airline based en route sequencing and spacing field test … · 2011-06-03 · previous tests (e.g....
TRANSCRIPT
© 2011 The MITRE Corporation. All rights reserved.
F044-B11-015
Airline Based En Route Sequencing and Spacing Field Test Results
Peter Moertl
June 2010
ATM Seminar, Berlin
1
Approved for
public release
Case number:
11-0212
© 2011 The MITRE Corporation. All rights reserved.
F044-B11-015
Interval Management
Interval Management –Spacing (IM-S)
Interval Management –Delegated Separation (IM-
DS)
Flight Deck Based IM-DS (FIM-DS)
FIM-DS with Wake Risk Management
Ground IM-S (GIM-S)
Ground IM-DS (GIM-DS)
Flight Deck Based IM-S
(FIM-S)
Interval Management
• Interval Management (IM) describes a set of applications to improve sequencing and spacing of converging flights
– Automation supported speed management on ground and/or flight deck)
2
© 2011 The MITRE Corporation. All rights reserved.
F044-B11-015
ABESS
• Airline Based En Route Sequencing and Spacing (ABESS) is a GIM-S application for airlines
– Early version of IM, prior to ATC implementation
– Development activity has been completed
– Preconditioning of flights through the uplink of minor speed advisories
(2) Conduct
Departure Terminal Enroute Terminal Landing
Top of
Climb
Setup
Enroute
Merge
Fix
Fine-tune
Spacing
Speed
Adjustability
Period (SAP)
Earl
y S
AP
Late
SA
P
(1) ABESS Setup
(3) ABESS Termination
Pre Dep.
3
© 2011 The MITRE Corporation. All rights reserved.
F044-B11-015
Related Concepts
• 3-D PAM
– Sweet (2008) “3D-PAM”
• Attila
– Presentation at Air Traffic Control Association (ATCA), March 22, 2005, “Self-Metering Airport Arrival System Gaining Recognition,” Alexandria, VA.
• AMAN Products
– Eurocontrol (2010), Arrival Manager: Implementation Guidelines and Lessons Learned
• Main differentiation with ABESS
– ABESS is airline based
– Extended metering horizon
4
© 2011 The MITRE Corporation. All rights reserved.
F044-B11-015
ABESS Expected Benefits
• Increase in aircraft efficiency via use of early, minor speed adjustments that avoid costly, low altitude maneuvering
• Provide appropriate spacing for the initiation of new applications such as Flight-deck based Interval Management (FIM) and Optimized Profile Descents (OPDs)
• Realization of desired arrival sequences that are optimized for airline needs
• Reduce both ATC workload and radio frequency congestion
5
© 2011 The MITRE Corporation. All rights reserved.
F044-B11-015
Information Exchange
6
© 2011 The MITRE Corporation. All rights reserved.
F044-B11-015
ABESS User Interface At Airline Operations Center
7
© 2011 The MITRE Corporation. All rights reserved.
F044-B11-015
ABESS Concept Development
• Development started in 2005 with cooperation between FAA, NASA, MITRE, and UPS
• Concept development was embedded in a series of exploratory field tests with UPS:
– October 2006
– May / November 2008
– Spring 2009
– June 2010
- Objective of exploratory field tests was to develop procedures and functional requirements
- Resulted in multiple iterations of the concept
8
© 2011 The MITRE Corporation. All rights reserved.
F044-B11-015
Exploratory Test Environment
9
Field Test Participants - UPS dispatchers - Pilots - Air traffic controllers - Software engineering - Human factors researchers
“Off-the-shelf” ground tools from NASA and MITRE were initially used and then refined
Field Test Traffic: UPS West Coast aircraft to UPS hub in Louisville
© 2011 The MITRE Corporation. All rights reserved.
F044-B11-015
ABESS Aircraft Position Data Feeds
• Track data from Enhanced Traffic Management System (ETMS)
– One minute updates
– Lower (truncated) position precision
• ADS-B data
– About 1 sec updates
• Radar data
– About 12 sec updates
10
© 2011 The MITRE Corporation. All rights reserved.
F044-B11-015
June 2010 Test
• June 7 – 9 and 14 – 16, 2010
– ABESS testing during night
• Objectives
– Measure ability to predict fix crossing times over extended look-ahead time periods (e.g. up to 80 min)
– Measure accuracy of spacing problem resolution
– Determine the operational acceptability of ABESS for controllers, pilots, and AOC personal
11
© 2011 The MITRE Corporation. All rights reserved.
F044-B11-015
RESULTS
12
© 2011 The MITRE Corporation. All rights reserved.
F044-B11-015
Results Overview
• Fix crossing time predictions
• Spacing predictions
• Metering conflict detections
– True / missed detections
– Nuisance detections
– Causal factors
• Speed advisory determination
• Observed spacing at fix
13
© 2011 The MITRE Corporation. All rights reserved.
