Terminal Decision Support Tool Aiman Al Gingihy Danielle Murray Sara Ataya OR/SYST 699 Fall 2013 Faculty Presentation December 13, 2013 Agenda • Introduction o Why are we here today? o Who cares? o Solving the problem • Analysis o Gap Analysis o Alternatives Analysis • • • • Recommendations Next Steps Acknowledgements Questions? 2 Introduction 3 Why Are We Here Today? • The Federal Aviation Administration (FAA) is transforming the national airspace • Next Generation Air Transportation System (NextGen) • What is NextGen? o Agency initiate to move from a ground based system to a satellite system • Why? o o o o Shorter routes Less time and fuel Less traffic delays More capacity • How is team TDST going to look into solving this? o By enabling the use of Performance Based Navigation in the Terminal Environment 4 What is Performance Based Navigation? • Defines performance requirements o More flexibility through NAS o Dynamic Management of Aircraft • How? o Through advanced procedures; RNAV/RNP Procedures o Optimize use of airspace!! Example of Complex Merge in Terminal Environment 5 Where is the Terminal Environment?? TRACON Boundary 30 to 50 miles from airport 6 Who Cares?? Stakeholders; that’s who! Terminal Controller 7 So What is the Problem? • Terminal Controllers; the primary stakeholder do not have a tool to allow aircraft to use these advanced procedures 8 How We Will Solve This Problem Methodology Approach!! 9 Gap Analysis 10 Current State Future State • Transition from conventional routes to advanced routes • Current aircraft equipage to fly these procedures is about 60% • Current utilization of advanced procedures NAS wide is 10% • Enhance safety aspects • Reduce carbon emissions • Reduce flight delays • Reduce noise impacts • Deliver a more efficient, consistent flow of traffic down to the runway 11 What is the “Formal” Gap?? 1. Inability to continue efficient arrival operations into terminal airspace 2. Lack of automation for Terminal controllers that can support mixed equipage operations RESULT: Pulling Aircraft off approaches….. Losing benefits from procedures… WHICH, Jeopardizes the investment the agency has made 12 Remedies?? 13 Actual Aircraft 2 nm A Near-term Solution: Relative Position Indicator (RPI) 25 NM Merge Point 7 nm 25 NM 2 nm 6 nm Projected Aircraft 8 nm Indicator 2 nm Approved for Public Release; Distribution Unlimited. Case # 09-0127 © 2009 The MITRE Corporation. All rights reserved. TSS Video 15 Relative Position Indicator TSS Lite + RPI Terminal Sequencing and Spacing R e m e d i e s Type of Display Aid Methodology behind Relative Absolute Absolute Calculated Relative Position Calculated Relative Position Schedule Based Position 1 (STARS) 2 (STAR, ERAM, and TBFM) 3 (STARS, ERAM, and TBFM) Mixed Equipage Mixed Equipage Mixed Equipage Aids in Complex Merges Aids in Complex Merges within Aids in Complex Merges within within Terminal Environment Terminal Environment Terminal Environment Complement TBFM System One piece of information Developed within TBFM System Display Aid Number of System Dependencies Equipage Environment Connection with TBFM inherent in TBFM User Benefits may vary based on Benefits may incrementally Precision of tools allows for controller experience improve with additional inexperienced/ experienced (inexperienced can gain information for controller controller to see benefits No Yes Yes No Yes via sequence Yes via speed and sequence greater benefit) Incorporation of Winds for Solution? Provide Trajectory Solution? Benefits Reduces controller workload Early Application of Speed Delay/Reduce Delay Vectoring Enables OPD operations Provides two more pieces of information to controllers to help sequence a/c Further reduces controller Workload Allow 95% a/c to stay on RNP curved path approach Provide streamlined arrival solution; increasing 16 predictability Alternatives Analysis 17 Methodology 18 Utility Analysis • System Level Analysis • Purpose o Help the decision maker identify which alternative will best meet the expectations of NextGen • Alternatives 1. 2. 3. TSS TSS Lite & RPI RPI • Attributes 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. Time to Mature Capability Time to Adapt/Train Capability Maintain/Increase Throughput RNP Utilization/Predictability Fuel/Mission Saving Reliability Controller Acceptability Systems Use Target Accuracy Collision 19 TSS TSS Lite + RPI RPI (Runway assignments and sequence numbers plus RPI) Time to Mature Capability 5 5 7 Time to Adapt/Train 1= 1 year or more, 5 = five months, and 10 = 1 month. 