1. National Aeronautics and Space Administration
Next Generation Air Transportation System
Airspace Project
NASA ASAS Activities: Decision Support for Airborne Trajectory Management & Self-Separation
Robert A. Vivona
AOP Lead Engineer
L-3 Communications
Billerica, MA
Greetings. Thanks and acknowledgements of authors.
This presentation is about a recent study on pilot response delay done at NASA as part of the NextGen airspace project.
ASAS TN2.5 Workshop, Rome, Italy, November 13, 2008

2. Presentation Overview
Background
Airborne trajectory management
Autonomous Operations Planner (AOP)
Provided ASAS functions
System interface
Capabilities
AOP Experimental Performance
Concluding Remarks

3. Background: Airborne Trajectory Management
Concept
Trajectory-oriented operations
En route & transition to terminal
Aircraft self-optimization
Mixed environment
Self-separating & ground managed aircraft
Self-separating aircraft:
Flight-deck decision support equipped
Autonomous Operations Planner
“Autonomous Flight Rules”
Self-separate (traffic & area hazards)
Conform to flow constraints
Don’t generate conflicts within 5 mins
Broadcast state & intent data
FMS & air/ground data link equipped
Ground managed aircraft
Similar to today’s operations
Broadcast state (and intent) data
Terminal Airspace (Merging & Spacing) Q
En Route Airspace
Ground managed
Self-separating
ATSP Responsibilities
Manage airspace resources
Generate flow constraints
Datalink to autonomous aircraft
Control ground managed aircraft
ASAS TN2.5 Workshop, Rome, Italy, November 13, 2008

4. Autonomous Operations Planner (AOP)
Purpose: enable trajectory management
Conformance to constraints
Separation from traffic aircraft
Avoidance of special use airspace
Minimum penetration of weather hazards
Conformance to time-based flow constraints
Path optimization
Supports navigation & guidance decisions
Concept of use
Detects & alerts need to modify trajectory
Supports trajectory modifications:
Strategic & tactical maneuver alternatives
“What-if” maneuver analysis
Generates user-optimal paths
Automatically adapts to flight-crew chosen guidance mode
Tactical Maneuver Restriction Region
Area of Conflict Along Current FMS Route
Resolution Maneuver Uploaded to FMS as Mod Route
Conflict
Aircraft
Navigation Display
ASAS TN2.5 Workshop, Rome, Italy, November 13, 2008

5. AOP: Provided ASAS Functions
AOP Approach
Traffic Data
Identify Reference Aircraft
Primary Function: Uses ADS-B and TIS-B information to track and predict traffic aircraft behavior.
Track Reference Aircraft
Provide Reference Aircraft’s Trajectory
Assessing Initiation Criteria
Not supported (though could be). Not yet required for current research efforts.
Assessing Continuation Criteria
Assessing Termination Criteria
Merge
Compute Maneuvers to Merge
Not required. Separate system (PDS) at Langley provides merging & spacing capability.
Compute Merge Location
Follow
Compute Dependent Following Trajectory
Manage Interval
Compute Speeds to Achieve and Maintain Interval
Monitor Interval Conformance
Separation Maintenance
Monitor Maintenance of Separation
Secondary Function: Uses state-based CD&R for blunder protection.
Compute Guidance to Maintain Separation
Conflict Detection
Probe Trajectory For Conflicts
Primary Function: Uses both intent-based and state-based approaches.
Alert Crew to Conflicts
Conflict Resolution
Compute Vertical Maneuvers to Provide Separation
Primary Function: Uses both strategic and tactical approaches.
Compute Lateral Maneuvers to Provide Separation
Compute User-Preferred Trajectory to Provide Separation
Trajectory Optimization
Compute Traffic-Constrained User-Preferred Trajectory
Primary Function: Provided by provisional trajectory analysis.
Compute Aircraft Performance
Compute Maneuvering Flexibility
Primary Function: Looks within and outside resolution horizon.
*Source: FAA/Eurocontrol Cooperative R&D AP23
ASAS TN2.5 Workshop, Rome, Italy, November 13, 2008

6. AOP: System Interface Inputs: Flight Management Guidance settings
Traffic data
Weather data
AOP MCDU data
Flight Management
System (FMS)
Mode Control Panel (MCP)
Flight Control
Computer (FCC)
Other Sources:
GNSS Clock
ADS-B
TIS-B
FIS-B
AOP
Inputs
Autonomous
Operations
Planner (AOP)
AVIONICS DATA BUS
Multi-Function Control
& Display Unit (MCDU)
Outputs:
Conflict alerts
FMS route mods
MCP setting mods
AOP MCDU data
AOP
Outputs
Integrates with
Existing Avionics
& Displays
Navigation Display
Primary Flight Display (PFD)
ASAS TN2.5 Workshop, Rome, Italy, November 13, 2008

