Detroit Deicing Decision Support Tool

1
Detroit DeicingDecision Support Tool

• Jonathan T. LeeDTS-53, Surveillance and
  Assessment Division
• CTUGApril 18, 2006

2
Acknowledgement

• NASA Langley Research Center, under contract
  DTRS57-03-C-10014
• Northwest Airlines
• System Operations Control Center, Minneapolis
• Detroit ramp tower
• Volpe Center
• Michael Geyer
• Jack Perkins (deceased)
• Anastasios (Chris) Daskalakis
• Suzanne Chen
• Jonathan Lee

3
Outline

• Overview and Problem Statement
• Model Description
• Simulation Results
• Future Work

4
Project Overview

• Started in October 2003
• 2 Labor Years
• 350K Budget
• Publications/Presentations
• 2005 Digital Avionics System Conference
• 2005 Air Traffic Control Association Annual
  Conference Proceedings
• 2005 Integrated CNS Conference Workshops
• 2004 Digital Avionics System Conference
• Patent pending

5
Deicing Overview

• Motivation
• Decrease impact of snow and ice on air carrier
  and airport operations
• Issues
• Non-linear mapping between initial conditions and
  total system time (total time in queue and in
  deicing pad)
• Deicing pad configurations
• A/C sequence
• Optimal utilization of scarce resources
• Benefits
• Maximize predictability
• Minimize operating cost
• Reduce environmental impact

6
Detroit (DTW) Map
Deicing Pad 3L
Deicing Pad 4R
7
Objective

• To assist in determining the deicing pad to send
  the outbound aircraft (A/C) to minimize the total
  time in queue and at the deicing position, given
  the current conditions and type of aircraft at
  the respective deicing pads. (Think supermarket
  checkout lines.)

Snow Type X
8
Outline

• Overview and Problem Statement
• Model Description
• Simulation Results
• Future Work

9
Deicing Queuing Model

• Queuing Model
• Parameters
• Taxi Time Statistics
• Pad Time Statistics
• Pad Configuration

Exponential/LogNormal
Gaussian
10
Deicing Queuing Model

• Queuing Model
• Inputs
• Type of A/C in deicing positions
• Type of A/C and order in queue
• Type of snow event

11
Aircraft List
B747
A330/DC-10
B757
A319/A320
DC-9
RJs(RJ-85, CRJ, Saab-340)
12
DTW Deicing Pads Configuration

• Deicing Pad 4R
• Deicing Pad 3L

13
Snow Types
Northwest Airlines, Weather Categories,
internal document.
14
Model Parameters
S.D. is only valid for Gaussian distribution.
15
DTW Deicing DST, Build 2
16
Outline

• Overview and Problem Statement
• Model Description
• Simulation Results
• Future Work

17
A320 Capable Positions

• Deicing Pad 4R
• Deicing Pad 3L

18
Outbound A/C A320
19
Snow Type A vs. C
20
A320 Capable Positions

• Deicing Pad 4R
• Deicing Pad 3L

21
B757 Capable Positions

• Deicing Pad 4R
• Deicing Pad 3L

22
Outbound A/C A320 vs. B757
23
Order of Queue
24
Outline

• Overview and Problem Statement
• Model Description
• Simulation Results
• Future Work

25
Future Directions

• Detroit Deicing DST
• Real-Time Surveillance Data
• Strategic Decision Making
• Other Locations
• Other Surface Planning Tool
• Louisville Integrated Arrival Planning Tool
• NASA Airportal Project

26
Deicing DSTTechnical Future Directions

• Real-Time Surveillance Input
• Type of A/C in position
• Elapsed time in position
• Type of A/C in queue
• Type of A/C departing from gate and runway
  assignment
• Holdover time for deiced A/C
• Runway queue
• Real-Time Prediction of System Time
• Real-Time Performance Monitor
• Strategic Decision Making Multiple Departing A/C
  and Optimal Ordering and Dispatch
• Other Locations

27
Backup Slides
28
Outline

• Overview and Problem Statement
• Model Description
• Validation Results
• Simulation Results
• Future Work

29
Modeling Overview

• Mathematical Methods
• Deicing queuing model
• Monte Carlo simulation
• Metrics
• Average total system time (queue and deicing)
• Confidence interval on total system time
• Model Elements
• Deicing process
• Spray time
• Taxi time from queue to deicing position
• Pad configuration

30
Model Parameters
Mean Time
Snow
Type A
Type B
Type C
Type D
Type E
B747
A330/DC10
B757
10.98
8.91
16.93
22.87
22.75
Pad
A319/A320
10.72
9.25
14.93
18.15
19.82
DC9
10.04
8.78
14.30
16.39
19.79
RJs
9.42
6.65
11.95
7.47
13.50
Standard Deviation
Snow
Type A
Type B
Type C
Type D
Type E
B747
A330/DC10
B757
4.75
2.70
7.95
10.89
8.40
Pad
A319/A320
4.16
3.83
7.91
9.16
8.50
DC9
4.44
2.61
7.16
6.94
7.95
RJs
3.93
1.29
6.18
0.13
4.89
31
Outline

• Overview and Problem Statement
• Model Description
• Validation Results
• Simulation Results
• Future Work

32
Pad Time Validation

• Eliminated outliers
• Pad Time
• Pad Time 60 min.
• Normality tests on log(Pad Time)
• parametric test (e.g. Chi-square test) or
• non-parametric test (e.g., Kolmogorov-Smirnov
  test)
• Pad Time under all non-widebody A/C and snow
  types passes, at least one, normality test

33
Example of LogNormal Result
After Transformation
Before Transformation
DC-9, E-snow events
34
Outline

• Overview and Problem Statement
• Model Description
• Validation Results
• Simulation Results
• Future Work

35
Widebody Capable Positions

• Deicing Pad 4R
• Deicing Pad 3L

36
Oversized Outbound A/C
37
Recent Updates

• Deicing Pad Re-Configuration
• Customize max A/C size
• Additional deicing positions
• Widebody A/C deicing (any two adjacent and open
  deicing positions at 4R deicing pad)
• Spray Time Re-Configuration
• Deicing Position Shutdown
• Model Validation

38
Configurable Deicing Pad
39
Configurable Spray Time
40
Position Shutdowns
41
Position Shutdown 6 vs. 4
6 Open Positions/Pad
4 Open Positions/Pad
42
Project Status

• Deicing tool validation on going
• Collaboration with Sensis Corp. under
  negotiation
• Deicing kick-off meeting under discussion
• Patent application file April 10, 2006