by the controllers is better able to be captured (e.g., is more consistent) than for the arrival runway prediction problem.However, these results should also be evaluated in a variety of other dimensions, as outlined below.
+Performance of Different AlgorithmsModels were trained using Logistic Regression for some airports.The performance of these models is compared against the XGBoost models in Table 9.The same training and evaluation samples were used for each algorithm.From these results, it is clear that the XGBoost models perform significantly better than those trained using Logistic Regression.For the purposes of this research, we are not concerned about the root cause of this differential, just the trend that XGBoost seems to produce better-performing models.Some preliminary results indicated that there was potential for tuning the hyperparameters of the Logistic Regression models to achieve better performance, but still not at
+Comparison to Expert-Driven approachAs described in the introduction, the legacy ATD-2 system used at several airports has decision trees for predicting runway assignments developed through data analysis and interviews with SMEs.In Table 10, we compare the accuracy of the XGBoost models with that of the expert-driven models.In this comparison, the accuracy metrics for the XGBoost models presented earlier are repeated, but the metrics for the legacy ATD-2 systems are sampled at the pushback event, where the accuracy should be highest.There are two interesting trends in this comparison between the ML model results and the expert-driven models.First, the performance of the expert-driven model at KCLT exceeds that of the ML model, reflecting how well-tuned the legacy ATD-2 system is at that facility.In contrast, for the Dallas-area airports, the ML model is able to achieve superior performance.The important difference between KCLT and KDFW is that, during the pandemic, KCLT mostly continued operating in the same fashion (albeit with reduced traffic) while at KDFW, several operational changes were implemented.These operational changes did not result in updates to the decision trees used in the legacy system.
+Confusion MatricesIn Figure 5, the relatively simply confusion matrix for KLGA is shown, and in Figure 6, the confusion matrix for KDFW is shown.
+Fig. 5 KLGA Confusion MatrixAgain the fraction of the evaluation dataset in each cell is shown as a percentage.It is clear that the bulk of the observations fall on the diagonal for each airport, for which the model correctly predicting the departure runway.From this confusion matrix, is is also clear that the model does face some confusion about the use of the diagonal runways at KDFW (i.e., the 13 and 31 runways).There are also several off-diagonal blocks corresponding to incorrect predictions at KDFW when these runways were (or were not) expected to be used.This is clearly an area where some additional features may improve model performance, since there is clearly some strategy in the controllers' runway assignment decisions to use those diagonal runways (e.g., GA versus airline flights, parking stand location).We have generalized this important notion about an incorrect prediction to a parallel runway being less bad than an incorrect prediction to a non-parallel runway.Table 11 shows the fraction of rows from the evaluation dataset for which the prediction was incorrect, but for which the numeric portion of the runway identifier (i.e., the direction) matched the true runway used.Only airports with some parallel runways are included in these data.From these data, two trends are clear.First, the Dallas-area airports continue to exhibit greater flexibility in runway utilization, as was observed for arrivals.Second, and perhaps more importantly, the value of this metric across all airports is lower than it was for arrivals.This suggests that departure runway assignments are more predictable (as reflected in the higher accuracy metrics) than arrival runway assignments.This is an important finding, because the primary focus of the experiments being conducted in the ATD-2 project is planning runway utilization for departures.
+Evolution of Predictions over TimeIn previous analysis sections, results have been evaluated together across all lookahead times.In some ways, this is quite a fair evaluation strategy, because much of the data used in making these predictions is static (excepting the TBFM-assigned runway).However, it is also important to acknowledge and evaluate the dynamic nature of these models.To that end, Figures 7 and8 depict the accuracy of models for KDAL and KIAH over time, as flights approach and land at the airport.These two airports are simply examples of consistent behavior we observe across airports: there are
+Use of Different Time PeriodsFinally, we present results demonstrating that our use of data from 2020 (during the pandemic) yielded models of similar quality (in aggregate) to models trained on data from earlier periods.To make this comparison, we trained a model on the same time period (April 25 through December 31) from 2019 for KDFW.This airport was selected for comparison because the degradation of ATD-2 decision tree model accuracy apparent in Tables 6 and10 suggested operational changes.The data in Table 12 compares the performance of the two models trained on different time periods, and evaluated using samples taken from their "own" year of data.Additional classification metrics besides accuracy (as shown previously) are presented, including precision, recall, and area under ROC curve (AUC).The results from this comparison suggest that the XGBoost algorithm is able to produce a model of equivalent quality using either data from a "normal" time period, or from the pandemic time period.Or, put another way, the problem of assigning flights to departure runways was equally predictable with the same features during each distinct period, even if those relationships may have changed somewhat.
+IV. Conclusion and Ongoing WorkIn this paper, we have described our work training ML models to predict arrival and departure runway assignments.This work shows initial promise for learning the heuristics used by controllers to assign flights to runways when landing or departing.Models have relatively high accuracy, likely high enough to support the use cases for which they are being evaluated on the ATD-2 project.The overall approach will enable broad deployment across a wide variety of U.S. airports using a standardized approach and dataset.In concert with models to predict other aspects of NAS operations, we believe this data-driven machine learning approach will enable rapid testing and deployment of advanced prediction and decision-support tools.Fig. 3 KJFK3Fig. 3 KJFK Model Accuracy over Time Fig. 4 KCLT Model Accuracy over Time
+Fig. 66Fig. 6 KDFW Confusion Matrix
+Fig. 7 KDAL7Fig. 7 KDAL Model Accuracy over Time
+
+
+
+Table 1 List of modeling features Feature Arrivals Departures1Aircraft engine classxxWake turbulence categoryxxPlanned fixxxFlight plan filedxxAirport configurationxxTime since airport configuration changedxxTime until estimated operationxxTBFM-assigned runwayx
+Table 2 Sample Data from ATD-2 Fuser2FlightLookahead (mins) Other FeaturesABC12374.3[set1]ABC12355.1[set2]ABC12331.9[set3]ABC12312.8[set4]
+Table 3 Cleaned, Carried-forward Sample Data3FlightLookahead (mins) Other FeaturesABC12374[set1]ABC12373[set1]ABC123...[set1]ABC12355[set2]ABC12354[set2]ABC123...[set2]ABC12331[set3]ABC12330[set3]ABC123...[set3]ABC12312[set4]ABC12311[set4]ABC123...[set4]ABC1230[set4]
+Table 4 Arrival Runway Accuracy Metrics for Various Airports Airport XGBoost Classifier Accuracy4KDFW0.618KDAL0.726KCLT0.777KIAH0.699KJFK0.765KEWR0.939KLGA0.973KPHL0.791KBOS0.915
+Table 5 Comparison of Different Algorithms for Arrival Runway Prediction Airport XGBoost Classifier Accuracy Logistic Regression Accuracy5KDFW0.6180.375KDAL0.7260.495KCLT0.7770.266KIAH0.6990.250KJFK0.7650.527KEWR0.9390.567KPHL0.7910.431
+Table 6 Comparison of ML and Expert-Driven Approaches for Arrival Runway Prediction Airport XGBoost Classifier Accuracy ATD-2 Decision Tree Accuracy6KDFW0.6180.524KDAL0.7260.578KCLT0.7770.911
+Table 7 Incorrect Prediction to Parallel Runways Airport Fraction Observations Incorrect, but on Parallel Runway7KDFW0.266KDAL0.227KCLT0.194KIAH0.229KJFK0.167KEWR0.041KPHL0.095KBOS0.012
+Table 8 Departure Runway Accuracy Metrics for Various Airports Airport XGBoost Classifier Accuracy8KDFW0.821KDAL0.813KCLT0.886KIAH0.797KJFK0.932KEWR0.971KLGA0.977KPHL0.902KBOS0.894
+Table 9 Comparison of Different Algorithms for Departure Runway Prediction Airport XGBoost Classifier Accuracy Logistic Regression Accuracy9KDFW0.8210.548KDAL0.8130.839KCLT0.8860.314KIAH0.7970.657KJFK0.9320.497KEWR0.9710.583a level comparable with even the stock implementation of the XGBoost classifier.
