8/9/2019 fa04_lex5
1/39
1.221J/11.527J/ESD.201J Transportation SystemsFall 2004
LECTURE 5
DISPLAYS
SPEAKER: Joseph M. Sussman
MIT
8/9/2019 fa04_lex5
2/39
The Elevator ExampleElevators are simple compared to some of the
more complex transportation systems, but they
can be instructive and illustrative.
With this simple example we can gain insight intooverall system behavior that we can apply to
more complex systems.
2
8/9/2019 fa04_lex5
3/39
Elevator System60.......1
ELEVATORS A B CFigure 6.1
3
8/9/2019 fa04_lex5
4/39
Key Point 6: Transfers
Intermodal and intramodal transfers are key
determinants of service quality and cost.
Transfers between elements of the
transportation system are often inefficient.
In the elevator example, a transfer from the
walk-mode as one comes into the building,to the elevator-mode, implies some waiting
and, hence, some inefficiency.
4
8/9/2019 fa04_lex5
5/39
Key Point 8: Capacity
Capacity is a complex, multi-dimensional
system characteristic affected by:infrastructure
vehicles
technology
labor
institutional factorsoperating policyexternal factors (e.g., clean air, safety,
regulation)
5
8/9/2019 fa04_lex5
6/39
Key Point 8: Capacity (continued)
In the elevator example, We could increase the number of
elevators.
We can also change vehicletechnology. For example, we could
have larger or faster elevators.
We could have capacityimprovements as a result of control
technologies and smarter algorithmsfor dispatching.
6
8/9/2019 fa04_lex5
7/39
Key Point 9: SupplyLevel-of-service = f(volume); Transportation Supply. As
volume approaches capacity, level-of-service deteriorates
dramatically -- the hockey stick phenomenon.
LOS vs. Volume: The Hockey Stick
LOS
HOCKEYSTICK
CAPACITY
VOLUME
Figure 6.37
8/9/2019 fa04_lex5
8/39
Direct Elevator Service
8
605958
............210
Figure 7.1
8/9/2019 fa04_lex5
9/39
Another Elevator Configuration
60
50
40
30
20
10
1 2 3 4 5 6Figure 7.2
9
8/9/2019 fa04_lex5
10/39
What configuration of elevators makes the mostsense?
What is the basic trade-off here from theviewpoint of the building owner?
10
8/9/2019 fa04_lex5
11/39
Key Point 11: Infrastructure Shape
The shape of transportation infrastructure
impacts the fabric of geo-economic structures.
11
8/9/2019 fa04_lex5
12/39
Key Point 14: Cost/Level-of-ServiceTrade-offsCost/level-of-service trade-offs are afundamental tension for the transportation
provider and for the transportation
customer, as well as between them.
12
8/9/2019 fa04_lex5
13/39
Key Point 15: Demand ConsolidationConsolidation of like-demands is often used as a cost-minimizing strategy.For example, when an airline runs a hub-and-spoke operation, it is consolidating people from different origins who have common destinations into airplanes to lowercosts.
13
8/9/2019 fa04_lex5
14/39
Key Point 16: Lumpy InvestmentInvestments in capacity are often lumpy (e.g.,
infrastructure).SIDINGS
Capacity ofSingle vs. Double Track Rail Line
WESTBOUND
EASTBOUND
(OPERATIONS IN BOTH DIRECTIONS )SINGLE TRACK
DOUBLE TRACK
Figure 7.7 14
8/9/2019 fa04_lex5
15/39
Key Point 17: Capacity, Cost andLevel-of-Service
The linkages between capacity, cost and level-of-
service -- the lumpiness of investment juxtaposedwith the hockey stick level-of-service function as
volume approaches capacity -- is the central
challenge of transportation systems design.
If we underinvest in capacity, our level-of-servicemay be uncompetitive. If we overinvest, level-of-
service may be fine, but costs will be high and our
prices may not be competitive. Making this
decision, faced with lumpy investments and the
hockey stick LOS/volume relation, is difficult
indeed. This leads to the central challenge of
transportation system design.
