ADVANCED AIR TRANSPORTATION TECHNOLOGIES DISTRIBUTED AIR/GROUND-TRAFFIC MANAGEMENT 5 th USA/Europe Air Traffic Management R&D Seminar Budapest June 2003 Autonomous Aircraft Operations and the Resolution of ‘Over-constrained’ Conflicts David J. Wing * † Dr. Karthik Krishnamurthy Richard Barhydt * Dr. Bryan Barmore * *NASA David J. Wing, NASA Langley Research Center, Hampton VA USA † Titan Corporation, Hampton VA USA /20
ADVANCED AIR TRANSPORTATION TECHNOLOGIES DISTRIBUTED AIR/GROUND-TRAFFIC MANAGEMENT 5 th USA/Europe Air Traffic Management R&D Seminar Budapest June 2003 Outline • • • DAG-TM Concept Element 5 Overview ‘Over-Constrained’ Metering Scenario and Objective Experimental Approach and Setup Results and Summary Planned Joint Simulation of Air-Ground Integration in DAG-TM David J. Wing, NASA Langley Research Center 2
ADVANCED AIR TRANSPORTATION TECHNOLOGIES DISTRIBUTED AIR/GROUND-TRAFFIC MANAGEMENT 5 th USA/Europe Air Traffic Management R&D Seminar Budapest June 2003 DAG-TM CE-5 “En Route Free Maneuvering” Concept Overview Concept Integrates: Cost management, Passenger comfort IFR trajectory management User-determined optimal trajectory David J. Wing, NASA Langley Research Center Managed (IFR) Aircraft Maneuver restrictions Terminal area Crossing restrictions Priority rules Special Use Airspace avoidance Airborne separation Hazard avoidance Fleet management IFR priority Aeronautical Operational Control Autonomous (AFR) Aircraft Mixed operations Operational constraints User flexibility $+J IFR and AFR traffic flow management Air Traffic Service Provider 3
ADVANCED AIR TRANSPORTATION TECHNOLOGIES DISTRIBUTED AIR/GROUND-TRAFFIC MANAGEMENT 5 th USA/Europe Air Traffic Management R&D Seminar Budapest June 2003 Arrival Metering of Autonomous Aircraft • Nominal procedure: – ATS Provider establishes an arrival flow schedule to meet but not exceed airport capacity, and issues terminal entry clearances consisting of ‘Required Time of Arrival’ (RTA) and crossing restrictions at assigned terminal entry fixes. – Autonomous aircraft fly self-selected trajectories to meet the constraints while maintaining separation from traffic. Separation zone Metering fix To Airport Schedule is set Metering interval • Design considerations to make constraints ‘achievable’ – What is the minimum metering interval consistent with airborne separation while converging to the fix? – How close to the fix can aircraft adjust to schedule changes? David J. Wing, NASA Langley Research Center 4
ADVANCED AIR TRANSPORTATION TECHNOLOGIES DISTRIBUTED AIR/GROUND-TRAFFIC MANAGEMENT 5 th USA/Europe Air Traffic Management R&D Seminar Budapest June 2003 Over-Constrained Situations • What about non-achievable constraints? – Constraints are often unrelated to each other, and could therefore be incompatible (e. g. , TFM, separation, weather/SUA, a/c perf. , company prefs. ) – Incompatibility could arise through failures in communication, data entry, or coordination – Example scenario of an over-constrained situation: separation loss at metering fix due to scheduling error – Result of significant system failure – “It would never happen” – Maybe, but studying extraordinary situations like this can help qualify inherent operational safety and stability characteristics of the basic concept • An assessment of robustness to such failures is important to assessing concept feasibility – CE 5 assumption: controller does not monitor autonomous aircraft separation – If monitoring is needed, this may seriously impact workload and procedures, and therefore cripple a key concept benefit – accommodating significant traffic growth David J. Wing, NASA Langley Research Center 5