F044-B11-015
Fix Crossing Time Predictions
14
-120
-60
0
60
120
180
240
300
Pre
dic
ted
Min
us
Act
ual
Err
ors
in
Seco
nd
s
Metering Fix Prediction Errors, June 9
-120
-60
0
60
120
180
240
300
Pre
dic
ted
Min
us
Act
ual
Err
ors
in
Seco
nd
s
Metering Fix Prediction Errors, June 14
-120
-60
0
60
120
180
240
300
Pre
dic
ted
Min
us
Act
ual
Err
ors
in
Seco
nd
s
Metering Fix Prediction Errors, June 15
-120
-60
0
60
120
180
240
300
UPS843 UPS857 UPS957 UPS907 UPS905 UPS945 UPS919
Pre
dic
ted
Min
us
Act
ual
Err
ors
in
Seco
nd
s
Metering Fix Prediction Errors, June 16
Only flights are shown here for which no speed advisories were
executed as changes in speeds would have impacted the actual
crossing times.
Calculated at the time when the last aircraft entered the SAP
© 2011 The MITRE Corporation. All rights reserved.
F044-B11-015
Average Fix Crossing Time Prediction Errors
• Average prediction accuracy is similar to that found in previous tests (e.g. Moertl et al. 2009)* where average prediction errors were on average less than 30 sec for the last 100 minutes prior to the metering point
15
*Moertl, P., Arthur, W. C., Pollack, M. P., Stein, J., Zheng, L. (2009). Field Test Results of an Airline Based En Route
Sequencing and Spacing Tool. Proceedings of the 28th Digital Avionics Conference, October 25 – 29, 2009.
Signed (sec) Unsigned (sec)
Average 26 44
Median 16 27
Minimum -62 1
Maximum 261 261
• Calculated at the time when the last aircraft entered the SAP
© 2011 The MITRE Corporation. All rights reserved.
F044-B11-015
Spacing Predictions
• The correlation between actual versus predicted spacing is an indicator for the quality of spacing predictions
• Correlation was calculated when the last aircraft in the flow entered the freeze horizon
16
y = 0.9594x + 24.16R² = 0.905
0
200
400
600
800
1000
1200
0 200 400 600 800 1000 1200 1400
Pre
dic
ted
Sp
acin
g (i
n
Seco
nd
s)
Actual Spacing (in Seconds)
Predicted versus Actual Spacing, June 9 - 16th
Correlation between predicted
and actual spacing r = 0.95 over
4 days combined
© 2011 The MITRE Corporation. All rights reserved.
F044-B11-015
Metering Conflict Predictions True Detections
• Over the 4 nights of regular ABESS operations, ABESS detected 142 “metering conflict” aircraft pairs
– Aircraft were predicted to reach the metering point with less than the target spacing of 120 sec.
• There were 32 true conflict aircraft pairs
– ABESS successfully identified all of these, except 3
– 92 % of conflict aircraft detection
17
© 2011 The MITRE Corporation. All rights reserved.
F044-B11-015
Metering Conflict Predictions Nuisance Detections
• Nuisance detections should be kept to a minimum
– likely to reduce the tools usefulness and may reduce operator trust in the tool
• Calculation of nuisance detections:
• Resulted in 75 % nuisance detections
– Only 25 % of ABESS identified conflicts were “true”
18
* Successful speed advisories removed the conflict spacing that caused
the speed advisory to be given.
*
© 2011 The MITRE Corporation. All rights reserved.
F044-B11-015
Identified Contributors to Nuisance Rate
• Data instability
– Variability of groundspeed (see Moertl et al. 2009)
– Requires heuristics to correctly “smooth” data
• Wind prediction inaccuracy
– Contributions of error sources to overall nuisance rate remain to be determined
• ATC clearances
– Unforeseen trajectory changes
19
© 2011 The MITRE Corporation. All rights reserved.
F044-B11-015
Observed Variability of Speed Reports
20
Altitude
© 2011 The MITRE Corporation. All rights reserved.
F044-B11-015
Working Around Variability of Speed Reports
• To account for the variability in ground speed, the ABESS trajectory modeler uses the speed that is filed in the flight plan
• Once non-conformance between trajectories and true position reports is detected, the trajectories are updated with consideration of reported ground speed
– However, even in cases of non-conformance, the modeled speed is still strongly influenced by filed speed
– This work-around procedure delays the modeler’s ability to detect speed changes for a flight
• This is graphically shown on the next slide
21
© 2011 The MITRE Corporation. All rights reserved.
F044-B11-015
Effect of Speed Report Variability
22
The trajectory modeler
is delayed in picking up
the increase in
airspeed
Once the modeler
detects the (unexpected)
increase in speed, the fix
crossing time errors decrease
© 2011 The MITRE Corporation. All rights reserved.
F044-B11-015
Speed Advisory Determination
• Two alternatives for speed advisories
– “Global” speed advisory solutions resolved not only the conflict between two aircraft but also between all other aircraft in that stream
• The ABESS tool created global solutions that consisted of multiple speed advisories, thereby resolving each conflict
– “Non-global” speed advisory solutions resolved only individual conflict pairs
• Sometimes “global” solutions required aircraft to fly speeds outside their allowable speed envelope
– In that case, the operator selected the speed advisory that resulted the most in the desired change
• Required repetitive user actions for each conflict
23
© 2011 The MITRE Corporation. All rights reserved.