+Table 10 Comparison of ML and Expert-Driven Approaches for Departure Runway Prediction Airport XGBoost Classifier Accuracy ATD-2 Decision Tree Accuracy10KDFW0.8210.828KDAL0.8130.654KCLT0.8860.950
+Table 11 Incorrect Prediction to Parallel Runways Airport Fraction Observations Incorrect, but on Parallel Runway11in accuracy leading up to the pushback event, because there are relatively few changes in the features input to the model leading up to pushback.KDFW0.121KDAL0.170KCLT0.081KIAH0.034KJFK0.017KEWR0.014KPHL0.071KBOS0.018relatively few changes
+Table 12 Comparison of Departure Runway Models from 2019 and 2020 Metric 2020 Model 2019 Model12Accuracy0.8210.851Misclassification to parallel runway0.1210.111Precision0.8240.841Recall0.8210.851AUC0.9210.913
+		
+		
+			
+
+				
+
+
+	
+		Knowledge-based runway assignment for arrival aircraft in the terminal area
+		
+			DouglasIsaacson
+		
+		
+			ThomasDavis
+		
+		
+			JohnRobinson, Iii
+		
+		
+			DouglasIsaacson
+		
+		
+			ThomasDavis
+		
+		
+			JohnRobinson, Iii
+		
+		10.2514/6.1997-3543
+		
+	
+	
+		Guidance, Navigation, and Control Conference
+		
+			American Institute of Aeronautics and Astronautics
+			1997
+		
+	
+	Isaacson, D., Davis, T., and Robinson, III, J., "Knowledge-based Runway Assignment for Arrival Aircraft in the Terminal Area," Guidance, Navigation, and Control Conference (GNC), 1997. https://doi.org/10.2514/6.1997-3543.
+
+
+
+	
+		Air Traffic Control Decision Support for Integrated Community Noise Management
+		
+			SanderJ.
+		
+		
+			HendrikusG.
+		
+		10.5772/25215
+		
+	
+	
+		Aeronautics and Astronautics
+		
+			InTech
+			2020
+		
+		
+			National Aeronautics and Astronautics Administration
+		
+	
+	National Aeronautics and Astronautics Administration, "Airspace Technology Demonstration 2 (ATD-2): Integrated Ar- rival/Departure/Surface (IADS) Traffic Management," , 2020. URL https://aviationsystems.arc.nasa.gov/research/atd2/index. shtml.
+
+
+
+	
+		Predicting Runway Allocation with Support Vector Machine and Logistic Regression
+		
+			KEisinga
+		
+		
+			July 2016
+		
+		
+			Tilburg University
+		
+	
+	Master's thesis
+	Eisinga, K., "Predicting Runway Allocation with Support Vector Machine and Logistic Regression," Master's thesis, Tilburg University, July 2016.
+
+
+
+	
+		Modeling of runway assignment strategy by human controllers using machine learning
+		
+			YoichiNakamura
+		
+		
+			RyotaMori
+		
+		
+			HisaeAoyama
+		
+		
+			HyuntaeJung
+		
+		10.1109/dasc.2017.8102099
+		
+	
+	
+		2017 IEEE/AIAA 36th Digital Avionics Systems Conference (DASC)
+		
+			IEEE
+			2017
+			
+		
+	
+	Nakamura, Y., Mori, R., Aoyama, H., and Jung, H., "Modeling of Runway Assignment Strategy by Human Controllers Using Machine Learning," 2017 IEEE/AIAA 36th Digital Avionics Systems Conference (DASC), 2017, pp. 1-7. https: //doi.org/10.1109/DASC.2017.8102099.
+
+
+
+	
+		The Multiple Runway Planner (MRP): Modeling and Analysis for Arrival Planning
+		
+			MatthewBerge
+		
+		
+			AslaugHaraldsdottir
+		
+		
+			JulienScharl
+		
+		10.1109/dasc.2006.313684
+		
+	
+	
+		2006 ieee/aiaa 25TH Digital Avionics Systems Conference
+		
+			IEEE
+			2006. 2006
+			
+		
+	
+	Berge, M. E., Haraldsdottir, A., and Scharl, J., "The Multiple Runway Planner (MRP): Modeling and Analysis for Arrival Planning," 2006 IEEE/AIAA 25th Digital Avionics Systems Conference (DASC), 2006, pp. 1-11. https://doi.org/10.1109/DASC. 2006.313684.
+
+
+
+	
+		Progress Toward Future Runway Management
+		
+			GaryLohr
+		
+		
+			SherilynBrown
+		
+		
+			SteveAtkins
+		
+		
+			SteveEisenhawer
+		
+		
+			TerryBott
+		
+		
+			DouLong
+		
+		
+			ShahabHasan
+		
+		10.2514/6.2011-6925
+		
+	
+	
+		11th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference
+		
+			American Institute of Aeronautics and Astronautics
+			2011
+		
+	
+	Lohr, G., Brown, S., Atkins, S., Eisenhawer, S., Bott, T., Long, D., and Hasan, S., "Progress Toward Future Runway Management," 11th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference, 2011. https://doi.org/10.2514/6.2011-6925.
+
+
+
+	
+		Development of a Runway Allocation Optimisation Model for Airport Strategic Planning
+		
+			SanderJHeblij
+		
+		
+			RolandA AWijnen
+		
+		10.1080/03081060801948191
+		
+	
+	
+		Transportation Planning and Technology
+		Transportation Planning and Technology
+		0308-1060
+		1029-0354
+		
+			31
+			2
+			
+			2008
+			Informa UK Limited
+		
+	
+	Heblij, S., and Wijnen, R., "Development of a Runway Allocation Optimisation Model for Airport Strategic Planning," Transportation Planning and Technology, Vol. 31, No. 2, 2008, pp. 201-214. https://doi.org/10.1080/03081060801948191.
+
+
+
+	
+		Fuser Deeper Dive (Mediation & Use Cases)
+		
+			SMGorman
+		
+		
+			JMBurke
+		
+		
+			IJRobeson
+		
+		
+			BSPhipps
+		
+		
+			2019
+		
+	
+	Gorman, S. M., Burke, J. M., Robeson, I. J., and Phipps, B. S., "Fuser Deeper Dive (Mediation & Use Cases)," 2019.
+
+
+
+	
+		Structure of situational tasks in the air traffic controllers� professional training
+		
+			KaterynaSurkova
+			0000-0002-1388-7611
+		
+		
+			MarynaLomakina
+		
+		10.33251/2522-1477-2020-8-124-129
+		
+	
+	
+		Scientific Bulletin of Flight Academy. Section: Pedagogical Sciences
+		SBFA
+		2522-1477
+		
+			8
+			
+			2020
+			Flight Academy of National Aviation University
+		
+	
+	Federal Aviation Administration, "National Flight Data Center 28 Day NASR Subscription," online, 2020. URL https: //www.faa.gov/air_traffic/flight_info/aeronav/aero_data/NASR_Subscription/.
+
+
+
+	
+		
+			RKille
+		
+		Summary of the July 1999 Internal Departure Delay Investigation for ZDV Along with New Departure Delay Data From ZOB, ZID, ZFW and ZDV
+		
+			2004
+		
+	
+	presentation
+	Kille, R., "Summary of the July 1999 Internal Departure Delay Investigation for ZDV Along with New Departure Delay Data From ZOB, ZID, ZFW and ZDV," presentation, 2004.
+
+
+
+	
+		Practice for Application of Federal Aviation Administration (FAA) Federal Aviation Regulations Part 21 Requirements to Unmanned Aircraft Systems (UAS)
+		10.1520/f2505-07
+		
+	
+	
+		Federal Aviation Administration
+		
+			ASTM International
+			February 2016
+		
+	
+	Federal Aviation Administration, "Wake Turbulence Recategorization," , February 2016. URL https://www.faa.gov/ documentLibrary/media/Order/JO_7110_659C.pdf.
+
+
+
+	
+		A Recursive Multi-step Machine Learning Approach for Airport Configuration Prediction
+		
+			SKhater
+		
+		
+			JRebollo
+		
+		
+			WCoupe
+		
+		
+			2021
+		
+	
+	Submitted to AIAA Aviation Forum
+	Khater, S., Rebollo, J., and Coupe, W., "A Recursive Multi-step Machine Learning Approach for Airport Configuration Prediction," Submitted to AIAA Aviation Forum, 2021.
+
+
+
+	
+		A Machine Learning Approach to Predict Aircraft Landing Times using Mediated Predictions from Existing Systems
+		
+			DanielWesely
+		
+		
+			AndrewChurchill
+		
+		
+			JohnSlough
+		
+		
+			WilliamJCoupe
+		
+		10.2514/6.2021-2402
+	
+	
+		AIAA AVIATION 2021 FORUM
+		
+			American Institute of Aeronautics and Astronautics
+			2021
+		
+	
+	Submitted to AIAA Aviation Forum
+	Wesely, D., Churchill, A., Slough, J., and Coupe, W., "A Machine Learning Approach to Predict Aircraft Landing Times using Mediated Predictions from Existing Systems," Submitted to AIAA Aviation Forum, 2021.
