Slide 1
ILLINOIS - RAILROAD ENGINEERING
Railroad Transportation
Energy Efficiency
Chris Barkan
Professor
Director - Railroad Engineering Program
Department of Civil and
Environmental Engineering
University of Illinois at
Urbana-Champaign
Slide 2
ILLINOIS - RAILROAD ENGINEERING
Acknowledgements
• Project Team
– Civil & Environmental Engineering
• Yung-Cheng Lai, J. Riley Edwards, Tristan Rickett
– Beckman Institute - Electrical & Computer Engineering
• Narendra Ahuja, John Hart, Avinash Kumar
• Research Sponsor
– BNSF Railway Technical Research and Development
• Mark Stehly, Larry Milhon, Paul Gabler
• Partial support for Yung-Cheng Lai from a CN Research
Fellowship in Railway Engineering
Slide 3
ILLINOIS - RAILROAD ENGINEERING
Presentation Outline
• Fundamentals of railroad transportation energy
efficiency
• Measuring railroad train resistance
• Intermodal train aerodynamics and train resistance
• Optimizing intermodal train loading to minimize
resistance
• Use of machine-vision to automatically monitor
intermodal train loading efficiency
Slide 4
ILLINOIS - RAILROAD ENGINEERING
Jeopardy
Whatisis
What
457?
457?
Slide 5
ILLINOIS - RAILROAD ENGINEERING
THE
NUMBER
OF
THE NUMBER OF
MILES
RAILROADS
MILES RAILROADS
CAN
CANMOVE
MOVEONE
ONETON
TON
OF
OFFREIGHT
FREIGHTON
ONONE
ONE
GALLON
FUEL
GALLONOF
OF FUEL
Slide 6
ILLINOIS - RAILROAD ENGINEERING
Energy efficiency truck vs. rail
• How does that compare to
truck transport?
Rail
Rail is 3.5 times
more efficient than truck
Truck
0
100
200
300
400
500
Freight Revenue Ton-Miles per Gallon
(AAR & FRA data)
Slide 7
ILLINOIS - RAILROAD ENGINEERING
US 20th Century was about
CONVENIENCE
The 21st must consider
EFFICIENCY as well
•
Then
– Abundant: energy, land, natural
resources, labor, dominant economy
•
Now
– Diminishing resources:
• Energy
• Air quality
• Water
• Land
– Congestion
• Need more efficient use of
transportation infrastructure
– Stronger global competition
Slide 8
ILLINOIS - RAILROAD ENGINEERING
Petroleum-derived
energy consumption
•
Transportation accounts for the
majority of petroleum energy
consumption in the U.S.
•
Of this cars, light trucks (including
SUVs) and heavy trucks account
for a large majority, followed by air
•
Rail consumes only a small
fraction of transportation
consumption (estimated to be
2.2% of total in 2009)
•
Contrast percentage consumed by
rail compared to heavy trucks,
•
Recall that rail moves about 42%
of intercity freight ton-miles
whereas trucks move about 30%
US DOE http://www1.eere.energy.gov/vehiclesandfuels/facts/2006_fcvt_fotw414.html
DOE 2003
Year
Slide 9
ILLINOIS - RAILROAD ENGINEERING
Transportation energy and work
• What are the two primary elements of transportation energy
requirements?
– Resistance
How much work is required to move something
– Energy efficiency
How efficiently energy is converted into useful work
Slide 10
ILLINOIS - RAILROAD ENGINEERING
RESISTANCE (lbs./ton)
Speed and Resistance by Transport Mode
Pipe
line
Boat
Airplanes
Trucks
Rail uniquely combines
High Speed & Capacity with
Low Resistance
SPEED (mph)
Slide 11
ILLINOIS - RAILROAD ENGINEERING
Railroad transportation efficiency
• Railroads produce transportation “output” more efficiently
than their principal competition: trucks
• What is transportation “output”
– Freight ton miles
– Passenger miles
• So why are railroads so efficient?
– Low rolling friction
– Large vehicles
– Trains
Slide 12
ILLINOIS - RAILROAD ENGINEERING
Rolling friction
FR = μRW
where:
μR < μR
FR = resistive force of rolling
friction
μR = coefficient of rolling
friction for the two
surfaces
• proportional to the
width of the wheel
μR < μR
• inversely proportional
to its radius
W = weight
FR is also inversely
proportional to rolling
surface hardness
FR < FR
Slide 13
ILLINOIS - RAILROAD ENGINEERING
Steel Wheel on Steel Rail permits low
coefficient of rolling friction (μR)
•
Steel wheel on steel rail has lower rolling friction
(μR) than rubber tire on pavement:
– Steel wheel on rail: μR = 0.001
– Truck tire on pavement: μR = 0.006 to 0.010
a dna ™e miTkciuQ
ro s serp moced )de s serp mocnU( FFIT
.erutcip siht ee s ot dedeen era
– Tire is 6 to 10 times greater than steel wheel
•
Consequently lower rolling resistance
•
Why?
