Download Collaborative Decision Making (CDM) Year 2000 and Beyond Growing CDM Participation: 33 members

Collaborative Decision Making
(CDM) Year 2000 and Beyond
November 1999
Robert Hoffman
Metron, Inc.
FSM Deployment to Centers
Growing CDM Participation: 33
members
CDM Members (Including Sub-carriers)
AAL
American Airlines
EGF
American Eagle Airlines
COA
Continental Airlines
BTA
Jet Link
DAL
Delta Airlines
CAA
COM
ComAir
SWR
Swiss Air
NWA
Northwest Airlines
MES
Mesaba Airlines
SWA
Southwest Airlines
TWA
Trans World Airlines
UAL
United Airlines
AWI
Air Wisconsin Airlines
BLR
Atlantic Coast Airlines
GLA
Great Lakes Aviation
UFS
Union Flights
FDX
Federal Express Airlines
GAA
Business Express
ACA
Atlantic Coast Airlines
UPS
United Parcel Service Airlines
CDM Members (Including Sub-carriers)
USA
United States Airlines
ALO
Allegheny Airlines
ASH
Air Shuttle
CDL
CHQ
JIA
PDT
Piedmont Airlines
UCA
AWE
American West Airlines
MEP
Midwest Express Airlines
LOF 1
Trans World Express
2
SKW
Skywest Airlines
Note 1: LOF is sub for UAL, USA, TWA.
Note 2: SKW is sub for UAL and DAL.
Not CDM Member, But MOA Signed
CDN
Canadian Airlines
ONT
Air Ontario
ROA
Reno Air
TRS
AirTran Airways
FSM Deployment to Centers
z Flight Schedule Monitor (FSM)
y Primary operational/monitoring tool for GDPs
y Already in use at ATCSCC and at AOCs
z Now at 36 (U.S.) ATM Locations
y ATCSCC, 20 ARTCCs, 11 TRACONs
y FAA Academy, Technical Center in Atlantic City,
Volpe Center
y NAV Canada, Canadian Centers, Toronto Center
CDM Airport Arrival Information
z CDM Strings
y confluence of Airline operational info with ETMS
y GDP-oriented: aggregate demand lists
z 1999 Enhancements
y dual strings: redundancy, more robust, stronger
firewall
y in addition: test string, development string
z Future direction
ATCSCC
AOC
ARTCC
AOC
Year 1999
FSM
AOC
ARTCC
AOC
ARTCC
Tactical GDPs
z TGDP serves as
y “short notice” GDP
y alternative to ground stops
z Advantages of TGDP over Ground Stops
y better planning for airlines
y smoother delivery of traffic into airport
y less delay (ground + air)
y merge fully with ETMS
1
Assigned Ground Delay
Comparisons - ATL
z On-going analysis
Histogram of ATC Delay
3 Hour ATL TGDP
Histogram of ATC Delay
90 Minute ATL Ground Stop
30
50
25
40
Frequency
Frequency
CDM Benefits and Performance
20
15
10
5
y Metron
y NEXTOR
y CDM Analysis Sub-group
30
20
z Benefits Assessment
10
0
0
15
30
45
60
15
75
30
45
Bin
Bin
86 flights
avg = 35
standard deviation = 19.91
variance = 396
range = 73
109 flights
avg = 26
standard deviation = 11.88
variance = 141.03
range = 53
EDCT Compliance during a GDP
60
y FFP1 Program Office
y NEXTOR Benefits Report
Compression Benefits Trend
Jan. 20, 1998 - Oct. 31, 1999
Cumulative Compression Benefits
Jan 20, 1998 - Oct 31, 1999
EDCT Compliance History
RBS++
1,000,000
8/20/1999
9/20/1999
10/20/1999
5/20/1999
7/20/1999
6/20/1999
2/20/1999
4/20/1999
1/20/1999
3/20/1999
12/20/1998
0
11/20/1998
500,000
8/20/1998
Month-Year
1,500,000
9/20/1998
Aug-99
May-99
Feb-99
Nov-98
Aug-98
Nov-97
May-98
Feb-98
Aug-97
Nov-96
May-97
Aug-96
Feb-97
Feb-96
0
May-96
10
All Airports
2,000,000
10/20/1998
20
2,500,000
5/20/1998
Late
30
Snow
Season
3,000,000
7/20/1998
40
3,500,000
6/20/1998
On-time
2/20/1998
Early
4/20/1998
50
4,000,000
1/20/1998
Percentage (%)
60
3/20/1998
70
Cumulative Minutes Reduced
4,500,000
80
Date
CDM Proven Track Record