+
+
+
+	
+		Time Based Flow Management (TBFM) as a service—With NextGen SOA
+		
+			KiranChittargi
+		
+		
+			LockheedMartin
+		
+		10.1109/icnsurv.2013.6548660
+		
+	
+	
+		2013 Integrated Communications, Navigation and Surveillance Conference (ICNS)
+		
+			IEEE
+			August 2020
+		
+	
+	Time Based Flow Management
+	Federal Aviation Administration, "Time Based Flow Management," , August 2020. URL https://www.faa.gov/nextgen/cip/tbfm/.
+
+
+
+	
+		Application of the Logistic Function to Bio-Assay
+		
+			JosephBerkson
+		
+		10.1080/01621459.1944.10500699
+	
+	
+		Journal of the American Statistical Association
+		Journal of the American Statistical Association
+		0162-1459
+		1537-274X
+		
+			39
+			227
+			
+			1944
+			Informa UK Limited
+		
+	
+	Berkson, J., "Application of the Logistic Function to Bio-Assay," Journal of the American Statistical Association, Vol. 39, No. 227, 1944, pp. 357-365.
+
+
+
+	
+		Scikit-learn: Machine Learning in Python
+		
+			FPedregosa
+		
+		
+			GVaroquaux
+		
+		
+			AGramfort
+		
+		
+			VMichel
+		
+		
+			BThirion
+		
+		
+			OGrisel
+		
+		
+			MBlondel
+		
+		
+			PPrettenhofer
+		
+		
+			RWeiss
+		
+		
+			VDubourg
+		
+		
+			JVanderplas
+		
+		
+			APassos
+		
+		
+			DCournapeau
+		
+		
+			MBrucher
+		
+		
+			MPerrot
+		
+		
+			EDuchesnay
+		
+	
+	
+		Journal of Machine Learning Research
+		
+			12
+			
+			2011
+		
+	
+	Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Prettenhofer, P., Weiss, R., Dubourg, V., Vanderplas, J., Passos, A., Cournapeau, D., Brucher, M., Perrot, M., and Duchesnay, E., "Scikit-learn: Machine Learning in Python," Journal of Machine Learning Research, Vol. 12, 2011, pp. 2825-2830.
+
+
+
+	
+		XGBoost
+		
+			TianqiChen
+		
+		
+			CarlosGuestrin
+		
+		10.1145/2939672.2939785
+		
+	
+	
+		Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining
+		the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data MiningNew York, NY, USA
+		
+			ACM
+			2016
+			
+		
+	
+	Chen, T., and Guestrin, C., "XGBoost: A Scalable Tree Boosting System," Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, Association for Computing Machinery, New York, NY, USA, 2016, p. 785-794. https://doi.org/10.1145/2939672.2939785.
+
+
+
+	
+		
+			LBălan
+		
+		
+			)Kiyo
+		
+		
+			KKDeriabin
+		
+		
+			DHoang
+		
+		
+			LIvaniuk
+		
+		
+			ADada
+		
+		
+			YDatta
+		
+		
+			DPatel
+		
+		
+			ZWrigley
+		
+		
+			GDanov
+		
+		
+			IStichbury
+		
+		
+			JKhan
+		
+		
+			NTsaousis
+		
+		
+			NTheisen
+		
+		
+			MWalker
+		
+		
+			WNguyen
+		
+		
+			TWestenra
+		
+		
+			RCarvalho
+		
+		
+			LTrevisani
+		
+		
+			MDBertoli
+		
+		
+			SMawjee
+		
+		
+			SNijholt
+		
+		
+			BVukolov
+		
+		
+			DFischer
+		
+		
+			KVijaykumar
+		
+		
+			YMinami
+		
+		10.5281/zenodo.4336685
+		quantumblacklabs/kedro: 0.17.0
+		
+		bru5, and dr3s
+		
+			Dec. 2020
+		
+	
+	Bălan, L., (Kiyo), K. K., Deriabin, D., Hoang, L., Ivaniuk, A., Dada, Y., Datta, D., Patel, Z., Wrigley, G., Danov, I., Stichbury, J., Khan, N., Tsaousis, N., Theisen, M., Walker, W., Nguyen, T., Westenra, R., Carvalho, L., Trevisani, M. D., Bertoli, S., Mawjee, S., sasaki takeru, Nijholt, B., Vukolov, D., Fischer, K., Vijaykumar, Minami, Y., bru5, and dr3s, "quantumblacklabs/kedro: 0.17.0," , Dec. 2020. https://doi.org/10.5281/zenodo.4336685, URL https://doi.org/10.5281/zenodo.4336685.
+
+
+
+	
+		Practice for Application of Federal Aviation Administration (FAA) Federal Aviation Regulations Part 21 Requirements to Unmanned Aircraft Systems (UAS)
+		10.1520/f2505-07
+		
+		
+			null
+			ASTM International
+		
+	
+	TFMData Service
+	Federal Aviation Administration, "TFMData Service," , ????. URL https://cdm.fly.faa.gov/?page_id=2288.
+
+
+
+	
+		Distributing net-enabled federal aviation administration (FAA) weather data
+		
+			MarkSimons
+		
+		10.1109/icnsurv.2008.4559189
+		
+	
+	
+		2008 Integrated Communications, Navigation and Surveillance Conference
+		
+			IEEE
+			
+		
+	
+	Federal Aviation Administration, "SWIM Terminal Data Distribution System (STDDS)," , ????. URL https://www.faa.gov/air_ traffic/technology/swim/stdds/.
+
+
+				
+			
+		
+	

file152.txt

@@ -0,0 +1,666 @@
+
+	
+	
+		
+I. IntroductionC onflict prediction is an integral part of maintaining safe separation between aircraft in the National Airspace System.This task is currently handled completely by human controllers, but as air traffic demand continues to rise, automation will play a more prominent role in conflict detection and possibly resolution.To enable this transition, it is important to develop automation tools that can operate reliably and with a high degree of accuracy in real-world settings.One of the primary challenges in predicting conflicts using automation in a realistic setting is dealing with the effects of uncertainty.Automated conflict probes often utilize some type of trajectory generator to build a predicted trajectory, and this predicted trajectory never exactly matches what the aircraft will actually fly.Depending on the type and magnitude of the uncertainty, as well as the capabilities of the trajectory predictor, these errors can range from relatively minor errors, such as differences in turn modeling, to major issues like intent errors that can greatly affect the accuracy of trajectory predictions. 1 These errors, in turn, affect the ability of the conflict probe to detect conflicts and to suggest resolutions that are free from conflict.There are two primary ways to address the issue of errors in the trajectory prediction: improve the accuracy of the trajectory generator, or compensate for the error by adapting the conflict probe.For this work, the focus is on compensating for errors, rather than trying to improve trajectory predictions.Specifically, the authors are detecting and resolving conflicts at greater than the legal standard for separation of five nmi horizontally and 1,000 ft vertically using a geometric conflict detection scheme.While improving trajectory prediction accuracy is beneficial for detecting conflicts when using automated conflict probes, 2,3,4,5 and validating the accuracy of a trajectory prediction is seen as a necessary part of the future National Airspace System 6 (NAS), improving trajectory prediction performance often requires some form of equipage on the aircraft and/or data sharing, which can make the solution more difficult and expensive to implement in actual operations.There are alternative approaches that try to use adaptive algorithms to improve the accuracy of trajectory predictions, particularly during climb, by adapting either the modeled aircraft thrust 7,8 or weight. 9These approaches show promise, but are not yet mature.On the mitigation side, there are many studies that look at using increased separation criteria, or "buffers," to deal with uncertainty or error.Some examples include studies examining enlarged horizontal detection ranges in the presence of cruise speed errors, 10 wind prediction errors, 11 and maneuver-initiation time errors. 12One common theme through much of this work is a primary focus on horizontal errors and the effectiveness of using horizontal buffers or probabilistic conflict detection schemes to account for those errors.Some simulation test beds, such as the Center-TRACON Automation System (CTAS) 13 use a vertical buffer for aircraft that are transitioning altitudes, but that buffer is often on the order of hundreds of feet, which is not enough to account for uncertainties during the descent phase.