•
Rubber tire on pavement
– Small effects of static friction and adhesion of
the rubber
– Major factor is the deformation of the tire while
rolling under load
– Pavement deflection also contributes
•
Steel wheel and rail experience elastic deformation
under load as well, … but much less
QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.
Slide 14
ILLINOIS - RAILROAD ENGINEERING
Operation of large, heavy vehicles is feasible because
track structure is STRONG and efficiently designed
•
Track system design and materials optimized to support very heavy loads
•
Normal North American railroad axle load is 39 tons, compared to 8.5 for trucks
Slide 15
ILLINOIS - RAILROAD ENGINEERING
Large Size of rail vehicles permits
economies of scale
• Strong railroad infrastructure allows
larger, heavier vehicles than
practical for highways
• Permits economies of scale
– Larger vehicles can transport
more weight with less resistance
per unit
– Larger engines convert energy
to work more efficiently
Slide 16
ILLINOIS - RAILROAD ENGINEERING
Trains permit two more important
economies of scale
•
The ability to operate many vehicles coupled together permits two more
substantial economies of scale
– Labor: one or two people can operate a single train with 100 to 150 cars
(or more). Considering that each railcar is roughly equivalent to three
trucks, the economies are substantial*.
– Energy: close spacing of cars in train substantially reduces aerodynamic
resistance compared to trucks. This effect is particularly important at
higher speeds (> 40mph)
Slide 17
ILLINOIS - RAILROAD ENGINEERING
TRUCK trains are possible, but would we
really want them?
Slide 18
ILLINOIS - RAILROAD ENGINEERING
Rail transport requires less land per
unit of transport
26 containers
(1 x land)
13 trailers
(4 x land)
•
Transportation output per unit of land is much
greater for rail compared to highway
•
More units per vehicle (tons or people)
•
Fewer vehicles, and they are consolidated
into trains
•
Easier to accommodate temporal differences
in directional traffic
Slide 19
ILLINOIS - RAILROAD ENGINEERING
Measurement of train resistance
•
Substantial research early in the 20th century led to the development of a
general formula for train resistance
•
Developed by W.J. Davis, often referred to as the “Davis Equation”
2
CAV
29
Ro = 1.3 +
+ bV +
w
wn
where:
Ro = resistance in lbs. per ton
w = weight per axle (= W/n)
W = weight of car
n = number of axles
b = an experimental friction coefficient for flanges, shock, etc.
A = cross-sectional area of vehicle
C = drag coefficient based on the shape of the front of the train and
other features affecting air turbulence etc.
•
The Davis Equation has been substantially updated to reflect modern
developments, but its basic form remains the same
Slide 20
ILLINOIS - RAILROAD ENGINEERING
Sources of rail vehicle resistance
C = resistances
that vary as the
square of speed
(affected by
aerodynamics of
the train)
A = resistances that vary with axle
load (includes bearing friction,
rolling friction and track
resistance)
B = resistances that vary directly
with speed (primarily flange friction
and effects of sway and oscillation)
A varies with weight ("journal" or "bearing" resistance)
B varies directly with velocity ("flange" resistance)
C varies with the square of velocity (air resistance)
The general expression for train resistance is thus:
R = AW + BV + CV2
where:
R equals total resistance
W = weight
V = velocity
Cross-section of the
vehicle, streamlining
of the front & rear,
and surface
smoothness all affect
air resistance
Slide 21
ILLINOIS - RAILROAD ENGINEERING