ADVANCED AIR TRANSPORTATION TECHNOLOGIES DISTRIBUTED AIR/GROUND-TRAFFIC MANAGEMENT 5 th USA/Europe Air Traffic Management R&D Seminar Budapest June 2003 Experiment Research Issues and Scenario • • How robust are autonomous (distributed) operations in an over-constrained situation with no explicit coordination or centralized monitor (e. g. , ATC)? Is a priority system needed to safely ensure the separation prevails over other constraints? Aircraft A SUA Aircraft B SUA Identical time and crossing altitude assignments David J. Wing, NASA Langley Research Center 6
ADVANCED AIR TRANSPORTATION TECHNOLOGIES DISTRIBUTED AIR/GROUND-TRAFFIC MANAGEMENT 5 th USA/Europe Air Traffic Management R&D Seminar Budapest June 2003 NASA Air Traffic Operations Lab Langley Research Center Display and Glareshield Control Panels Primary Flight Display • • Navigation / Traffic Display FMS / CDU Pilot and Researcher Stations in NASA Air Traffic Operations Lab for investigating multi-aircraft operations (no ATC positions) Up to 8 interacting single-pilot stations using networked desktop simulators ADS-B modeling of message set, range, rate Aircraft simulation includes research prototype decision support tool for AFR operations David J. Wing, NASA Langley Research Center 7
ADVANCED AIR TRANSPORTATION TECHNOLOGIES DISTRIBUTED AIR/GROUND-TRAFFIC MANAGEMENT 5 th USA/Europe Air Traffic Management R&D Seminar Budapest June 2003 Research Prototype Decision Support System for Airborne Conflict Management Strategic & Tactical 0 min 5 min Strategic 10 min Resolution: Intent and State Intent Only 2 min Detection: Time to Loss of Separation Strategic FMS resolution Intent-based conflict alert David J. Wing, NASA Langley Research Center Tactical GCP resolution Conflict prevention band State-based conflict alert 8
ADVANCED AIR TRANSPORTATION TECHNOLOGIES DISTRIBUTED AIR/GROUND-TRAFFIC MANAGEMENT 5 th USA/Europe Air Traffic Management R&D Seminar Budapest June 2003 Priority Implemented Through Staggered Alerting Alert level Symbology Implication Traffic point out 1 Conflict: Action optional 2 Conflict: Action required Colors and implication based on MD-11 alerting conventions Higher Lo. S Aircraft priority* 0 Lower 10 * Modified VFR rule set used David J. Wing, NASA Langley Research Center 5 2 Minutes to Loss of Separation (Lo. S) 9
ADVANCED AIR TRANSPORTATION TECHNOLOGIES DISTRIBUTED AIR/GROUND-TRAFFIC MANAGEMENT 5 th USA/Europe Air Traffic Management R&D Seminar Budapest June 2003 Experiment Approach 1 2 a All pilots have equal priority Odd numbered pilots have priority 2 b Even numbered pilots have priority No priority rules Priority rules • Pilot instructions – (#1) maintain separation; (#2) achieve waypoint constraints within tolerances – Receive an award for performing both most consistently (incentive) – If unable #2, ‘notify’ ATC at earliest time using ‘unable’ buttons – Perform secondary task: answering trivia questions when prompted (~90 sec. ) ‘Unable’ buttons • 16 airline pilots (MD-11 or Airbus 319+) – 6 hours training – Total of 10 data runs (includes Barmore and Barhydt data runs) – Questionnaires and debriefs David J. Wing, NASA Langley Research Center 10
ADVANCED AIR TRANSPORTATION TECHNOLOGIES DISTRIBUTED AIR/GROUND-TRAFFIC MANAGEMENT 5 th USA/Europe Air Traffic Management R&D Seminar Budapest June 2003 Pilot Perspectives on Safety • Pilot rankings – 77% of the pilots rated the level of safety above neutral – 36% of the pilots found the scenario completely safe – ‘Less than neutral’ feedback “Considering the complete start-to-end scenario, including the conflicts and your resolution actions, what was the level of safety? ” David J. Wing, NASA Langley Research Center om sa ple fe tel ” y “C l” eu tra “N “N ot sa at fe all ” • Frustration at false alert • Conflicts with no resolution guidance • Unexpected simulation autoflight behavior • Misunderstanding of separation requirements • Frustration with finding conflict at RTA waypoint 11