F044-B11-015
Process for Speed Advisory Determination
24
ABESS detects
spacing conflict
ABESS provides
speed advisory that
resolves conflicts
for ALL
aircraft in stream
ABESS does NOT provide
speed advisory
Operator selects
appropriate
speed advisory
Operator assesses speed
advisory feasibility based
on flight’s speed envelope
and indicated airspeed
Operator views list of possible
speed advisories and associated
predicted meter time changes
Operator selects
speed advisory
with predicted meter time
that resolves conflict
Repeat for
every conflict
pair
2. N
on
-G
lob
al
So
luti
on
s
1. G
lob
al S
olu
tio
ns
OR
Operator communicates
speed advisory
to dispatcher
Next Slide
© 2011 The MITRE Corporation. All rights reserved.
F044-B11-015
Comparison with Indicated Airspeed
• Knowledge of indicated airspeed would have allowed the ABESS operator to determine speed advisory feasibility
– No indicated airspeeds were available to the ABESS operator
• Even if indicated airspeed had been available, instantaneous readouts (e.g. all 9 min at UPS) fluctuate considerably such that they can be confusing
• Operator had two ways to estimate speed advisory feasibility
1. Request readout of indicated airspeed from dispatcher
• Workload intensive and sometimes inaccurate
2. Rely on ABESS “modeled” indicated airspeed
• Not always reliable
25
© 2011 The MITRE Corporation. All rights reserved.
F044-B11-015
Speed Advisory Coordination
• Complex process of speed advisory communication resulted in average duration of 12 minutes, ranging between 4 to 31 minutes
– Uplink path + feedback path combined
26
ABESS provides
speed advisory
ABESS operator
- dispatch supervisor
Dispatch supervisor
- dispatcher
Dispatcher - flight deck Dispatcher - flight deck Dispatcher - flight deck
Uplink Path Feedback Path
© 2011 The MITRE Corporation. All rights reserved.
F044-B11-015
Observed Spacing at Fix Slow Downs
27
-0.01
-0.03
-0.01
-0.02 -0.02
-0.02
-0.02-0.02
-0.02
-0.02
-0.02
-0.02
-0.04
-0.02
-0.01-0.03
-0.02
0
60
120
180
240
300
UPS921 UPS903 UPS957 UPS945 UPS973 UPS941 UPS905 UPS921 UPS913 UPS907 UPS857 UPS905 UPS921 UPS973 UPS797 UPS913 UPS903 UPS941
Seq
ue
nci
ng
(se
con
ds)
Aircraft Spacing-Slow Downs
Predicted
Actual
Goal
-0.01
© 2011 The MITRE Corporation. All rights reserved.
F044-B11-015
Observed Spacing at Fix Speed Ups
28
+0.01
+0.04
+0.01
+0.01+0.02
0
20
40
60
80
100
120
140
UPS913 UPS921 UPS913 UPS919
Se
qu
en
cin
g (
seco
nd
s)
Aircraft Spacing - Speed Ups
PredictedActualGoal
© 2011 The MITRE Corporation. All rights reserved.
F044-B11-015
Summary
• Exploratory field test results indicate promise and limitations for using speed advisories during extended metering
– Average (signed) fix crossing time prediction errors were 26 sec for up to 80 min prior to the metering fix
– Actual and predicted spacing correlates at r = 0.95
– 92 % conflict detection rate
– 75 % nuisance detection rate
– 17 % success rate for spacing achievement
• Further research and development for extended metering algorithms
– Development of data processing heuristics
• Utilization of indicated airspeed information
– E.g. facilitate detection and correction of wind prediction errors
– Speed advisory algorithm refinement
• Achieve acceptable balance between missed / nuisance advisories
29
© 2011 The MITRE Corporation. All rights reserved.
F044-B11-015
© 2011 The MITRE Corporation. All rights reserved.
F044-B11-015 31
Approved for Public Release: 11-0212
This is the copyright work of The MITRE Corporation and was produced for the U.S. Government
under Contract Number DTFA01-01-C-00001 and is subject to Federal Aviation Administration
Acquisition Management System Clause 3.5-13, Rights in Data-General, Alt. III and Alt. IV (Oct.
1996). No other use other than that granted to the U.S. Government, or to those acting on behalf of
the U.S. Government, under that Clause is authorized without the express written permission of The
MITRE Corporation. For further information, please contact The MITRE Corporation, Contract
Office, 7515 Colshire Drive, McLean, VA 22102, (703) 983-6000.
The contents of this material reflect the views of the author and/or the Director of the Center for
Advanced Aviation System Development, and do not necessarily reflect the views of the Federal
Aviation Administration (FAA) or Department of Transportation (DOT). Neither the FAA nor the DOT
makes any warranty or guarantee, or promise, expressed or implied, concerning the content or
accuracy of the views expressed herein.
2011 The MITRE Corporation. The Government retains a nonexclusive, royalty-free right to
publish or reproduce this document, or to allow others to do so, for “Government Purposes Only”.