7 8 9 Maintain/Increase Throughput 7 6 5 9 7 6 8 6 5 6 7 8 9 8 6 5 5 10 9 7 6 9 10 10 1 = TRL 1, 2, 5=TRL 4, 10= TRL 9 1= 0% increase, 5= 5% increase, 10= 10% increase RNP Utilization/Predictability 1 = 50% of a/c stay on approach, 5 = 75%, 10= 100% Fuel/Emissions 1 = 5% savings on fuel/emissions, 5 = 10%, 10 = 15% Reliability 1= reliable 10% of the time, 5= reliable 75% of the time, 10= 100% of the time Controller Acceptability 1= no buy in, 5 = somewhat buy in, 10 = greatly buy in System Use 1= 0 facilities able to use capability, 5 = 35 facilities able to use capability, 10 = 70 facilities or more able to use capability Target Accuracy 1 = not accurate, 5 = somewhat accurate, 10 = very accurate Collision Risk 1 = .001% risk 5 = .0001%, 10 = .00001% risk 20 Hierarchy Decision Tree • What is needed to address the current gap and meet NextGen expectations • Weights generated elicitation – swing weights • Are attributes independent? Best System to Enable Use of PBN Time Benefits 0.36 0.20 Op Suitability 0.44 Maturity Throughput Reliability 0.56 0.34 0.22 Adapt/Train Utilization/Predicta bility Acceptability 0.44 0.43 Fuel/ Emissions 0.23 0.16 System Use 0.14 Target Accuracy 0.23 Collision 0.25 21 Value Function Attribute Value Function Maturity 0.20 Adapt/Train 0.16 Collision 0.11 Target Accuracy 0.10 Reliability 0.10 RNP Utilization/Predictability 0.09 Acceptability 0.07 Throughput 0.07 System Use 0.06 Fuel/Emissions 0.05 22 Alternative Ranking Results of MAVT 23 Cost vs. Utility Total Cost vs. Utility Fixed Cost 0.72 TSS $70M TSS Lite/RPI $12M RPI $10M RPI 0.71 0.70 0.69 TSS Utility 0.68 0.67 Reoccurring Cost TSS Lite & RPI 0.66 TSS $450K TSS Lite/RPI $350K RPI $200K 0.65 $10 $30 $50 Total Cost (Million Dollar) $70 24 Scenario Analysis Alternatives User Scenario Agency Scenario SE Scenario Benefits > Time Scenario Equal Weights for Level 1 Attributes TSS 0.69 0.73 0.71 0.72 0.69 TSS Lite + RPI 0.66 0.65 0.68 0.66 0.64 RPI 0.72 0.66 0.72 0.67 0.68 25 Sensitivity Analysis Performed “what if” analysis to study the behavior of the attribute change as the score changes Ran k Attribute Steepness/Slo pe 1 Maturity 0.198 2 Adapt/Train 0.158 3 Collision 0.112 4 Target Accuracy 0.101 5 Reliability RNP Utilization/Predictability 0.098 7 Acceptability 0.073 8 Throughput 0.068 9 System Use 0.061 10 Fuel/Emissions 0.046 6 0.085 26 Final Conclusions/Recommendations • The scores differ considerably between the ATC perspective of values and FAA headquarters' perspective. • Recommend a meeting at the decision maker level to set clear priorities on what is MOST important • If cost is not an issue, one potential recommendation is a phased approach of RPI followed by TSS o Allow agency to realize some sort of benefits in near term If Decision Maker Priority Is…. Capability Time RPI Benefits TSS Cost RPI 27 Next Steps • Deliver final report • Place emphasis that decision makers need to determine clear priorities for defined attributes • Provide alternatives analysis spreadsheet formulas to key decision makers so they can see test different scenarios 28 Acknowledgements The Terminal Decision Support team would like to thank our FAA Sponsor as well as team of Subject Matter Experts Subject Matter Expert Organization En Route Controller/TBFM SME Federal Aviation Administration Former Airline Pilot/Current FAA Manager Federal Aviation Administration Terminal Automation SME MITRE Corporation Terminal/PBN Automation SME Federal Aviation Administration Dr Lance Sherry, Executive Director of the Center for Air George Mason University Transportation Systems Research Paula Lewis, PA FAA - Assistant Administrator for Regions George Mason University and Center Operations Dr. Andrew Loerch, Associate Professor/Associate Chair SEOR Department George Mason University 29 Questions? Feedback Aiman: aalgingi@gmail.com Danielle: danielle.murray3@gmail.com Sara: ataya.sara@gmail.com Full List of References Available in Final Report 30 Backup Slides 31 Description of Criteria • • • • Time to Mature Capability: This metric represents how mature the actual capability is at this point in time. This is a quantitative metric as both tools have undergone a maturity assessment as recently as September 30, 2013. In terms of the analysis, 1 = TRL 1, 2, 5=TRL 4, 10= TRL 9. TRL speaks to the Technical Readiness Level of the Capability. We are assuming that each capability would be brought to a max level of a TRL 9 before the next stage in the lifecycle. The figure below describes each level in the TRL framework [27]. Time to Adapt/Train: This metric is based upon research