Resistances that vary with weight
•
Journal resistance
– Friction between journal
and bearing
•
Rolling Friction
– Friction between wheel and rail due
to “creepage” at interface
– Minute elastic deformation of wheel
and rail surfaces
•
Track Resistance
– Deformation of track structure
– Consequent “uphill” running
Slide 22
ILLINOIS - RAILROAD ENGINEERING
Resistances that vary directly with speed
•
Flange contact
– consequent friction and impacts
– rail lubrication reduces resistance
on both curved and tangent track
•
Wheel/rail interface friction
– lateral movement between wheel tread and rail head
•
Oscillation can also induce various other energy losses into:
– vehicle suspension system (sway, bounce, buff, draft)
– track structure
Slide 23
ILLINOIS - RAILROAD ENGINEERING
Resistances that vary as the square of speed
•
Streamlining of vehicles and train has a
substantial effect on air resistance as
speeds increases
•
Front and rear of train, as well as
smoothness of sides affect air resistance
•
Empty, open-top cars create turbulence
that increases drag
•
Wide spacing between cars also creates
turbulence that increases drag
•
The aerodynamics of the whole train may
be more important than that of individual
vehicles
Non-aerodynamic
Aerodynamic
Slide 24
ILLINOIS - RAILROAD ENGINEERING
Conventional Freight Train Resistance
16
Air Resistance (C)
14
Rolling Resistance (B)
12
Bearing Resistance (A)
10
8
6
C
4
2
0
10
B
A
20
30
40
50
60
70
80
Speed (mph)
Using Modified Davis Equation with coefficients from Hay pg. 79
90
100
Slide 25
ILLINOIS - RAILROAD ENGINEERING
Intermodal (TOFC) Freight Train Resistance
16
14
Air Resistance (C)
Rolling Resistance (B)
12
Bearing Resistance (A)
10
8
6
C
4
2
0
10
B
A
20
30
40
50
60
70
80
Speed (mph)
Using Modified Davis Equation with coefficients from Hay pg. 79
90
100
Slide 26
ILLINOIS - RAILROAD ENGINEERING
Pretty picture,
but aerodynamic
nightmare
TOFC silhouette
on bridge
…and empty slots even worse!
Slide 27
ILLINOIS - RAILROAD ENGINEERING
Improving intermodal train energy efficiency
•
Intermodal (IM) freight is the largest source of revenue for US freight railroads
and the fastest growing segment of their business
•
IM freight is the least energy efficient in comparison to other types of rail freight
– High speed necessary to compete with trucks
– Poor aerodynamics due to large air gaps between and beneath loads
•
Continued growth of this business indicates need to improve its energy efficiency
Slide 29
ILLINOIS - RAILROAD ENGINEERING
“Gap Length” and “Position in Train” are the two
most important factors to IM train aerodynamics
Based on the wind tunnel testing of rail equipment, three important
factors to IM train aerodynamics were identified:
1. Gap Length between the IM loads
2. Position in Train
3. Yaw Angle: wind direction (canceled out over the whole route)
Slide 30
ILLINOIS - RAILROAD ENGINEERING
BNSF Transcontinental Mainline
Ne
lso
I
D
rtla
Po
nd
ok
a
CANAD
A
vr
Ha
e
ne
n
W
A
Sn
o
M
wd
e
Glendiv
e
e
ur
La
l
M
T
o
in
t
M
da
an
n
d
er
Ab
n
W
Y
t
lle
Gi
te
O
C RN
A V
Biebe
r
sp
Ca
r
Bridg
er Jct
lia
Al
e
UT
A
Z
o
rst
Ba
w
eb
Pu
o
C
O
N
Richmo
nd
San
Diego
Topek
a
La nta
Ju
Wichit
a
K
O
S
K
Avar
d
n
xli
Te
e
Amarill
o
Kansas
City
bb
Lu
k
o
El so
Pa
er
stl s
Ru ring
Sp
M
O
Tuls
a
Oklahoma
City
tw
ee
Sw
er
at
I
N
K
Y
Paduca
h
hi
mp
Me
s
oc
T
X
Fort
Worth
n
Be
nt aum
o
Ga
lv e
st o
M
S
T
NA
L
Birmingha
m
A
R
L
A
Longvie
w
DeRidd
er
y
Ba y
t
Ci
LA to Chicago over 2,000 miles
St
r
or eat
Pe
or
i
St
Louis
Phoeni
x
nc
Ri
n
a
nv
e
l
Albuquerq
ue
Bele
n
Chicag
o
L
GaSavann
les
g a
bu
r
a
w
Io ct
J
Long Beach/
Los Angeles
De