ADVANCED AIR TRANSPORTATION TECHNOLOGIES DISTRIBUTED AIR/GROUND-TRAFFIC MANAGEMENT 5 th USA/Europe Air Traffic Management R&D Seminar Budapest June 2003 Loss of Separation (Lo. S) Case 1 First alert @ 10 minutes to conflict Alert terminated when pilot changed heading Aircraft B’s path relative to Aircraft A Closest Point of Approach: 4. 67 nm, 979 ft. Aircraft A Probable Cause of Lo. S Brief Lo. S Aircraft B Pilot climbed on seeing other aircraft pass behind him Initial descent to solve earlier conflict David J. Wing, NASA Langley Research Center Premature climb by Aircraft B after observing Aircraft A pass behind his wing. Relevant Factors Pass-behind counted as Lo. S in this simulation. Aircraft B declared ‘unable ALT’. Normally ATC would have assigned new altitude, precluding the climb. 12
ADVANCED AIR TRANSPORTATION TECHNOLOGIES DISTRIBUTED AIR/GROUND-TRAFFIC MANAGEMENT 5 th USA/Europe Air Traffic Management R&D Seminar Budapest June 2003 Loss of Separation (Lo. S) Case 2 Alert @ 5 minutes To conflict Pilot climbs to FL 355 to resolve Erroneous descent by autoflight system Closest Point of Approach : 1. 92 nm, 998 ft. First alert @ 10 minutes to conflict Aircraft A Brief Lo. S Pilot returned to FL 345, since other aircraft stayed high Aircraft B Probable Cause of Lo. S Simulation error: Unexpected descent by Aircraft A’s autoflight system below altitude limit set by pilot. Important Lesson Learned First alert @ 5 minutes to conflict Initial descent to solve a different conflict David J. Wing, NASA Langley Research Center Aircraft A’s path relative to Aircraft B • Fix the simulation errors • Link the autoflight system to the conflict management tools, and use the ‘command’ trajectory for primary alerts. 13
ADVANCED AIR TRANSPORTATION TECHNOLOGIES DISTRIBUTED AIR/GROUND-TRAFFIC MANAGEMENT 5 th USA/Europe Air Traffic Management R&D Seminar Budapest June 2003 Who Met Constraints • Over-constrained conflicts – Setup: At least 1 pilot in each pair had to miss at least 1 constraint, by experiment design Assigned Constraints: RTA <30 seconds Altitude < 500 ft Position < 2. 5 nm • Key Findings – (1) All data runs: One-third of pilots met all constraints regardless of priority rules – (2) No priority rules: RH aircraft met all constraints more often than LH aircraft – (3) Priority rules: Only privileged aircraft met all constraints – increased predictability • • But still 1/3 of these did not False alerts and unnecessary maneuvering due to receiving traffic trajectory constraints (assigned RTA and altitude) in ADS-B message rather than the commanded trajectory David J. Wing, NASA Langley Research Center 14
ADVANCED AIR TRANSPORTATION TECHNOLOGIES DISTRIBUTED AIR/GROUND-TRAFFIC MANAGEMENT 5 th USA/Europe Air Traffic Management R&D Seminar Budapest June 2003 Who First Yielded Right-of-way • Key Findings Higher Lo. S – Objective: Reduce the probability of both aircraft moving simultaneously to resolve conflicts Aircraft priority • Priority-based alerting Lower 10 5 2 Minutes to Loss of Separation (Lo. S) – No priority rules: RH and LH aircraft were equally likely to maneuver first – Priority rules: Low priority aircraft was always first to maneuver – increased predictability • Priority-based alerting successfully introduced a bias governing which aircraft would yield – Technique is suitable for complex priority rule sets David J. Wing, NASA Langley Research Center 15
ADVANCED AIR TRANSPORTATION TECHNOLOGIES DISTRIBUTED AIR/GROUND-TRAFFIC MANAGEMENT 5 th USA/Europe Air Traffic Management R&D Seminar Budapest June 2003 Experiment Summary • Autonomous (distributed) operations was robust in this simulation of an over-constrained conflict scenario (potential hazard) – The absence of centralized monitoring and/or explicit coordination were not observed to be liabilities in protecting safety • Priority rules were not observed to increase or decrease separation risks – Broadcasting state/intent information was sufficient coordination for airborne separation – Priority rules did increase predictability • Broadcast of ‘command’ trajectory would reduce undesired maneuvering due to false alerts – Broadcasting assigned constraints was a factor in 1 of 2 separation loss events, and contributed to unnecessary disruption to traffic flow • Implementation of priority system through staggered alerts was effective – Technique is suitable for complex priority rule sets, if needed David J. Wing, NASA Langley Research Center 16