+II. BackgroundThe authors have examined the effectiveness of a buffer to aid automated conflict detection and resolution in the presence of trajectory prediction errors using a geometric conflict detection algorithm, but that work was focused on using buffers in the horizontal plane of up to two nmi. 14In that work, the authors looked at a range of uncertainties and tried to determine what effects these uncertainties had on missed-alert rates, false-alert rates, and losses of separation (LOS) using automated conflict detection and resolution tools in a non-real-time simulation.Missed alerts are conflicts that should have been detected but are not, and false alerts are conflicts that are predicted to occur but do not.That study found, among other things, that there was a very high rate of missed detections and LOS cases near an aircraft's top-of-descent (TOD) point.Additionally, increasing the horizontal separation for these cases did not have a noticeable effect on the number of LOS observed.Follow-on, unpublished work revealed the difficulty the conflict probe had in the vertical plane due to the very small legal vertical separation requirement.With many aircraft descending fast enough to completely pass through the legal separation in under 20 seconds, it does not take much uncertainty to create a situation that results in an LOS using standard vertical separation.This danger can be reduced by asking the aircraft to provide intent information about its planned descent, such as the anticipated TOD point or its desired descent profile, but an aircraft might not exactly fly the stated profile anyway.As an example, there can be an error on the order of a few nautical miles between the Flight Management System's (FMS) predicted and actual TOD point. 15Additionally, depending on the airspace class, aircraft in level flight can pass over another level aircraft 1,000 feet below and be perfectly legal and safe.It is only when aircraft are transitioning altitudes that detections in the vertical plane become a concern.One solution is to simply clear all of the airspace beneath an aircraft that might be descending soon.This would ensure safety, but could decrease airspace capacity and increase the total delay experienced by aircraft near either their own or some other aircraft's TOD point.Therefore, it was decided to design vertical conflict detection buffers that would provide just enough warning to ensure safe separation, while minimizing the amount of extra airspace that would need to be cleared in addition to the amount required by the legal separation standard.The question of "How much warning is enough?" is one that is not totally answered.For this study, the vertical separation buffers have an altitude range that covers roughly four minutes of descent at the nominal predicted descent rate for each aircraft.The exact sizes of the buffers, therefore, vary from flight to flight.Four minutes was chosen as the look-ahead time for this study as a reasonable minimum that should allow a resolution tool to resolve a potential conflict before it becomes an LOS.
+III. Simulation Environment
+A. ACES and AAC AutoresolverThe simulation test bed used for the study is the Airspace Concepts Evaluation System (ACES). 16This is a non-real-time simulation that uses a four-degree-of-freedom model to create aircraft trajectories based off performance data and stored flight plan information.The aircraft performance data are derived from the Base of Aircraft Data (BADA) 17 and the flight plans are created from the filed flight plans for days in the National Airspace System (NAS).For the current study, the traffic scenario for all the data runs consisted of the flight plans for 9,272 flights across the US, representing about four hours worth of takeoffs during the busiest part of a day in 2005.The wind data is RUC data recorded from a day in May of 2002.The simulation includes a conflict probe that uses knowledge of those flight plans to check for conflicts along an aircraft's predicted trajectory.Conflict resolution is handled by the Autoresolver, 18 which is a component of the Advanced Airspace Concept 19 (AAC).The simulation only examines aircraft during their flight through Center airspace, from departure fix to arrival fix, and does not examine flights inside terminal airspace.The current version of AAC attempts to resolve conflicts at roughly eight minutes until predicted LOS, referred to as its action time, though if both aircraft are headed to the same arrival fix and within 20 minutes of that fix, AAC can attempt to resolve the conflict at 20 minutes to predicted LOS.Typically, AAC tries to issue a resolution that is free of conflicts for up to four minutes beyond its action time.
+B. Trajectory PredictionIn this study, two trajectories are created for each aircraft.The first is the "real" trajectory, which is the one that the aircraft will actually fly.From that trajectory a "perturbed" trajectory is created, which includes the prediction errors being tested (figure 1(a) has an example of this).Every minute, the perturbed trajectory is sent to the conflict detection algorithm and then on to the AAC Autoresolver if a resolution is required.
+IV. Experiment SetupThe study consists of two sections.The first section examines the performance of the vertical buffer in terms of conflict detection only.This allows the vertical buffers to be examined in repeatable data sets using multiple error types, with every case having the exact same number of actual conflicts because each aircraft will fly the same "true" trajectory in each run.The second study uses the vertical buffers for both detecting and resolving conflicts.Both studies use the same error types.The legal separation requirement used for all runs is defined as five nmi of horizontal separation and 1,000 ft of vertical separation.For conflict detection, a six nmi horizontal range is used for all cases.When issuing resolutions, the Autoresolver attempts to obtain seven nmi of horizontal separation.The simulations that were run without a vertical buffer use 1,000 ft vertically for the entire flight, while the cases with vertical buffers use a specialized buffer near TOD and during descent, and 1,000 ft elsewhere.The vertical buffers used in this study are described in detail in the next subsection.
+A. MethodTable 1 shows the five types of trajectory uncertainty used in this study.The error rages were chosen to be roughly in line with values used in other studies, though they were chosen to be slightly larger overall.Cruise speed, descent speed, and TOD location are modeled as uniform distributions around zero.In this simulation, "descent speed" includes an error in the predicted descent Mach number and descent CAS.Wind speed errors are modeled as a prediction that is 25% stronger than the actual wind, as read from a RUC wind file, with the direction unchanged.The aircraft fuel weight is used to adjust the aircraft's weight, and is modeled as a uniform distribution applied to predicted aircraft fuel weights around a nominal value.The references for the error ranges are included in the table, while a more detailed description of how the errors are implemented in ACES can be found in previous work by the authors.
+Error SourceError Range Top of Descent Location roughly +/-10 nmi 20,21 Descent Speed +/-10% The conflict detection portion of this study examined six error configurations and had two vertical buffer settings, for a total of 12 data runs.The test matrix for this portion of the study is shown in table 2. Four of the errors were examined individually (TOD location, descent speed, cruise speed, and wind speed), with a fifth case that had all of the errors, including weight, enabled together.Aircraft weight was not examined on its own because previous work has shown it to have a mild, though non-zero, impact on the aircraft trajectory near TOD. 14A case with no error was also examined to establish the baseline behavior without trajectory prediction errors.All the error configurations had one simulation run with the enhanced vertical buffers disabled and one with them enabled.The resolution portion of the study consisted of five simulations with all the errors enabled.The test matrix for this part of the study is shown in table 3. The first three runs used the full error range with vertical buffers disabled, set to the full size they were in the detection runs, and set to 80% of full size.The last two runs used 50% of the values shown in table 1, and included a case with no vertical buffer and one with a buffer set to 50% of the full value used in the detection runs.This last pair of runs was used to roughly simulate how effective a vertical buffer combined with improved trajectory predictions would be.The vertical conflict detection buffer consists of two parts (see figure 1(a) and figure 1(b)).The first part was a buffer around the predicted top of descent point.This buffer was constructed by assuming the aircraft might descend as much as four minutes early or late.Using the predicted average descent for each aircraft, this buffer is extended along the descent for four minutes and should provide a minimum of three minutes warning.The second part of the vertical detection buffer is implemented after the aircraft has started to descend and is shown in figure 1(b).This buffer is created at the aircrafts current position during each conflict detection cycle and extends forward along the predicted trajectory.The buffer takes the predicted descent rate at the temporal midpoint of its remaining descent, creates a "fast-descent" and "slow-descent" profile, and extends those four minutes into the future from the aircraft's current position.The fast-descent profile assumes the aircraft is descending 400 fpm faster than predicted, while the slow descent profile assumes a descent rate 200 fpm slower than predicted.These values were based on results from preliminary data  collected for this study.These two buffers comprise the "full" buffer case.The simulations with reduced buffer size simply scaled the early/late descent time and fast/slow climb rate by a percentage value.The look-ahead time was four minutes along the aircrafts predicted descent.Alternative look-ahead times and buffer shapes will be examined in a future study.
+B. Conflict Detection MetricsThe main metrics being used in the detection part of the study are the number of missed and false alerts for a specific predicted time until LOS.A missed alert is defined as a case where there is a loss of separation along an aircraft's true trajectory that is not detected by looking at the perturbed, predicted trajectory.This is recorded by the time until the aircraft would actually have a loss of separation.Depending on the error type, there are generally more missed detections when the aircraft are still 20 minutes apart than when they are closer.In this study, we are most concerned with the missed detections that occur with 3 minutes or less until the time of first loss of separation.These late detections can be very difficult to solve, as neither aircraft has much time to move out of the way.In addition, even though these conflicts do not always result in a loss of separation, they are cases that could more easily become losses, depending on the situation in the surrounding airspace and the capabilities of the person or automation attempting to resolve the conflict.False alerts are defined similarly to missed, except that false alerts are cases where the perturbed predicted trajectory detects a potential conflict that the true trajectory reveals will not actually occur.Furthermore, for all cases in this study, the legal separation requirement (five nmi horizontally and 1,000 ft vertically) is used to determine whether or not a conflict actually occurred.This means that using any enlarged conflict detection criteria will produce false alerts, even with zero trajectory prediction error.These false alert cases do not impact safety directly, but they can have a large impact on efficiency, as a high rate of these alerts means that many aircraft are being moved to resolve conflicts that would not have actually occurred.This, in turn, adds to the delay for aircraft flying through the area.Therefore, even though some non-zero value should be expected any time an enlarged detection criteria is used, it is desirable to keep this rate as low as possible without degrading safety.