W
II
Sioux
City
Fort
Madison
M
Los
Angeles
M
N
I
A
N
E
nc
Chicago
St
Paul
co
Lin
ln
r
n
r
ke
Ba
ld
ee
in
erl
St
g
Br
h us
e
sfi
D
h ulu
t
e
Keddi
e
es
Fr
o
rg
Fa
o
N
DS
D
ye
fa
La
e
F
L
tt
New
Orleans
Pensac
ola
Sp
elb
Sh
y
Springfi
eld
le
t
n
Ho
on us
Sumas
Om
ha a
Se
att
Pl
oo enty
d
w
Vancouv
er
Slide 31
ILLINOIS - RAILROAD ENGINEERING
Slot utilization measures the percentage of slots
on IM cars are used for loads
Maximizing slot utilization improves train energy efficiency because
it eliminates empty slots and the consequent large gaps
However, it does not account for the size of the space compared
to the size of the load
Two trains may have identical slot utilization, but different loading
patterns and aerodynamic resistance
100 %
100 %
Not the “Right Car” for the “Right Loads”
Slide 32
ILLINOIS - RAILROAD ENGINEERING
Train Energy Model and Aerodynamic Subroutine
were used to conduct efficiency analysis

General Train Resistance Equation: R = A + BV + CV2

Bearing and rolling resistance are primarily affected by weight

Aerodynamic coefficient is determined according to loading pattern

Fuel Consumption: AAR Train Energy Model (TEM)

Representative Train:

3 locomotives + 20 five-unit IM cars
1 five-unit spine car
Slide 33
ILLINOIS - RAILROAD ENGINEERING
Larger gaps resulting in a higher aerodynamic
coefficient and greater resistance
0.05
0.04
0.03
0.02
0
2
4
6
8
10
Gap Length (ft)
12
14
Slide 34
ILLINOIS - RAILROAD ENGINEERING
Equipment matching matches IM loads with railcars
so as to minimize gaps and maximize efficiency
Capacity of well and spine cars
is usually constrained by the length
of the loads
40’
40’ well
larger gap
48’ well
Equipment matching matches
IM loads so as to minimize gaps
Example:
– 40’ container in 40’ well car,
rather than a 48’ well car
– 48’ trailer in 48’ slot spine car,
rather than car with 53’ slot
48’
48’ slot
larger gap
53’ slot
Slide 35
ILLINOIS - RAILROAD ENGINEERING
Aerodynamic Coefficient of 40’ IM Loads
in Various Sized-well Cars
Aerodynamic Coefficient
(lbs/mph/mph)
5.2
5.0
4.8
4.6
0.0
40'
48'
Length of Well (ft)
53'
Slide 36
ILLINOIS - RAILROAD ENGINEERING
Total Train Resistance = Weight + Aerodynamics
55,000
Train Resistance (lbs)
40'-well
48'-well
53'-well
45,000
35,000
25,000
15,000
-10
0
10
20
30
40
Speed (mph)
50
60
70
80
Slide 37
ILLINOIS - RAILROAD ENGINEERING
Fuel Consumption over an 103-mile Route
Fuel Consumption (gal)
800
+40
+13
750
700
0.0
40’
48’
Length of Well (ft)
53’
Slide 38
ILLINOIS - RAILROAD ENGINEERING
Matching can save as much as 1 gal/mile
Train Resistance (lbs)
60,000
40
40
40
40
48' well
40' well
48
50,000
48
48' slot
53' slot
40,000
30,000
20,000
10,000
18,000
27%
Fuel Savings ~ 1 gal/mile/train
Fuel Savings = 0.13 gal/mile/train
15,000
12,000
9,000
6,000
3,000
0
-10
3%
1%
0
10
20
30
40
Speed (mph)
50
60
70
80
-10
0
10
20
30
40
Speed (mph)
50
60
70
80
Slide 39
ILLINOIS - RAILROAD ENGINEERING
Loads should be assigned not only based on
slot utilization but also “slot efficiency”
Aerodynamic Coefficient (lbs/mph/mph)
12.00
DS Containers on Well Cars
Trailers on Spine Cars
3%
9.48
10.00
9.20
36%
8.00
6.56
23%
6.00
5.05
5.90
5%
4.82
4.00
Slot Efficiency
2.00
0.00
90% Slot Utilization
100% Slot Utilization
Loading Pattern
Equipment Matching
Slide 40
ILLINOIS - RAILROAD ENGINEERING
Using “Slot Efficiency” to monitor
the loading efficiency of IM trains
IM
Loads
Length of actual load
Length of ideal load
IM Equipment
& Loading Capability
Slot Efficiency
Calculator
= 100%
Slot Efficiency
Slot efficiency represents the loading efficiency by comparing the difference