7. Maneuver without conflict (conflict prevention)
AOP: Capabilities
Conflict management
Probe multiple maneuvers for situation awareness
Conflict detection and resolution details
Intent-based and state-based approaches
Maneuver without conflict (conflict prevention)
Provisional (“what-if”) planning
Maneuver restriction bands
ASAS TN2.5 Workshop, Rome, Italy, November 13, 2008 7

8. Conflict Management Approach:
Probe all relevant maneuvers (trajectories) requiring evaluation by flight crew
Display all conflicts to provide complete situation awareness
Provide resolution capabilities for each maneuver
AOP can simultaneously predict/evaluate multiple ownship maneuvers
Maintaining current guidance (Commanded Prediction)
Evaluate impact of current guidance settings
“What happens if I don’t change the guidance settings?”
Reconnecting to strategic route (Planning Prediction)
Advise & evaluate maneuver to re-establish FMS active route
“How do I get back to my long-range plan?”
Stop maneuvering (State-vector projection)
Evaluate impact of maintaining current state
“What happens if I stop or don’t start/continue maneuvering?” (e.g., blunder)
Not all maneuvers always relevant
ASAS TN2.5 Workshop, Rome, Italy, November 13, 2008

9. Conflict Management Commanded Prediction
FMS Predicted
Top-of-Descent
State vector projection
VNAV PATH
LNAV/VNAV
VNAV ALT
Commanded prediction
MCP altitude limit
Planning prediction
VNAV PATH
Active Route
Altitude Constraint
Primary CD Alerting
Secondary CD Alerting
Commanded Prediction
Predicts impact of current guidance
mode settings
Initiates VNAV PATH descent
Predicts guidance switch to VNAV
ALT at MCP altitude
Primary CR impacts active guidance
Planning Prediction
Predicts impact of most strategic path
Predicts VNAV PATH descent
Ignores MCP altitude
Secondary CR impacts non-active path
State vector projection
Predicts impact of not initiating descent
State projection at cruise altitude
Point out / override CD&R
ASAS TN2.5 Workshop, Rome, Italy, November 13, 2008

10. Conflict Management Primary conflicts on the commanded prediction
Secondary conflicts on planning and blunder (state-vector) predictions
LNAV
On path
TRACK HOLD
Off path
Within capture
TRACK HOLD
Off path
Beyond capture
Commanded
Commanded
Blunder
Protection
Commanded
Last
LOS
Planning
First
LOS
ASAS TN2.5 Workshop, Rome, Italy, November 13, 2008

11. Conflict Management: Conflict Detection & Resolution
Intent-based conflict detection
Trajectory prediction
Ownship
Traffic
Trajectory prediction uncertainty buffers
Intent-based conflict resolution
Strategic & tactical
State-based CD&R
ASAS TN2.5 Workshop, Rome, Italy, November 13, 2008

12. Conflict Management: Intent-Based CD
1xN probing of ownship versus all hazards (traffic and area)
Probes ownship 4D trajectory against all traffic aircraft 4D trajectories and area hazard geometries
Configurable research parameters
Required separation zone
Independent values for AFR & IFR traffic
Look-ahead
Typically 10 minutes
Uses prediction uncertainty bounds
Independent definitions for
ownship and traffic
different maneuvers (flight modes)
Conflict = predicted loss between uncertainty regions
ownship
traffic
ASAS TN2.5 Workshop, Rome, Italy, November 13, 2008

13. Intent-Based CD: Trajectory Prediction
ADS-B
Mode Status Report (MSR)
Air Referenced Velocity Report (ARV)
Trajectory Change Report (TCR)
State Vector Report (SVR)
Target State Report (TSR)
Ownship
Generated from aircraft guidance settings
Primarily based on
Sensors: initial condition
MCP, FMS, FCC, MCDU settings
Numerical integration using internal trajectory predictor
FMS quality prediction for all guidance modes
Trajectory generated for CD application (commanded, etc.)
Traffic
Generated from ADS-B data
Primarily based on:
SVR: initial condition
TCR: represents predicted trajectory
TCP+N approach
No numerical integration
Exploring using TSR with integration for tactical guidance modes
One trajectory per traffic aircraft (used for all CD applications)
ASAS TN2.5 Workshop, Rome, Italy, November 13, 2008