+C. Conflict Resolution MetricsFor the portion of the study looking at resolutions, the primary metrics are the number of losses of separation, the number of resolutions issued, and the total delay for aircraft in flight due to conflict resolution maneuvers.The average delay per resolution is also reported.The LOS metric is the driving one for this study, as it represents failures of the system that could affect safety.It is defined as any case where two aircraft in enroute airspace pass within the legal separation requirement of each other.It should be noted that LOS cases are expected in this simulation, because the vertical buffers presented here are only a partial solution aimed at significantly reducing or eliminating the number of losses seen in the descent phase of flight.Also, the Autoresolver is only the first level of the multi-layered AAC, so LOS cases in this study could be more accurately described as conflicts that would not be resolved by Autoresolver, and would fall through to the next layer of an overall system.Analysis of those other systems is beyond the scope of the current work, so for simplicity, conflicts that the Autoresolver fails to resolve will be called LOS cases.The number of resolutions and total delay are ways to quantify the effect of the vertical buffers on system efficiency, as compared to a system with no buffers.Any changes to the system that improve robustness will likely have efficiency penalties, but keeping track of the number of extra resolutions and the amount of extra delay allow for comparisons between options, both now and in future work.
+V. Results
+A. Conflict DetectionThe goal of the first part of the study is to check how many conflicts are detected at least three minutes before predicted loss of separation.As a reminder, a "missed alert" is an alert where the flown "truth" trajectory predicts a loss of separation while the perturbed predicted trajectory does not.A "false alert" is the case where the perturbed predicted trajectory identifies a conflict that would not have occurred, based on the "true" flown trajectory.Also as a reminder, conflicts were only detected, not resolved in this portion of the study, so it is possible for a single conflict to produce multiple false alerts and/or missed alerts.Figure 2 details the percentage of potential conflicts that were missed conflict alerts, plotted as a function of time until loss for each of the five error cases.The chart on the left uses the standard legal separation vertically for conflict detection, while the chart on the right includes the vertical detection buffers.Both charts use the same horizontal detection range of six nmi, and both charts show all error types, with the case that had no trajectory prediction errors included as a reference.The case without the vertical buffer illustrates the problem caused by these uncertainties.While there are fewer missed detections as the time to loss decreases, many of those missed detections persist until there is very little time to resolve them.This is especially true for the descent speed and top-of-descent position errors which do not really show a decrease in the number of missed alerts until there is less than five minutes until LOS.These two errors help drive the curve for the "all error" case up, so that at four minutes to actual LOS for the all-error case, 8.7% of conflicts are missed by the detection algorithm without the vertical buffer.The right chart shows the effectiveness of the current iteration of the vertical detection buffer.It should be noted that these are missed alerts for all phases of flight, so there are some in climb or cruise that vertical detection buffers for descent simply will not address, especially in the case with all the errors combined.There is a noticeable drop in missed alerts for all error types due to the addition of the buffers, especially with less than five minutes until the loss of separation.To continue the example from the previous paragraph, enabling the buffer drops the missed detection value at four minutes to LOS to 2.4% for conflicts in all phases of flight.Considering that the vertical buffer will not directly impact conflicts that do not involve at least one arriving aircraft, that reduction is significant.Looking at an error that directly affects arriving aircraft, the number of missed detections at eight minutes to LOS for TOD error drop from 7.5% without the vertical buffer, to 1.2% with the buffer enabled.At four minutes, there 4.7% of conflicts are missed with TOD error and no vertical buffer and 0.1% missed for the same error with vertical buffers enabled.These results further strengthen the position that conflicts involving aircraft descending into their arrival fix make up a large portion of the conflicts that are difficult to detect, and that a vertical buffer can largely mitigate this.Figure 3 shows the percentage of alerts that were false alerts for the conflict detection study.This figure shows that adding vertical buffers has a significant effect on false alerts, especially below 10 minutes until predicted LOS.As these were the results with no conflict resolutions implemented, every aircraft flew the same true trajectory in both the left and right charts.Therefore, the overall increase should be entirely attributable to the added vertical detection buffer.To give an example of the scope of the increase, in the case with all errors, at 8 minutes until predicted LOS, the percentage of detections that were false alerts is 30% without the vertical buffers and 53% with the buffers enabled, a jump of 23%.The case without error saw the biggest jump, with the percentage of false alerts moving from 17% without the vertical buffers to 49% with buffers, or a jump of 32%.The magnitude of these increases implies that there was a lot of traffic around aircraft that were near or past their TOD point.This, in turn, implies that there are many aircraft that could be at risk if an aircraft deviates much from its predicted trajectory near TOD.The net result of a large increase in the number of resolutions issued due to these false alerts is supported by results presented later in the paper.The fact that there are so many more detections also means that it will likely be difficult to find a strategy for mitigating uncertainties in trajectory prediction for this level of trajectory prediction error that does sharply increase the number of false alerts or add significant delay, as there are simply many aircraft in relative proximity during the descent phase for aircraft that are arriving.
+B. Conflict ResolutionThe results with conflict resolutions enabled are discussed in this section.The primary metrics analyzed are the number of losses of separation, the number of resolutions issued, and the total delay and delay per resolution.This section will explore the effectiveness of the vertical buffers and the penalty for using the buffer in terms of the amount of extra delay created and number of resolutions issued.Figure 4 shows the losses of separation, categorized by flight phases of the two aircraft involved, for the simulations with and without the vertical buffers.All cases detect conflicts at six nmi horizontally, and attempt to obtain at least seven nmi horizontally when issuing resolution maneuvers.The left group of columns are cases where at least one aircraft was climbing, the second group is the case where both aircraft were roughly in their cruise segment, the third group is for cases where one aircraft was an arriving aircraft descending into the Terminal area, and the fourth group is the special case where one aircraft was a descending arrival and the other was a departing aircraft still climbing to cruise altitude.The main point of this plot is to show that simply by using increased vertical separation criteria near and after an aircraft's top of descent point, one can dramatically decrease the number of times uncertainty induces an LOS.However, these buffers alone are not enough to completely remove the problem.The case without the vertical buffer once again emphasizes the difficulties the conflict detection and resolution algorithms have with descending aircraft, with 209 of the 276 total LOS cases involving at least one aircraft that was descending near its arrival airport, as shown by the right two blue columns.When the buffer is enabled, the number of total LOS cases drops to 49, while the number of losses involving arriving aircraft drops to 12 (right two light blue columns combined).Ideally, the number of LOS for arriving aircraft would be zero with the vertical buffer enabled, but there were a few cases that slipped through.While it should be possible to keep increasing the size of the vertical buffer to cover all of those cases, it is worth investigating a different approach for dealing with those last few losses of separation.The idea that it might not be efficient to continually increase the vertical buffer to remove all LOS cases came from the results of the 80% buffer run, shown as the green columns in the figure.This scenario actually produced slightly better results than the full buffered case as far as dealing with LOS cases occurring for arriving aircraft, for reasons that are not clear.Unfortunately, due to timing constraints, exploring the causes of these losses of separation involving arriving aircraft that are not resolvable even with the vertical buffers is beyond the scope of this paper.However, this analysis will be done in future work, as accounting for these cases is necessary for making a system that is truly robust to trajectory prediction errors.One interesting result was the decrease in the number of LOS cases between two aircraft in cruise for both buffer cases.This could simply be the result of these buffers being used for temporary, in cruise descents to avoid conflicts, but further analysis would be required before it could be claimed as a benefit.As the focus of this study is on LOS cases involving an arriving aircraft, the examination of cruise LOS cases will be also deferred to a future study.The number of resolutions as well as the delay per resolution are summarized in table 4. It is immediately obvious that the addition of these vertical buffers produces a steep increase in both the number of resolutions as well as the total delay experienced by aircraft in the system, with the number of resolutions issued increasing by 53% and the total delay by 106%.However, one has to consider that the unbuffered case also had a very large number of losses of separation that need to be addressed, so some penalty is likely unavoidable.The 80% buffer results are more promising, showing an appreciable decrease in the number of resolutions and amount of delay added to the system, increasing the number of resolutions by 42% and the total delay by 69%.The fact that this 80%buffer was able to perform just as well as the full buffer in regards to accounting for losses of separation involving arriving aircraft was a major point.It strengthened the idea that alternative options might be best suited to dealing with the few losses that the vertical detection buffer does not catch, as there seems to be diminishing returns in regards to catching LOS cases when increasing the vertical buffer beyond a certain size.Planned future work includes developing alternative methods for dealing with these last few LOS cases, as well as exploring the use of smaller buffers.This could lead to a system with no LOS involving arriving aircraft and with less system-wide delay caused by resolution maneuvers than would be possible using the enhanced vertical buffers alone.This section describes a pair of simulations run with all of the trajectory prediction error ranges set to 50% of their full values, to roughly simulate the effects of improving trajectory prediction accuracy.The first run used no vertical buffer, while the second used a vertical buffer that was also set to 50% of the full value, to take advantage of the reduced error range.As trajectory accuracy improves, one would expect to see fewer losses, a smaller buffer requirement, and more efficiency in terms of the number of resolutions and the delay per resolution, though it is difficult to predict the amount of savings without simply collecting the data.The results of these runs are shown in figure 5, along with the original, full error case with no vertical buffer.It is immediately apparent that, even without the addition of the vertical buffer, cutting the trajectory prediction error in half cuts more than half of the losses off separation in all flight regimes.Furthermore, the addition of a vertical buffer (also half the size of the first one tested in this study), cuts the overall number of LOS cases from 117 in the unbuffered case to 23, and the cases involving arriving aircraft from 87 to 3.Table 5 shows the results of the half-error cases, with the full-error case (no vertical buffer) provided for reference.Compared to the half-error, no vertical buffer case, implementing the half-sized vertical buffer increased the number of resolutions issued by 42% and increased the total delay due to resolution maneuvers by 82%.These increases are in line with the percentages seen for implementing the vertical buffers in table 4, though the overall numbers are lower because of the reduced number of resolutions and delay in the half-error case without vertical buffers.This illustrates the effectiveness of combining the approaches of improving trajectory prediction accuracy and implementing vertical buffers, as the overall number of LOS cases, the number of added resolutions, and the increase in delay are all significantly reduced when the buffer is implemented in the half-error case.