between the actual and ideal loading configuration
Every slot in each type of railcar has an ideal load that can be determined
by using the loading capability of each railcar acquired from UMLER
Slot efficiency is similar to slot utilization except that it also factors in
the energy efficiency of the load-slot combination
Right Cars for the Right Loads
Slide 41
ILLINOIS - RAILROAD ENGINEERING
Aerodynamic Loading Assignment Model (ALAM)
optimizes the aerodynamic efficiency of IM trains
Minimize Total Adjusted Gap Length
subject to:
Railcar Loading Capability
(including railcar accommodation ability, train schedule & blocking plan)
Double Stack Constraints
Weight Constraints for Every Unit
Length Constraints for Every Slot
Railcar Inventory
(car initial & #)
Optimal Loading Pattern
Blocking Plan
ALAM
Train Schedule
Load Inventory
(weight & Length)
Updated Load Inventory
Slide 42
ILLINOIS - RAILROAD ENGINEERING
Z is the total “adjusted” gap length within the train
N
Ak +1
∑



 

A1
k =1
z=
 Uk − ∑∑ y ij 1k Li  +  Uk +1 − ∑∑ y ij 1k +1Li  
 U1 − ∑∑ y ij 11Li  +
2
2 
i
j
i
j
i
j

 
 
∑∑ L × y
i
i
j
ij1k
∑∑ L × y
j
1.P1
1.P2
Where:
i
j
k
p
Ak
Uk
Li
yijkl
ij1k +1
i
i
U1
= Type of the load (40’, 48’, 53’ etc.)
= Load number within the specific type
= Unit number (1,2,…,N)
= Position in the unit (P1 or P2)
= Adjusted factor of kth gap
= Length of kth unit
= Length of ith type load (ft)
= 1 if jth load in i type was assigned to kth unit Lth position; 0 otherwise
2.P1
2.P2
U2
Slide 43
ILLINOIS - RAILROAD ENGINEERING
Aerodynamic Loading Assignment Model (ALAM)
Min
z
s.t.
∑∑ y
p
minimize total adjusted gap length
ijpk
Ripk ≤ 1
∀i , j
k
y ijpk ≤ Ripk
loading capabilities
∀i , j , p, k
40 − ∑ ∑ y ij 2k Li ≤ Φ(1 − xk )
i ∈CL
∑ ∑y
i ∈CL
j
≤ xk
ij 2 k
≤ 2 × (1 − ∑ ∑ y ij 1k )
j
∑∑∑ y
i
j
double-stack rules
i ∈CL
ijpk
w ij ≤ Wk
1)
∀k (such that δ k =
j
∀k
weight constraint
p
∑∑ y
i
1)
∀k (such that δ k =
ij 1k
j
∑∑y
i ∈TL
∀k (such that δ k =1)
ijpk
Li ≤ Qkp
∀k , p
length constraint
j
y ijpk , xk = 0, 1
Where:
Ripk
Wk
δk
xk
= Loading capability
wij = Weight of jth load in i type
= Weight limit of kth unit
Qkp = Length limit of position p in kth unit
= 1 for well-car unit; 0 otherwise Ф = A large positive number
= 1 if the top slot of kth Unit can be used; 0 otherwise
Slide 44
ILLINOIS - RAILROAD ENGINEERING
The front of the train experiences the greatest
aerodynamic resistance
Unit (k)
#
1 (locomotive)
2
3
4
5
6
7
8
9
10
100
Drag area (CDA) Adjusted factor
(ft2)
31.618
1.5449
28.801
1.4073
26.700
1.3046
25.133
1.2280 Placing loads with shorter
23.963
1.1709 gaps in the frontal position
23.091
1.1283 generates less
22.440
1.0964 aerodynamic resistance
21.954
1.0727
21.591
1.0550
21.320
1.0418
20.466
1.0000
base value
Objective:
Minimize the total “adjusted” gap length within the train
(adjusted gap length = adjusted factor x actual gap length)
Slide 45
ILLINOIS - RAILROAD ENGINEERING
Applying ALAM to the example train can
save 0.95 gallons per mile for one train
Loads: fifty
40’
, fifty
48’
, fifty
53’
Train: ten 5-unit 53-foot-slot spine cars followed by
ten 5-unit 48-foot-slot spine cars
Optimum based on ALAM = 514 (ft)
Worst case by manual assignment = 1170 (ft)
Fuel savings is 0.95 gallons/mile/train
150 loads
100 slots
Slide 46
ILLINOIS - RAILROAD ENGINEERING
Applying ALAM to four general types of IM trains
Type
C20 C40
32 184
Intl. Stack Train
Domestic Stack Train 28 88
0
0
TOFC/COFC Train
32 22
Mixed Train
C45
8
9
0
0
C48
0
17
0
6
C53
0
102
0
59
T20 T28 T40 T45 T48 T53 Total Loads
0
0
0
0
0
0
224
0
0
0
0
0
0
244
11 31 0 30 35 24
131
0
2
2
2 15 33
173
International
Stack Train
0.00 gal/mile