14. Intent-Based CD: Trajectory Prediction Uncertainty Buffers
Vertical
Time
Altitude
Cross-track
Cross-track
Time
4D “tube” around 4D trajectory
Encapsulates prediction uncertainties unique to each segment type
Time
Lateral path
ASAS TN2.5 Workshop, Rome, Italy, November 13, 2008

15. Conflict Management: Intent-Based Conflict Resolution
Strategic
Develops FMS-compatible routes
Uploaded directly into the FMS
Crew requested (semi-automatic)
Solution:
Independent lateral and vertical maneuver options
Approach:
Resolves all conflicts and constraints, non-cooperative (priority rule-based)
Pattern-Based Genetic Algorithm
Tactical
Develops MCP setting advisories
Automatic when FMS decoupled or short time to conflict
Independent altitude, vertical rate, heading/track options
Resolves all conflicts, non-cooperative (priority rule-based)
Sweep until first conflict free setting found
ASAS TN2.5 Workshop, Rome, Italy, November 13, 2008

16. Conflict Management: Intent-Based Conflict Resolution
Nav Display
Primary Flight
Display
Track
Advisory
Vertical Rate
Altitude
Active Route: Magenta
Resolution Route: Blue
Nav Display
Strategic Intent-Based CR
Tactical Intent-Based CR
ASAS TN2.5 Workshop, Rome, Italy, November 13, 2008

17. Conflict Management: State-Based CD&R
Independent system from intent-based CD&R
Supports
Blunder protection
Override for short term conflicts
Approach
Two options
NLR (Modified Voltage Potential)
Langley (KB3D)
Resolution advisories
Independent MCP settings
track/heading, vertical rate, altitude
Automatically displayed when needed
Similar characteristics
Resolves most immediate conflict
Cooperative/Non-cooperative (configurable)
Maneuver to increase to minimum separation standard
Implicitly coordinated with other traffic maneuvers CD
Look-ahead at 5 minutes (configurable)
No area hazard detection
Does not consider uncertainty
Modified Voltage Potential

18. Maneuvering Without Conflict (Conflict Prevention)
Provisional (what-if) planning
Non-conflict generated maneuver
Probe for conflicts before execution
FMS provisional
Automatic probe of FMS MOD route
MCP provisional
Automatic probe of non-active MCP inputs
Maneuver restriction (MR) bands
Protect against unallowable maneuvers
Conflicts generated within 5 mins
Bands show unallowable MCP settings
Lateral (track/heading)
Vertical (vertical rate)
Automatically generated for:
Non-active MCP setting change
Switch to tactical guidance mode
FMS
Provisional
Conflict
Manual MOD
Route in FMS
Lateral MR
Band
MCP
Provisional
Conflict
Non-active
Change in MCP
Track Setting
(e.g., in LNAV)
ASAS TN2.5 Workshop, Rome, Italy, November 13, 2008

19. AOP Experimental Performance
Experiment 1: Lateral Only, Random Routes, All Autonomous
10X playback speed
Sustained Mean Density1
Sim. Hours
Simulated flights
Traffic conflicts
LOS2
2x 3
10x
3.45 36
881
195
6.11
1527
550
8.61
2195
1018
11.64
3000
1788
15.24 12
1302
963
17.18
1560
1256
Totals
168
10,465
5770
Lateral Strategic CR Only
NASA Air Traffic Operations Lab
All aircraft co-altitude, circle diameter 160 NM
Sustained Mean Density
1 Aircraft per 10,000 NM2
2 Loss of separation (5 NM)
3 Ref. sector ZOA31 – median density, 19 Feb 2004
Consiglio, Hoadley, Wing, Baxley: Safety Performance of Airborne Separation -
Preliminary Baseline Testing. AIAA
ASAS TN2.5 Workshop, Rome, Italy, November 13, 2008