z GDP-E Proven Track Record
y
y
y
y
y
Increased flexibility
Increased predictability
More efficient tools
Situational awareness
Community involvement
Departures and Arrival Balancing
z Departure and Arrival Balancing
y So far, CDM has been arrival oriented
y CDM approach must be preserved
y MIT PhD Thesis by W.Hall (under A.Odoni)
x integrates E.Gilbo Model (Volpe), CDM paradigms
y MIT models: N.Puget, JPClarke, L.Kang
z The New Frontiers
y Departure/Arrival Balancing
y Collaborative Routing
2
Collaborative Routing
z Apply GDP concepts/paradigms to the enroute environment
y RBS, compression algorithms
y User preferences
y Situational Awareness
z Main problems
y weather
y congestion
Why CR is harder than GDP
z Stochastic demand
y Deterministic at airport: just need good data
y But 20% of sector/route traffic is “unpredictable”
z Multi-dimensional nature
y space allocation as well as time allocation
y rolling planning horizon
y multiple resources (arrival fixes, sectors, routes)
En-Route Issues
z Rerouting around congested areas
y routes vs. sectors
y aggregate demand lists (ADLs)
y off-loading to other sectors/routes
z Ground Delays
y Use GDP-like tools to ground delay flights?
z Air Delays (assigned)
y Use of Miles-in-Trail (MIT) restrictions
y Integration with other control measures
Why CR is harder than GDP (2)
z Lack of RBS in the SKY
z Integration with other ATM initiatives
y linking of multiple centers with distinct operational
behaviors
z Data Issues
y Order of magnitude more data necessary than for
runway resources
z Capacity Metrics
y airspace harder than arrival resource
Community Solutions
z Air Transport Community Involvement
y collaborative paradigms/rationing schemes
y mixture of user rerouting and ATC delay
z Can GDP concepts carry over?
y Rationing of en-route resources
y Compression analog (T.Vossen, T.Butler, M.Ball,
UMD)
z Coded Swap Routes (CDR)
Stochastic Solutions
z Stochastic Modeling
y ADL modifier for demand (R.Hoffman, M.Ball,
UMD)
y Statistical Models for Capacity Prediction (T.Inniss,
M.Ball, UMD)
y Departure predictions (L.Kang, JPClarke, MIT)
z General Education:
y ATM/AOC use of probabilistic information
y need for simplified identification of routes
3
The Proper Role of Optimization
z What is “optimal”?
Other Solutions
z Simulation
y Balance between global efficiency and user
preferrence
y The Yardstick: How far off are we?
z Integer Programming/Network Flow Models
y J.Goodhart, (C.Yano) PhD Thesis, UCB
y S.Stock, D.Bertsimas, MIT
CR Analysis Tools (2)
y Historical playback
y , the fifth dimension
y What-if modeling
z Game Theory
y How to specify options/preferences
y Choices made w/o knowing choices of competitors
The Working Paradigm
z Post-operative Analysis Tools
y Metron/OSU POET software for en-route anlaysis
y Off-loading to other sectors/routes
z Off-loading to other sectors/routes
Academia
COE
CDM Community
y identify potential off-loads
Airlines
ATCSCC
GDP-E
CR
Problems
z Constraint identification products
y MITRE’s CCFP
CDM Academic-Industry
Coordination
Year 2000 and Beyond
COE
Academic Community
CDM Community
Academia
Airlines
GDP-E CR
ATCSCC
Background Info:
CDM http://www.metsci.com/cdm/
NEXTOR http://www.isr.umd.edu/NEXTOR/home.html
Metron Contact:
Dr. Robert Hoffman
Senior Analyst, Aviation Division
703-787-8700 [email protected] (Metron)
301-405-6622 [email protected] (UMD)
Problems
4