+VI. Future WorkAs stated previously, these vertical buffers are only a first step to making a system that can reliably predict and resolve all conflicts in the presence of uncertainty.The major problem area considered prior to this work was when one or more aircraft were descending towards their arrival fix.However, uncertainties and their resulting trajectory prediction errors can lead to losses of separation in climb and cruise, as well.Additionally, even our relatively large vertical buffers were not sufficient to completely deal with trajectory prediction errors during descent.The next step is to explore ways to remove those last few LOS cases for arriving aircraft.Additionally, more work needs to be done examining different vertical buffer sizes and lookahead times.Following that, the focus will shift to using buffers and perhaps an adaptive climb algorithm for removing LOS cases during climbs, and then figuring out a way to remove LOS cases in cruise as efficiently as possible.Increasing the range of types of uncertainty is also part of the planned work, with the eventual goal of producing a system that can be made robust to varying levels of trajectory prediction uncertainty in all phases of fight.A secondary goal is to do that while limiting the decrease in system efficiency in terms of delay caused by resolution maneuvers.
+VII. ConclusionsThe results of this study point to vertical buffers as an effective first step for detecting potential losses of separation involving aircraft descending into terminal airspace in the presence of trajectory prediction errors.Results showed that errors involving predictions during the descent phase are difficult to detect with much more than a few minutes until loss with just a conventional horizontal buffer.Further, implementing vertical separation buffers can be an effective technique for reducing the number of cases that are missed by the detection algorithm, especially for errors in TOD location and descent speed.Results also showed that enabling the vertical buffers decreased the number of missed alerts with four minutes to LOS from 8.7% for all errors without the vertical buffers to 2.4% with buffers.For TOD errors, the missed alerts at 4 minutes until LOS dropped from 4.7% without the vertical buffer to 0.1% with the buffers.Enabling the buffers also increased the percentage of detections that were false alerts by around 20% to 30% for predicted LOS times in the eight minute range for all error types.When AAC was allowed to resolve detected conflicts, implementing the full vertical buffer produced a significant reduction in the number of losses of separation in the simulation.The number of LOS cases involving arriving aircraft was reduced from 207 in the case with full error and no vertical buffer to 12 when the full-sized vertical buffer was used.Setting the vertical buffer to 80% of the full size actually reduced the number of LOS for arriving aircraft compared to the full buffer, dropping the number of cases to 10.In terms of the other metrics, the full buffer increased the number of resolutions issued by 53% and the amount of delay accumulated by aircraft executing resolution maneuvers by 106%, while using the 80% buffer increased the number of resolutions by 41% and the delay by 69%.These results imply that vertical buffers can be effective in reducing the number of LOS cases involving arriving aircraft, but there is a point beyond which increases to the buffer size results in larger delay and more resolutions with no real reduction in the number of LOS cases, and alternative methods for catching the remaining LOS cases for arriving aircraft need to be developed.The study with 50% uncertainty showed that improvements in trajectory prediction significantly improve the ability of the vertical buffer to account for all LOS cases with arriving aircraft.Additionally, the system as a whole runs more efficiently with less resolutions and total delay when there is reduced uncertainty.While this is expected, it suggests that a joint approach of reducing trajectory prediction error and building robust detection schemes is likely the most viable way to achieve a system that has no LOS cases in the presence of multiple trajectory prediction errors.Figure 1 .1Figure 1.Vertical conflict detection buffer (a) before and (b) after the top-of-descent.
+Figure 2 .2Figure 2. Missed alerts for multiple uncertainties using conflict detection only.
+Figure 3 .3Figure 3. False alerts for multiple uncertainties using conflict detection only.
+Figure 4 .4Figure 4. Losses of separation with all uncertainty enabled.
+Figure 5 .5Figure 5. Losses of separation with 50% uncertainty enabled.
+14 Table 1 .141Trajectory Prediction Error Source and Range.
+Table 2 .2Test Matrix: Conflict Detection.Error TypeEnhanced Vertical Buffer SettingNo ErrorDisabled; Full BufferTop of Descent LocationDisabled; Full BufferDescent Speed (CAS and Mach)Disabled; Full BufferCruise SpeedDisabled; Full BufferWind SpeedDisabled; Full BufferAll Errors (including Weight)Disabled; Full Buffer
+Table 3 .3Test Matrix: Conflict Detection and Resolution.Error TypeBuffer SizeEnhanced Vertical Buffer SettingAll ErrorsFull Error Range Disabled; Full Buffer; 80% BufferAll Errors50% Error RangeDisabled; 50% Buffer
+Table 4 .4Resolution efficiency metrics by vertical buffer, full uncertainty.ConfigurationResolutions IssuedTotal Delay, minAverage Delay per Resolution, sAll Error; No Buffer12,574(Base)3,844(Base)18.3All Error; Full Buffer 19,180 (Base+53%) 7,912 (Base+106%)24.8All Error; 80% Buffer 17,794 (Base+42%) 6,492 (Base+69%)21.9
+Table 5 .5Resolution efficiency metrics by vertical buffer, 50% uncertainty.ConfigurationResolutions IssuedTotal Delay, minAverage Delay per Resolution, sAll Error; No Buffer12,574(Full)3,844(Full)18.350% Error; No Buffer9,860(Base)2,866(Base)17.450% Error; 50% Buffer 13,967 (Base+42%) 5,202 (Base+82%)22.3
+			of 12 American Institute of Aeronautics and Astronautics Downloaded by NASA AMES RESEARCH CENTRE on April 17, 2013 | http://arc.aiaa.org| DOI: 10.2514/6.2012-5644
+			of 12 American Institute of Aeronautics and Astronautics Downloaded by NASA AMES RESEARCH CENTRE on April 17, 2013 | http://arc.aiaa.org| DOI: 10.2514/6.2012-5644
+			of 12 American Institute of Aeronautics and Astronautics Downloaded by NASA AMES RESEARCH CENTRE on April 17, 2013 | http://arc.aiaa.org| DOI: 10.2514/6.2012-5644
+			of 12 American Institute of Aeronautics and Astronautics Downloaded by NASA AMES RESEARCH CENTRE on April 17, 2013 | http://arc.aiaa.org| DOI: 10.2514/6.2012-5644
+			Downloaded by NASA AMES RESEARCH CENTRE on April 17, 2013 | http://arc.aiaa.org| DOI: 10.2514/6.2012-5644
+			of 12 American Institute of Aeronautics and Astronautics Downloaded by NASA AMES RESEARCH CENTRE on April 17, 2013 | http://arc.aiaa.org| DOI: 10.2514/6.2012-5644
+			of 12 American Institute of Aeronautics and Astronautics Downloaded by NASA AMES RESEARCH CENTRE on April 17, 2013 | http://arc.aiaa.org| DOI: 10.2514/6.2012-5644
+		
+		
+			
+
+				
+
+
+	
+		Assessing Trajectory Prediction Performance – Metrics Definition
+		
+			SMondoloni
+		
+		
+			SSwierstra
+		
+		
+			MPaglione
+		
+		10.1109/dasc.2005.1563347
+	
+	
+		24th Digital Avionics Systems Conference
+		Baltimore, Maryland
+		
+			IEEE
+			2005
+		
+	
+	Mondoloni, S. and Bayraktutar, I., "Impact of Factors, Conditions and Metrics on Trajectory Prediction Accuracy," 6th USA/Europe ATM R&D Seminar , Baltimore, Maryland, 2005.