Domestic
Stack Train
TOFC/COFC
Train
Mixed Train
0.33 gal/mile
0.96 gal/mile
0.22 gal/mile
Slide 47
ILLINOIS - RAILROAD ENGINEERING
Using “Adjusted Slot Efficiency (ASE)” to
monitor the loading efficiency of IM trains
IM
Loads
ASE Calculator
IM Equipment
& Loading Capability
ASE = Adjusted Factor =
Slot Efficiency
Calculator
Length of actual load
Length of ideal load
Slot Efficiency
= 100%
*ASE accounts for both “Gap Length” and “Position in Train” effects
Slide 48
ILLINOIS - RAILROAD ENGINEERING
Conclusions & Recommendations
A train can be more efficiently operated if loads are assigned based
on adjusted slot efficiency
Filling empty slots with empty containers or trailers also reduces
aerodynamic resistance thereby improving energy efficiency
Uncoupling empty railcars from the end of loaded intermodal
trains when practical to reduce weight and fuel consumption
By using the aerodynamic loading assignment model (ALAM),
the potential fuel savings can be as much as 0.96 gal/mile/train
Questions?
Slide 49
ILLINOIS - RAILROAD ENGINEERING
Automated monitoring of IM loading using
machine vision
• Due to the high volume, speed
and length of their intermodal
traffic, BNSF was interested in
developing technology to
automatically monitor how well
they were loading their trains
• Sponsored research at UIUC to
develop this capability.
• Railroad Engineering Program
teamed with Computer Vision and
Robotics Lab in Beckman
Institute to develop an automated,
machine vision system for this
purpose
Q u i c k T im e ™
T IF F (U n c o m p r e s
a r e n e e d e d to s
Slide 50
ILLINOIS - RAILROAD ENGINEERING
Primary Machine Vision System Components
Incoming
Train
Image
Acquisition
System
Machine
Vision
Algorithms
Data
Analysis
System
Pertinent
Information
for
Railroads
Slide 51
ILLINOIS - RAILROAD ENGINEERING
Automated Monitoring System at BNSF’s
Logistics Park Chicago Facility
Slide 52
ILLINOIS - RAILROAD ENGINEERING
Support Equipment Layout
ANTENNA TOWER
CAMERA TOWER
EQUIPMENT
ENCLOSURE
(COMPUTER, ETC.)
LOOP
DETECTORS
Plan
Profile
9 FT
37.5 FT
Slide 53
ILLINOIS - RAILROAD ENGINEERING
Background Subtraction and Update Model
Background
Updating Module
Difference (XOR) Image
Slide 54
ILLINOIS - RAILROAD ENGINEERING
Panorama Image Generation
• The middle section of every frame in the video contributes to
the construction of the panorama
• The panorama is constructed, a patch at a time, using each
frame’s central patch and it’s pixel velocity relative to the
frame before
BNSF IM Train Panorama
Slide 56
ILLINOIS - RAILROAD ENGINEERING
Processed Train Panorama
Container (Red)
Trailer (Green)
Load Edges
in Blue
Slide 57
ILLINOIS - RAILROAD ENGINEERING
Results: Detection of Single Stack
Edges correctly
detected
Single Stack
Correctly
Detected
Edges correctly
detected
Slide 58
ILLINOIS - RAILROAD ENGINEERING
Results: Detecting Double Stacks
Double
Stack
Correctly
Detected
Slide 59
ILLINOIS - RAILROAD ENGINEERING
Results: Detection of Trailers
Trailers
Correctly
Detected
Slide 60
ILLINOIS - RAILROAD ENGINEERING
Evaluation of Loading Efficiency
Train
Monitoring
Module
Gap
Document
and
Report
Train Video
Train Panorama Reconstruction From
Input Video with Identified Gaps and
Detection of Containers/Trailers
Slide 61
ILLINOIS - RAILROAD ENGINEERING
Data Analysis System
• Edge detection and gap width data are matched with AEI equipment
type data and also with timestamp data
• The aerodynamic efficiency of each train is calculated and a score is
given to its loading pattern
• A report detailing the loading efficiency can then be developed and
sent to the appropriate personnel