and development performed to date. As RPI is incrementally more mature, this capability requires a much shorter timeframe than TSS. As such, it will take a longer time for site adaptation and training. For the purpose of this analysis, we recognize this time to be a reoccurring measure as this step will need to take place at each site. This number is quantitative based upon analysis. In terms of the analysis, 1= year or more, 5 = five months, and 10 = 1 month. Maintain/Increase Throughput: Throughput is a measure of number of landings per hour on a given runway. This metric is a qualitative relationship based upon individual data derived from both TSS and RPI simulations. In terms of the analysis, 1= 0% increase to throughput, 5= 5% increase, 10= 10% increase or more to throughput. RNP Utilization/Predictability: This metric represents a key objective – making arrivals as efficient as possible using PBN procedures. TSS provides a toolset which makes things as efficient as possible being that it is based upon an absolute schedule. RPI does provide greater efficiency compared to baseline operations but is not as efficient as TSS being that it is a relative tool. Included in this metric, is the ability of controllers to keep aircraft on RNP approaches. TSS has proved to be extremely efficient in keeping airplans on their RNP curved path approaches. While RPI has also proven effectiveness with allowing controllers to keep aircraft on PBN procedures, an evaluation of how many aircraft have been taken off their RNP curved path approach has not been conducted. Nonetheless, TSS demonstrates a clear gain in efficiency with controllers keeping aircraft on 95% of the time. In terms of the analysis, 1 = 50% of a/c stay on approach, 5 = 75% of a/c stay on 32 approach, 10= 100% of a/c stay on approach. Description of Criteria • • • • • • Fuel/Emissions: This metric is based on both qualitative and quantitative data. While an apples to apples comparison of the two capabilities cannot be performed, data and subject matter expertise opinion demonstrates that TSS will provide more fuel and emissions savings than RPI. In terms of the analysis, 1 = 5% savings on fuel/emissions, 5 = 10% savings on fuel/emissions, 10 = 15% savings on fuel/emissions. Reliability: This is the ability of the system to perform and maintain its functions in routine circumstances, as well as unexpected circumstances. This includes off nominal situations where controllers are being faced with difficult situations where the system is being tested in terms of sensitivity and flexibility. This is a qualitative assessment based upon subject matter expertise. In terms of the analysis, 1= reliable 10% of the time, 5= reliable 75% of the time, 10= 100% of the time.. Controller Acceptability: This metric represents the amount of buy in controllers have provided in regards to both capabilities. Human factors element (reduce workload, etc)This metric is based upon controller involvement in both RPI and TSS simulations and their subsequent feedback which has been documented in simulation result reports. In terms of the analysis, 1= no buy in, 5 = somewhat buy in, 10 = greatly buy in. System Use: This metric represents how many facilities will be able to use the capability. TSS is dependent on the facility having TBFM whereas RPI does not have a similar constraint. Both capabilities have a dependency on STARS. The factor of what facilities will gain benefit from either/or is also taken into account. The weights associated with this metric are qualitative based upon subject matter expertise of all factors listed above. In terms of the analysis, 1= 0 facilities able to use capability, 5 = 35 facilities able to use capability, 10 = 70 facilities or more able to use capability. Target Accuracy: In specific terms, accuracy is a degree of closeness to the actual value. For this analysis, we focus on the level of accuracy the system gives in terms to the information it displays to the controllers. The more accurate the information, the more precisely they can deliver aircraft to the runway. This is also a qualitative assessment based upon subject matter expertise. In terms of the analysis, 1 = not accurate, 5 = somewhat accurate, 10 = very accurate. Collision Risk: This metric was included to show that none of these capabilities truly have a collision risk. All of these tools are decision support tools to the controllers and controllers are ultimately responsible for separation of aircraft. In terms of the analysis, 1 = .001% risk 5 = .0001%, 10 = .00001% risk. 33