+
+
+
+	
+		Field evaluation of descent advisor trajectory prediction accuracy for en-route clearance advisories
+		
+			StevenGreen
+		
+		
+			RobertVivona
+		
+		
+			MichaelGrace
+		
+		
+			Tsung-ChouFang
+		
+		10.2514/6.1998-4479
+	
+	
+		Guidance, Navigation, and Control Conference and Exhibit
+		
+			American Institute of Aeronautics and Astronautics
+			1998
+		
+	
+	Green, S. M., Vivona, R. A., and Grace, M. P., "Field Evaluation of Descent Advisor Trajectory Prediction Accuracy for En-route Clearance Advisories," AIAA Guidance, Navigation, and Control Conference, 1998.
+
+
+
+	
+		Improved Lateral Trajectory Prediction Through En Route Air-Ground Data Exchange
+		
+			DavidSchleicher
+		
+		
+			JJones
+		
+		
+			DarrenDow
+		
+		
+			RichardCoppenbarger
+		
+		10.2514/6.2002-5845
+	
+	
+		AIAA's Aircraft Technology, Integration, and Operations (ATIO) 2002 Technical Forum
+		Forum, Los Angeles, California
+		
+			American Institute of Aeronautics and Astronautics
+			2002
+		
+	
+	Schleicher, D. R., Jones, E., and Dow, D., "Improved Lateral Trajectory Prediction through En Route Air-Ground Data Exchange," AIAA Aviation Technology, Integration and Operations (ATIO) Forum, Los Angeles, California, 2002.
+
+
+
+	
+		A methodology for the performance evaluation of a conflict probe
+		
+			KarlDBilimoria
+		
+		
+			MMPaglione
+		
+		
+			HQLee
+		
+		10.2514/6.1998-4238
+	
+	
+		Guidance, Navigation, and Control Conference and Exhibit
+		
+			American Institute of Aeronautics and Astronautics
+			2004
+		
+	
+	24th International Congress of the Aeronautical Sciences
+	Bilimoria, K. D., Paglione, M. M., and Lee, H. Q., "Performance Analysis of a Conflict Probe Utilizing Only State Vector Information," 24th International Congress of the Aeronautical Sciences, 2004.
+
+
+
+	
+		Lateral Intent Error’s Impact on Aircraft Prediction
+		
+			MikePaglione
+		
+		
+			IbrahimBayraktutar
+		
+		
+			GregMcdonald
+		
+		
+			JesperBronsvoort
+		
+		10.2514/atcq.18.1.29
+	
+	
+		Air Traffic Control Quarterly
+		Air Traffic Control Quarterly
+		1064-3818
+		2472-5757
+		
+			18
+			1
+			
+			2009
+			American Institute of Aeronautics and Astronautics (AIAA)
+			Napa, California
+		
+	
+	8th USA
+	Paglione, M., McDonald, G., Bayraktutar, I., and Bronsvoort, J., "Lateral Intent Error's Impact on Aircraft Prediction," 8th USA/Europe Air Traffic Management R&D Seminar , Napa, California, 2009.
+
+
+
+	
+		Implementation and Metrics for a Trajectory Prediction Validation Methodology
+		
+			MikePaglione
+		
+		
+			RobertOaks
+		
+		10.2514/6.2007-6517
+	
+	
+		AIAA Guidance, Navigation and Control Conference and Exhibit
+		Hilton Head, South Carolina
+		
+			American Institute of Aeronautics and Astronautics
+			2007
+		
+	
+	Paglione, M. M. and Oaks, R. D., "Implementation and Metrics for a Trajectory Prediction Validation Methodology," AIAA Guidance, Navigation and Control Conference and Exhibit, Hilton Head, South Carolina, 2007.
+
+
+
+	
+		Adaptive improvement of aircraft climb performance for air traffic control applications
+		
+			GLSlater
+		
+		10.1109/isic.2002.1157831
+	
+	
+		Proceedings of the IEEE Internatinal Symposium on Intelligent Control
+		the IEEE Internatinal Symposium on Intelligent ControlVancouver, British Columbia, Canada
+		
+			IEEE
+			2002
+		
+	
+	Slater, G. L., "Adaptive Improvement of Aircraft Climb Performance for Air Traffic Control Applications," Proceedings of the IEEE International Symposium on Intelligent Control, Vancouver, British Columbia, Canada, 2002.
+
+
+
+	
+		Adaptive Improvement of Climb Performance, Master's thesis
+		
+			AAGodbole
+		
+		
+		
+			University of Cincinnati
+		
+	
+	Godbole, A. A., Adaptive Improvement of Climb Performance, Master's thesis, University of Cincinnati.
+
+
+
+	
+		Adaptive Trajectory Prediction Algorithm for Climbing Flights
+		
+			CharlesSchultz
+		
+		
+			DavidThipphavong
+		
+		
+			HeinzErzberger
+		
+		10.2514/6.2012-4931
+	
+	
+		AIAA Guidance, Navigation, and Control Conference
+		
+			American Institute of Aeronautics and Astronautics
+			2012-4931, 2012
+		
+	
+	Schultz, C., Thipphavong, D., and Erzberger, H., "Adaptive Trajectory Prediction Algorithm for Climbing Flights," AIAA Guidance, Navigation, and Control Conference, No. AIAA 2012-4931, 2012.
+
+
+
+	
+		The Effects of Speed Uncertainty on a Separation Assurance Algorithm
+		
+			ToddALauderdale
+		
+		10.2514/6.2010-9010
+	
+	
+		10th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference
+		Fort Worth, Texas
+		
+			American Institute of Aeronautics and Astronautics
+			2010
+		
+	
+	Lauderdale, T. A., "The Effects of Speed Uncertainty on a Separation Assurance Algorithm," AIAA Aviation Technology, Integration, and Operations Conference, Fort Worth, Texas, 2010.
+
+
+
+	
+		Impact of Wind Prediction Errors on an Automated Separation Assistance System
+		
+			MariaConsiglio
+		
+		
+			SherwoodHoadley
+		
+		
+			DanetteAllen
+		
+		10.2514/6.2009-7016
+	
+	
+		9th AIAA Aviation Technology, Integration, and Operations Conference (ATIO)
+		8th USA/Europe Air Traffic Management R&D Seminar
+		Napa, California
+		
+			American Institute of Aeronautics and Astronautics
+			2009
+		
+	
+	Consiglio, M., Hoadley, S., and Allen, B. D., "Estimation of Separation Buffers for Wind-Prediction Error in an Airborne Separation Assistance System," 8th USA/Europe Air Traffic Management R&D Seminar , Napa, California, 2009.
+
+
+
+	
+		Effect of conflict resolution maneuver execution delay on losses of separation
+		
+			AndrewCCone
+		
+		10.1109/dasc.2010.5655379
+	
+	
+		29th Digital Avionics Systems Conference
+		Salt Lake City, Utah
+		
+			IEEE
+			2010
+		
+	
+	Cone, A., "Effect of Conflict Resolution Maneuver Execution Delay On Losses of Separation," 29th Digital Avionics Systems Conference, Salt Lake City, Utah, 2010.
+
+
+
+	
+		Conclusion
+		
+			DMcnally
+		
+		
+			EMueller
+		
+		
+			DThipphavong
+		
+		
+			RPaielli
+		
+		
+			JCheng
+		
+		
+			CLee
+		
+		
+			SSahlman
+		
+		
+			JWalton
+		
+		10.1002/9781119006954.oth1
+	
+	
+		Aeronautical Air-Ground Data Link Communications
+		Nice, France
+		
+			John Wiley & Sons, Inc.