15
8/9/2019 fa04_lex5
16/39
Our Next Concept -- PeakingVolume vs. Time of Day
-UP
VOLU MEARRIVINGAT WORK
RETURN FROM
LUNCH PEAKPACKAGEPICK
TIME OF
7 AM NOON 4 PM DAY
UP DIRECTION
VOLUME
K
LEAVING
WORGOING TO
LUNCH PEAK
7 AM NOON 4 PM TIME OFDAY
DOWN D IRECTION
16Figure 8.1
8/9/2019 fa04_lex5
17/39
How much capacity should we provide?Different Capacity Decisions
VOLUME
CAPACITY1
CAPA Y3
CAPA Y2
=
CIT
CIT
AVERAGE
DEMAND
TIME O F DAY
So what do we do? We cannot choose such a low capacity that
customer levels-of-service during peak periods are unacceptable.
At the same time, however, we cannot provide a level-of-service
such that nobody ever has to wait-- its not economical. So, capacity3 may be a good
compromise. The question of design capacity and how we
accommodate temporal peaks in demand is Key Point 18.
Figure 8.2
17
8/9/2019 fa04_lex5
18/39
Key Point 18: Peaking
Temporal peaking in demand: a fundamental issue is design capacity -- how often do we not satisfy demand?
18
8/9/2019 fa04_lex5
19/39
Our Next Concept
The volume that a transportation serviceattracts is a function of the level-of-serviceprovided to customers. If the level-of-service
deteriorates, less people will want to use the
service.
This is simply a micro-economic concept.For example, if a movie theater doubles its
price, therefore making its service less
attractive -- in this case, more expensive -
fewer people will go to that movie theater. Ifa movie theater halves its price, more people
will go.
19
8/9/2019 fa04_lex5
20/39
Key Point 19: Volume = (level-of-
service); Transportation DemandTransportation Demand: LOS vs. Volume
LOS
DEMAN D
VOLUM E
Figure 8.3
20
8/9/2019 fa04_lex5
21/39
Key Point 21: Different Time ScalesDifferent transportation systemcomponents and relevant external
systems operate and change at different
time scales, e.g.,
Short run -- operating policy
Medium run -- auto ownership
Long run -- infrastructure, land use.
21
8/9/2019 fa04_lex5
22/39
Equilibrium between Supply and
Demand Level-of-service is a function of volume. As
a facility gets congested, the level-of-service
for customers deteriorates, as discussed in
Key Point 9.
But, as level-of-service changes, demandchanges as well. Demand increases aslevel-of-service improves and decreases as
level-of-service deteriorates.
22
8/9/2019 fa04_lex5
23/39
Key Point 22: EquilibriumLEVEL OF
SERVICE
*
v*
DEMAND
SUPPLY
LOS
VOLUME
Equilibration of transportation supply anddemand for transportation service to predictvolume is a fundamental network analysismethodology.
Figure 8.423
8/9/2019 fa04_lex5
24/39
The Mechanics of Supply/DemandEquilibrium
24
8/9/2019 fa04_lex5
25/39
Iteration to Find EquilibriumLEVEL OFSERVICE
DEMAND
1
2
V*, *
S
LOS
LOS
LO S
UPPLY
v1 v3 v2 VOLUME
v
v1 -- first guess at volume
LOS1 implies demand v2
2 implies LOS2
LOS2 implies v3 , etc., until v*, LOS* are found.
Figure 8.525
8/9/2019 fa04_lex5
26/39
Changing SupplyLEVEL OFSERVICE
D
**
*
S 2
EMAND
LOS
LOS
UPPLY
SUPPLY1
V* V** VOLUME
People respond to incentives. If highwaytransportation becomes cheaper, people buy
more.
Figure 8.626
8/9/2019 fa04_lex5
27/39
Equilibrium: A Second LookLEVEL OFSERVICE
DEMAND1
DEMAND2(AFTER ACTIVITYSHIFT)
AFTER NEW
CONSTRUCTION)
*
S 2(S
1
LOS***
LOS
UPPLY
UPPLY
VOLUME
Figure 8.9 27
8/9/2019 fa04_lex5
28/39
Increased Volume: Better or Worse?
So, at a higher volume, the customers are seeingthe same level-of-service. Are we worse or
better off than we were before we started?
CLASS DISCUSSION
28
8/9/2019 fa04_lex5
29/39
Key Point 26: Stochasticity
Stochasticity -- in supply and demand -- ischaracteristic of transportation systems.
