FACET: Future Air Traffic Management Concepts Evaluation Tool

1
FACETFuture Air Traffic Management Concepts
Evaluation Tool
Banavar Sridhar Shon Grabbe
First Annual Workshop NAS-Wide Simulation in
Support of NEXTGEN 10 December, 2008
2
Outline

• FACET Description
• FACET uses in NEXGEN analysis
• Tube Designs
• Optimization
• Network Analysis
• Issues
• Lack of methodology
• Simulation tools
• Integration of existing tools

3
Future ATM Concepts Evaluation Tool (FACET)

• Environment for exploring advanced ATM concepts
• FACET design balances fidelity and flexibility
• Utilizes less complex models of aircraft
  performance and terminal airspace
• Enables zoom from national to regional to single
  aircraft level
• FACET architecture enables modeling of 15,000
  aircraft trajectories at the national level in a
  few seconds
• Runs on a desktop computer (Linux, Solaris, Mac
  OS X, Win XP)
• Works with existing FAA systems on an enterprise
  server
• Accessible via Web to users of Flight Explorer,
  Matlab, and Jython
• 3 Operational Modes Playback, Simulation, Hybrid
• Used for visualization, off-line analysis and
  real-time planning applications

4
Animation A Day in the Life of Air Traffic

• Smithsonians National Air Space Museum is
  using FACET in America by Air exhibit

5
FACET Displays
6
FACET Software Architecture
7
Concept of Tube Network

• Dynamic airspace configuration is a key element
  ofthe Next Generation Air Transportation System
• Flexible airspace boundaries that are dynamically
  configured
• New airspace classes such as tube airspace
• Tube network connects regions with high traffic
  volume
• Network is dynamic tailored to demand, winds,
  and weather
• Tube airspace segregated from other airspace
  classes
• Tube traffic gets benefits, e.g., better routes
  and arrival slots
• Control mode inside tube may include self
  spacing/merging
• Concept of operations is not well defined at
  present
• Initial study to expose key research issues
• Develop a common analysis method
• Define and evaluate performance metrics

8
Design of Tube Structures

• Implemented in Future ATM Concepts Evaluation
  Tool by simulating traffic above 12,000 ft
• Historical air traffic data from Aug. 24, 2005
  used in four 6-hour blocks
• Five designs based on different methods
• Jet routes
• Delaunay triangulation (Sridhar, et al.)
• Traffic density
• Hough transform (Xue, et al.)
• Network cost optimized (Gupta, et al.)

9
Hough Transform
Network Cost Optimization

• 50 great circle tubes
• Maximize use by 5 additional travel distance

• Cost of each node, link and flight travel time
  of the network optimized (67 links)

10
Performance Metrics

• Instantaneous occupancy
• Utilization and activation/deactivation trigger
• Volume occupancy
• Capacity and duration
• Number of conflicts
• Communication and workload
• Frequency of tube crossings
• Communication and workload
• Encounter angles of tube crossings
• Communication and workload

11
Number of Conflicts

• Number of conflicts with and without tubes
  (simulation 5 nmi, 1000 ft)

Worse
Nominal
Delaunay Triangles
Jet Routes
Conflict count
Cost Optimized
Traffic Density
Hough Transform
Better
Center
12
Optimization-Simulation Environment
13
Strategic Departure Control Model

• Objective function Minimize the total system
  delay
• Inputs
• Scheduled departure times and flight plans
• Sector and airport capacities
• Outputs departure delay assigned to each flight

• 2-hr Planning Horizon
• 4,500 flights, 949 airports and 987 sectors
• 600,000 variables and 650,000 constraints

Bertsimas and Stock-Patterson, 1998
14
Strategic Weather Translation

• Active area of research
• Four reduced capacity scenarios considered (0,
  20, 40, and 60) if Convective Weather
  Avoidance Model (CWAM) 60 deviation probability
  contours existed

15
Tactical Weather Translation
Avoided Convective Weather Avoidance Model (CWAM)
60 deviation contours at FL300
16
Rerouting vs Ground Holding Delays
Benefits of departure control model limited
without accounting for flow-based weather impacts
17
Model Validation
18
Additional viewgraphs
19
US Air Traffic Network

• From current flight plan structure of one day,
  Airport
• and Airspace Network (AAN) has 8000 nodes

4
5
3
2
6
7
1

251
250G Node

• AAN has 225 nodes with gt 250 links (250G) and ten
  Centers have more than ten 250G nodes each
• There are 22 1000G nodes in the system today

20
Future Traffic Scenarios

• Projected growth of tripling of passengers by
  2025 along with increased air taxis and UAVs
• Terminal Area Forecast (TAF) generated growth
  rates used to create 3X current traffic
• 3X AAN has 1443 250G nodes and all Centers have
  more than forty 250G nodes
• There are 262 1000G nodes in the future system

21
Impact of Weather

• Convective weather related delay days of more
  than 200,000 minutes are increasing
• Weather is considered a disturbing agent, either
  random or selective

• The density of 250G nodes is seen much higher in
  some regions