+			2010
+			
+		
+	
+	McNally, D., Mueller, E., Thipphavong, D., Paielli, R., Cheng, J., Lee, C., Sahlman, S., and Walton, J., "A Near-Term Concept for Trajectory-Based Operations with Air/Ground Data Link Communication," 27th Iternational Congress of the Aeronautical Sciences, Nice, France, 2010.
+
+
+
+	
+		Relative Significance of Trajectory Prediction Errors on an Automated Separation Assurance Algorithm
+		
+			TALauderdale
+		
+		
+			ACone
+		
+		
+			ABowe
+		
+		
+			2011
+			Berlin, Germany
+		
+		
+			USA/Europe Air Traffic Management Research and Development Seminar
+		
+	
+	Lauderdale, T. A., Cone, A., and Bowe, A., "Relative Significance of Trajectory Prediction Errors on an Automated Separation Assurance Algorithm," 9th USA/Europe Air Traffic Management Research and Development Seminar , Berlin, Germany, 2011.
+
+
+
+	
+		Predictability of Top of Descent Location for Operational Idle-Thrust Descents
+		
+			LaurelLStell
+		
+		10.2514/6.2010-9116
+	
+	
+		10th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference
+		Berlin, Germany
+		
+			American Institute of Aeronautics and Astronautics
+			2011
+		
+		
+			USA/Europe Air Traffic Management Research and Development Seminar
+		
+	
+	Stell, L. L., "Prediction of Top of Descent Location for Idle-thrust Descents," 9th USA/Europe Air Traffic Management Research and Development Seminar , Berlin, Germany, 2011.
+
+
+
+	
+		Build 4 of the Airspace Concept Evaluation System
+		
+			LarryMeyn
+		
+		
+			RobertWindhorst
+		
+		
+			KarlinRoth
+		
+		
+			DonaldVan Drei
+		
+		
+			GregKubat
+		
+		
+			VikramManikonda
+		
+		
+			SharleneRoney
+		
+		
+			GeorgeHunter
+		
+		
+			AlexHuang
+		
+		
+			GeorgeCouluris
+		
+		10.2514/6.2006-6110
+		No. AIAA 2006-6110
+	
+	
+		AIAA Modeling and Simulation Technologies Conference and Exhibit
+		
+			American Institute of Aeronautics and Astronautics
+			2006
+		
+	
+	Meyn, L., Windhorst, R., Roth, K., Drei, D. V., Kubat, G., Manikonda, V., Roney, S., Hunter, G., and Couluris, G., "Build 4 of the Airspace Concepts Evaluation System," AIAA Modeling and Simulation Technologies Conference and Exhibit, No. AIAA 2006-6110, 2006.
+
+
+
+	
+		User Manual for the Base of Aircraft Data (BADA) Revision 3.8
+		
+			ANuic
+		
+		
+			April 2010
+			EUROCONTROL Experimental Centre
+		
+	
+	Tech. Rep. 2010-003
+	Nuic, A., "User Manual for the Base of Aircraft Data (BADA) Revision 3.8," Tech. Rep. 2010-003, EUROCONTROL Experimental Centre, April 2010.
+
+
+
+	
+		Automated conflict resolution, arrival management, and weather avoidance for air traffic management
+		
+			HErzberger
+		
+		
+			TALauderdale
+		
+		
+			Y-CChu
+		
+		10.1177/0954410011417347
+	
+	
+		Proceedings of the Institution of Mechanical Engineers, Part G: Journal of Aerospace Engineering
+		Proceedings of the Institution of Mechanical Engineers, Part G: Journal of Aerospace Engineering
+		0954-4100
+		2041-3025
+		
+			226
+			8
+			
+			2010
+			SAGE Publications
+			Nice, France
+		
+	
+	Erzberger, H., Lauderdale, T. A., and Cheng, Y., "Automated Conflict Resolution, Arrival Management and Weather Avoidance for ATM," 27th Iternational Congress of the Aeronautical Sciences, Nice, France, 2010.
+
+
+
+	
+		Concept for Next Generation Air Traffic Control System
+		
+			HeinzErzberger
+		
+		
+			RussellAPaielli
+		
+		10.2514/atcq.10.4.355
+	
+	
+		Air Traffic Control Quarterly
+		Air Traffic Control Quarterly
+		1064-3818
+		2472-5757
+		
+			10
+			4
+			
+			2004-212828 , 2004
+			American Institute of Aeronautics and Astronautics (AIAA)
+		
+	
+	NASA/TP-
+	Erzberger, H., "Transforming the NAS: The Next Generation Air Traffic Control System," NASA/TP-2004-212828 , 2004.
+
+
+
+	
+		Trajectory Modeling Accuracy for Air Traffic Management Decision Support Tools
+		
+			SMondoloni
+		
+		
+			MPaglione
+		
+		
+			SGreen
+		
+		
+			2002
+		
+	
+	The 23th Congress of the International Council of the Aeronautical Sciences (ICAS
+	Mondoloni, S., Paglione, M., and Green, S., "Trajectory Modeling Accuracy for Air Traffic Management Decision Support Tools," The 23th Congress of the International Council of the Aeronautical Sciences (ICAS), 2002.
+
+
+
+	
+		Predictability of Top of Descent Location for Operational Idle-Thrust Descents
+		
+			LaurelLStell
+		
+		10.2514/6.2010-9116
+	
+	
+		10th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference
+		Fort Worth, Texas
+		
+			American Institute of Aeronautics and Astronautics
+			2010
+			11
+			12
+		
+	
+	Stell, L. L., "Predictability of Top of Descent Location for Operational Idle-Thrust Descent," AIAA Aviation Technology, Integration, and Operations Conference, Fort Worth, Texas, 2010. 11 of 12
+
+
+
+	
+		Downloaded by NASA AMES RESEARCH CENTRE on April 17
+		10.2514/6.2012-564422
+		
+		
+			2013
+			American Institute of Aeronautics and Astronautics
+		
+	
+	American Institute of Aeronautics and Astronautics Downloaded by NASA AMES RESEARCH CENTRE on April 17, 2013 | http://arc.aiaa.org | DOI: 10.2514/6.2012-5644 22
+
+
+
+	
+		
+			RACoppenbarger
+		
+		
+			GKanning
+		
+		
+			RSalcido
+		
+		Real-Time Data Link of Aircraft Parameters to the Center-TRACON Automation System (CTAS)," 4th USA/Europe ATM R&D Seminar
+		Santa Fe, New Mexico
+		
+			2001
+		
+	
+	Coppenbarger, R. A., Kanning, G., and Salcido, R., "Real-Time Data Link of Aircraft Parameters to the Center-TRACON Automation System (CTAS)," 4th USA/Europe ATM R&D Seminar , Santa Fe, New Mexico, 2001.
+
+
+
+	
+		Accuracy of RUC-1 and RUC-2 Wind and Aircraft Trajectory Forecasts by Comparison with ACARS Observations
+		
+			BarryESchwartz
+		
+		
+			StanleyGBenjamin
+		
+		
+			StevenMGreen
+		
+		
+			MatthewRJardin
+		
+		10.1175/1520-0434(2000)015<0313:aorarw>2.0.co;2
+	
+	
+		Weather and Forecasting
+		Wea. Forecasting
+		0882-8156
+		1520-0434
+		
+			15
+			3
+			
+			2000
+			American Meteorological Society
+		
+	
+	Schwartz, B. E., Benjamin, S. G., Green, S. M., and Jardin, M. R., "Accuracy of RUC-1 and RUC-2 Wind and Aircraft Trajectory Forecasts by Comparison with ACARS Observations," Weather and Forecasting, Vol. 15, No. 3, 2000, pp. 313-326. 12 of 12
+
+
+
+	
+		Robust Conflict Detection and Resolution around Top of Descent
+		
+			AndrewCone
+		
+		
+			AishaBowe
+		
+		
+			ToddLauderdale
+		
+		10.2514/6.2012-5644
+		
+	
+	
+		12th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference and 14th AIAA/ISSMO Multidisciplinary Analysis and Optimization Conference
+		
+			American Institute of Aeronautics and Astronautics
+			2013
+		
+	
+	American Institute of Aeronautics and Astronautics Downloaded by NASA AMES RESEARCH CENTRE on April 17, 2013 | http://arc.aiaa.org | DOI: 10.2514/6.2012-5644
+
+
+				
+			
+		
+