Aircraft Boarding Data, Validation, Analysis Dr. Michael Schultz Institute of Flight Guidance German Aerospace Center (DLR e.V.)
Aircraft Boarding
Data, Validation, Analysis
Dr. Michael Schultz
Institute of Flight Guidance
German Aerospace Center (DLR e.V.)
Introduction
aircraft trajectory, ground operations, boarding
• Aviation System Block Upgrades (ASBU)
• timeline to implement efficient flight paths
• Trajectory Based Operations (TBO)
• Global ATM Operational Concept (Doc. 9854)
• Flight & Flow Information for a Collaborative
Environment (FF-ICE Concept, Doc. 9965)
• need for a System Wide Information
Management (SWIM, Doc. 10039)
• Aircraft trajectory (Air-to-Air vs. Gate-to-Gate)
• A-CDM - milestone concept
• 4D ground trajectory - turnaround management
• critical path Boarding (± 3 min)
> ATM Seminar • Seattle > Michael Schultz • Aircraft Boarding - Data, Validation, Analysis > 27.06.2017 DLR.de/FL • Chart 2
Boarding Time
• Boarding time measured in the field
• 282 events, single aisle aircraft (B737, A320)
• boarding with 29 - 190 passengers
• Analysis of boarding rate (passenger per time)
• assumption of slow, medium, fast progress
• Q-Q plot to differentiate measurements and
expected distribution
• Classification of boarding rates
• linear progress (y = mx + n)
• average m = 4.5 pax/s, n = 2.3 min
• slow m = 1.0 pax/s, n = 12.3 min
• medium m = 1.2 pax/s, n = 8.2 min
• fast m = 2.2 pax/s, n = 3.5 min
> ATM Seminar • Seattle > Michael Schultz • Aircraft Boarding - Data, Validation, Analysis > 27.06.2017
0
5
10
15
20
0 50 100 150 200
bo
ard
ing
tim
e (
min
)
passengers
Boarding Time
DLR.de/FL • Chart 3
0
15
30
45
0 50 100 150 200
fre
qu
en
cy o
f m
easu
rem
en
t
passengers
Boarding Time Classification
fast medium slow
> ATM Seminar • Seattle > Michael Schultz • Aircraft Boarding - Data, Validation, Analysis > 27.06.2017 DLR.de/FL • Chart 4
Model Assumptions
• Problem: Boarding is owned by the passenger
• Boarding model covers operational reality of airline
boarding strategy by:
• seat load factor of current flight
• passenger arrival rate at the aircraft
• one door/ two doors configuration
• amount of hand luggage
• passenger conformance to boarding strategy
• individual passenger behavior
(speed, seat shuffle, time to store luggage)
• passenger group constellation
> ATM Seminar • Seattle > Michael Schultz • Aircraft Boarding - Data, Validation, Analysis > 27.06.2017
0
5
10
15
20
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
pas
sen
ger
pe
r m
inu
te
boarding time (min)
Boarding Progress
Scenario A (fast) Scenario B (medium) Scenario C (slow)
DLR.de/FL • Chart 5
Boarding Strategies
> ATM Seminar • Seattle > Michael Schultz • Aircraft Boarding - Data, Validation, Analysis > 27.06.2017 DLR.de/FL • Chart 6
Validation
arrival rate
• Initial assumption of 14 passengers per minute arrival, equally distributed
• Measurement indicates exponential distribution of inter-arrival times between passengers (mtime = 3.7s)
• Deboarding is 50% faster than boarding
> ATM Seminar • Seattle > Michael Schultz • Aircraft Boarding - Data, Validation, Analysis > 27.06.2017
0%
100%
200%
300%
400%
0
5
10
15
20
0 5 10 15 20
pax
pe
r m
inu
te
time (min)
Boarding - Arrival Rates
Q.25 median Q.75 m flights coveredfl
igh
ts c
ove
red
0
25
50
75
100
am
ou
nt
of
pa
ss
en
ge
rs
arrival time intervalls (s)
Field measurements Exponential Distribution
0%
100%
200%
300%
400%
0
10
20
30
40
0 2 4 6 8 10
pax
pe
r m
inu
te
time (min)
Deboarding - Outflow Rates
Q.25 median Q.75 m flights covered
flig
hts
co
vere
d
DLR.de/FL • Chart 7
Validation
hand luggage storage
• Initial assumption: triangular distribution
• 323 values from field trials, Weibull distribution
• With a = 1.7, b = 16.0s, xmin = 0s
> ATM Seminar • Seattle > Michael Schultz • Aircraft Boarding - Data, Validation, Analysis > 27.06.2017 DLR.de/FL • Chart 8
• Comparison of assumption and field measurements
0%
10%
20%
30%
40%
0-5 5-10 10-15 15-20 20-25 25-30 30-35 35-40 40-45
Pro
bab
ility