20. AOP Experimental Performance
Experiment 2: Lateral Only, Random Routes, All Autonomous, Pilot Delay
Average
Density(*)
Average Pilot Delays (Seconds)
Flight Hours
Total Conflicts
Total LOS(**)
11.2
5X/3X
3.5
240.73
583
16.3
8X/5X
90.71
316 1
21.4
12X/7X
572.76
2307 2
Totals:
904.20
3206 3
This chart shows the effect of extreme density (~complexity) test conditions with normal pilot delays and all pilots responding on LOS
At the highest density level and under the constrained conditions of the experiment scenarios, the resolution function was able to find a strategic (lateral) solution in all but two cases.
In fact, out of 2307 conflicts generated in 572 simulated flight hours there were 2 instances where a resolution was not found. The case involved a 4-aircraft conflict in which one of them lost separation with two of the intruders
The full implementation of AOP would revert to a tactical resolution algorithm to before three minutes to go.
Removing constraints:
tactical resolutions
vertical degree of freedom to the resolution computation
Additional resolution maneuver patterns.
Lateral Strategic CR Only
(CPA < 0.02 nmi)
(*)Relative to mean and peak 1X densities of 1.8 and 3 aircraft, normalized to nmi2, at the most populated flight level of the median-density sector on 19 Feb 2004.
(**) All 3 LOS events were from high-complexity multi-aircraft conflicts. The 2 LOS events at the 21.4 density involved a 4-aircraft conflict in which one aircraft lost separation with two of the intruders.
Consiglio, Hoadley, Wing, Baxley, Allen: Impact of Pilot Delay and Non-Responsiveness on the
Safety Performance of Airborne Separation. ATIO 2008.
ASAS TN2.5 Workshop, Rome, Italy, November 13, 2008

21. AOP Performance: Batch Experiment 2
Experiment 2: Lateral Only, Random Routes, All Autonomous, Pilot Delay
Number of intruder aircraft per conflict resolutions.
As traffic density increases, so does the number of multi-aircraft conflicts, which reflects increased traffic complexity
Understanding the effect of response delay and traffic density:
Both graphs correspond to pilot delays of 240 sec.
AOP Intent-Based CD&R Studied
Under Highly Complex Traffic Scenarios
ASAS TN2.5 Workshop, Rome, Italy, November 13, 2008

22. Concluding Remarks: Current NASA Efforts
SPAS (Safety Performance of Airborne Separation)
PI: María Consiglio
Studying effects of major error sources (e.g., wind error) on safety performance
SPCASO (Safety & Performance Characterization of Airborne Self-Separation Operations)
Prediction uncertainty
PI: Danette Allen
Studying use of trajectory prediction uncertainty bounds to mitigate prediction error
Mixed operations
PI: David Wing
Studying impact of mixed AFR and IFR traffic
Investigating approaches to mitigating traffic complexity using AOP
ADS-B Performance
PI: Will Johnson
Studying impacts of ADS-B range, interference and limited intent
ASAS TN2.5 Workshop, Rome, Italy, November 13, 2008

23. Concluding Remarks: AOP References
System Concept
Ballin, M.G., Sharma, V., Vivona, R.A., Johnson, E.J., and Ramiscal, E.: “A Flight Deck Decision Support Tool for Autonomous Airborne Operations,” AIAA Guidance, Navigation, and Control Conference, AIAA , August 2002.
CD&R
Vivona, R., Karr, D., and Roscoe, D., “Pattern-Based Genetic Algorithm for Airborne Conflict Resolution”, AIAA Guidance, Navigation and Control Conference, AIAA , August 2006.
Karr, D., and Vivona, R., “Conflict Detection Using Variable Four-Dimensional Uncertainty Bounds to Control Missed Alerts,” AIAA Guidance, Navigation and Control Conference, AIAA , August 2006.
Mondoloni, S., Ballin, M., and Palmer, M.: “Airborne Conflict Resolution for Flow-Restricted Transition Airspace,” 3rd AIAA Aviation Technology, Integration and Operations (ATIO) Conference, AIAA , November 2003.
Experiments
Consiglio, M., Hoadley, S., Wing, D., and Baxley, B., “Safety Performance of Airborne Separation: Preliminary Baseline Testing,” 7th AIAA Aviation Technology, Integration and Operations (ATIO) Conference, AIAA , September 2007.
Consiglio, M., Hoadley, S., Wing, D., Baxley, B., and Allen, D., “Impact of Pilot Delay and Non-Responsiveness on the Safety Performance of Airborne Separation,” 8th AIAA Aviation Technology, Integration and Operations (ATIO) Conference, AIAA , September 2008.
ASAS TN2.5 Workshop, Rome, Italy, November 13, 2008

24. Demonstration Demo ~10x traffic density 10x playback speed
Long range display
No display filtering
Lateral Intent-Based CR
Navigation
Display
ASAS TN2.5 Workshop, Rome, Italy, November 13, 2008