ADVANCED AIR TRANSPORTATION TECHNOLOGIES DISTRIBUTED AIR/GROUND-TRAFFIC MANAGEMENT 5 th USA/Europe Air Traffic Management R&D Seminar Budapest June 2003 Planned Joint Simulation in Spring 2004 • Objectives: – To investigate the feasibility of integrated (non-segregated) Managed (IFR) and Autonomous (AFR) operations – To test new CE 5 roles/procedures for minimizing air/ground interaction in en-route and terminal transition – To test terminal arrival procedures related to CE 11 merging and in-trail spacing • Approach: Connect 2 NASA ATM research labs for real-time HITL simulation – Langley Air Traffic Operations Lab – joint sim focus is AFR operations – Ames Airspace Operations Lab – joint sim focus is ATC operations for mixed IFR/AFR traffic environment, and airborne human factors • Environment: En-route, transition, and terminal arrival traffic flows – – – ZAB/ZFW Centers and DFW TRACON ~5 subject controllers, ~20 subject pilots (+ pseudo-pilots and automated traffic) En-route / transition CD&R (including AFR/IFR conflicts) Metering of AFR/IFR arrivals (including disruptions) Merging and in-trail spacing in TRACON (DAG-TM Concept Element 11) David J. Wing, NASA Langley Research Center 17
ADVANCED AIR TRANSPORTATION TECHNOLOGIES DISTRIBUTED AIR/GROUND-TRAFFIC MANAGEMENT 5 th USA/Europe Air Traffic Management R&D Seminar Budapest June 2003 New CE-5 Roles & Procedures for Air/Ground Interaction *Key changes Autonomous Flight Rules (AFR) Aircraft · Maintains separation from all* aircraft » Extra separation margin given to IFR aircraft where feasible to minimize impact on ATS Provider » Ensures no near-term conflicts are created by maneuvering or changing intent · Selects and flies user-preferred trajectory » Maneuvers by AFR aircraft do not require clearance from ATS Provider (similar to VFR) » Trajectories selected to meet flight safety, fuel efficiency, performance limitations, and company preferences » Includes avoiding convective weather and maximizing passenger comfort » Unrestricted route & altitude except SUA’s established by ATS Provider · Conforms to TFM constraints » Adjusts path and speed to meet Required Time of Arrival (RTA) received from ATS Provider » Notifies ATS Provider if unable to meet RTA or crossing restrictions; request new assignment or alternative » Conformance required to gain terminal area access David J. Wing, NASA Langley Research Center Air Traffic Service (ATS) Provider · Separates IFR aircraft only* and monitors IFR conformance to flow/airspace constraints » Uses advanced tools and data link for enhancing IFR operations efficiency and tightening TFM tolerances · Establishes flow & airspace constraints for system-wide & local TFM » Meters AFR and IFR arrivals by assigning RTA’s (AFR) and speeds/vectors or data link trajectories (IFR) » Provides AFR aircraft an IFR clearance to enter terminal area (at which time AFR becomes IFR) · Not responsible for monitoring AFR ops* » Exception: Avoids creating near-term conflicts between AFR/IFR aircraft when maneuvering IFR aircraft » AFR aircraft treated much like VFR aircraft; relies on AFR aircraft to separate from IFR aircraft » Not responsible for ensuring AFR aircraft meet RTA 18
Скачать презентацию ADVANCED AIR TRANSPORTATION TECHNOLOGIES DISTRIBUTED AIR GROUND-TRAFFIC MANAGEMENT 5
0d3a7fc9eb63bf790d2a1f4e98d9dbc5.ppt