Time Categories (s)
Baggage Storage Time
Measurement
Weibull Distribution
Triangular Distribution(old model)
Validation
passenger seat shuffle
• Different kind of seat occupation pattern demands
for a specific amount of individual movements
• Assumption: Triangular distribution for single
movement (min = 1.8 s, mode = 2.4 s, max = 3.0 s)
> ATM Seminar • Seattle > Michael Schultz • Aircraft Boarding - Data, Validation, Analysis > 27.06.2017 DLR.de/FL • Chart 9
• Comparison of assumption and field measurements
• But: only 10 - 15 measurements per category
0
10
20
tim
e (
s)
seat shuffles
field measurement
simulated distribution
1 4 5 9
Validation
comparison to prior results
> ATM Seminar • Seattle > Michael Schultz • Aircraft Boarding - Data, Validation, Analysis > 27.06.2017
Boarding Strategies
boarding time (%)
random outside-in back-to-front block
1 door 100.0 80.9 110.5 96.2
calibrated 100.0 79.5 109.2 95.3
2 doors 74.2 63.8 75.3 76.2
calibrated 74.1 62.5 75.0 76.2
standard deviation (%)
1 door 7.1 5.5 7.9 6.6
calibrated 7.3 5.7 8.1 6.9
2 doors 4.6 2.9 4.8 5.3
calibrated 5.9 5.5 5.9 5.8
DLR.de/FL • Chart 10
Field Trials (I)
specific boarding strategies
• Field trials with two different scenarios (13 flights)
• A320/B738 aircraft
• standard gate position
• families are not separated
> ATM Seminar • Seattle > Michael Schultz • Aircraft Boarding - Data, Validation, Analysis > 27.06.2017
• Comparison of simulation and field measurements
• Accuracy of the boarding model: ± 5%
Boarding Strategies
Boarding time (%)
data sim. diff. Q.10 Q.25 Q.75 Q.90
random 102.6 100.0 2.6 -10.6 -5.7 6.2 11.8
airline - S1 94.8 98.7 -3.9 -11.5 -6.3 6.6 12.7
airline - S2 88.0 83.4 4.6 -8.9 -4.8 5.2 10.2
*airline - S2 80.8
DLR.de/FL • Chart 11
Field Trials (II)
specific boarding strategies
• Field measurements with 64 trials
• A320/B738 aircraft
• one door and two doors configuration
• seat load factor in three groups:
• A with 60%-80% (27 flights),
• B with 80%-90% (20 flights), and
• C with more than 90% (17 flights)
• remote, gate, and apron positions
• passenger classification: tourist, EU, Germany
• amount of pre-boarding passengers
• Problem: too many variations
> ATM Seminar • Seattle > Michael Schultz • Aircraft Boarding - Data, Validation, Analysis > 27.06.2017
• Comparison of simulation and field measurements
boarding time regarding to random reference
40%
60%
80%
100%
120%
140%data - outside-in data - block simulation
SLF 60% - 80%1door | 2doors
SLF 80% - 90%1door | 2doors
SLF 90% - 100%1door | 2doors
DLR.de/FL • Chart 12
Infrastructural Changes
side-slip seat
• Staggered seat approach
• Wider aisle enables passengers to pass each other
• Development of appropriate boarding strategy
• Up to 20% savings in boarding time (one door)
> ATM Seminar • Seattle > Michael Schultz • Aircraft Boarding - Data, Validation, Analysis > 27.06.2017 DLR.de/FL • Chart 13
> ATM Seminar • Seattle > Michael Schultz • Aircraft Boarding - Data, Validation, Analysis > 27.06.2017 DLR.de/FL • Chart 14
Summary / Outlook
• Stochastic model covers all operationally relevant aspects of aircraft boarding
• Field trials and measurements provide a solid database for validation
• Measurements in the field indicate both reliable set of input parameters and valid simulation approach (± 5%)
Next steps
• design of reliable, fast turnaround for short-haul flights
• first approach drafted
• online prediction of boarding progress for 4D ground trajectory
• model developed and used as input for deep learning approach
• dynamic seat allocation to regain control of the boarding sequence
• concept developed and already tested at Cologne-Bonn airport
> ATM Seminar • Seattle > Michael Schultz • Aircraft Boarding - Data, Validation, Analysis > 27.06.2017 DLR.de/FL • Chart 15
Aircraft Boarding
Data, Validation, Analysis
Deutsches Zentrum für Luft- und Raumfahrt e.V. (DLR)
German Aerospace Center
Institute of Flight Guidance
Dr.-Ing. Michael Schultz
Head of Department Air Transportation
Phone +49 531 295-2570