29
8/9/2019 fa04_lex5
30/39
StochasticityStochasticity in traffic volume is differentthan peaking.
VOLUME
STOCHASTIC VARIATION IN PEAK
3,000
AVERAGEAM
PEAK
TIME
Figure 9.1 30
8/9/2019 fa04_lex5
31/39
Measuring Transportation System
Performance What do your customers perceive?Our ways of measuring performance have to
relate to the ways in which our customers make
decisions about whether to use us or our
competitors.
31
8/9/2019 fa04_lex5
32/39
Performance Measures and Cost
A second concept is that performance measures have torelate to costs of operations and revenues derived from
operations of those systems.
The customer cares about system performance (e.g.,overall travel time), but
A third concept: System vs. Component Performance.The operation is often managed on a component basis.
The hope is, if we do a good job of operating thecomponents, the system as a whole will operate well.
Usually, it is a necessarycondition that component
operation be effective for system operation to be
effective, but it is very often not a sufficientcondition. We
can have a poorly meshed system in which thecomponents are operating well, but the system
performance is still poor.
32
8/9/2019 fa04_lex5
33/39
Key Point 28: Performance MeasuresPerformance measures shape transportation
operations and investment.
CLASS DISCUSSION
Some examples from transportation systems you
are familiar with.
33
8/9/2019 fa04_lex5
34/39
Key Points -- Summary (1)1. People and organizations alter behavior based on transportation
service expectations.
2. Transportation service is part of a broader system -- economic,social and political in nature.
3. Competition (or its absence) for customers by operators is acritical determinant of the availability of quality transportation
service.4. Analyzing the flow of vehicles on transportation networks, and
defining and measuring their cycle, is a basic element of
transportation systems analysis.
5. Queuing for service and for customers and storage forvehicles/freight/travelers, etc., are fundamental elements of
transportation systems.
34
8/9/2019 fa04_lex5
35/39
Key Points -- Summary (2)6. Intermodal and intramodal transfers are key determinants of
service quality and cost.7. Operating policy affects level-of-service.8. Capacity is a complex, multi-dimensional system characteristic
affected by: infrastructure vehicles
technology labor institutional factors operating policy external factors (e.g., clean air, safety, regulation)
9. Level-of-service = (volume); Transportation Supply. As volumeapproaches capacity, level-of-service deteriorates dramatically -the hockey stick phenomenon.
10. The availability ofinformation (or the lack) drives system
operations and investment and customer choices. 35
8/9/2019 fa04_lex5
36/39
Key Points -- Summary (3)11. The shape of transportation infrastructure impacts the fabric of
geo-economic structures.
12. The cost of providing a specific service, the price charged for thatservice, and the level-of-service provided may not be consistent.
13. The computation of cost for providing specific services is complex
and often ambiguous.
14. Cost/level-of-service trade-offs are a fundamental tension for thetransportation provider and for the transportation customer, as
well as between them.
15. Consolidation of like-demands is often used as a cost-minimizing
strategy.
36
8/9/2019 fa04_lex5
37/39
Key Points -- Summary (4)16. Investments in capacity are often lumpy (e.g., infrastructure).17. The linkages between capacity, cost and level-of-service -- the
lumpiness of investment juxtaposed with the hockey stick level-of-service function as volume approaches capacity -- is the
central challenge of transportation systems design.
18. Temporal peaking in demand: a fundamental issue is designcapacity -- how often do we not satisfy demand?
19. Volume = (level-of-service); Transportation Demand.20. Level-of-service is usually multi-dimensional. For analysis
purposes, we often need to reduce it to a single dimension, which
we call utility.
37
8/9/2019 fa04_lex5
38/39
Key Points -- Summary (5)21. Different transportation system components and relevant external
systems operate and change at different time scales, e.g., Short run -- operating policy Medium run -- auto ownership Long run -- infrastructure, land use
22. Equilibration of transportation supply and demand for transportationservice to predict volume is a fundamental network analysis
methodology.
23. Pricing of transportation services to entice different behavior is amechanism for lowering the negative externalities caused by
transportation users on other users and society-at-large.
24. Geographical and temporal imbalances of flow are characteristic intransportation systems.
25. Network behavior and network capacity, derived from link and nodecapacities and readjustment of flows on redundant paths, are important
elements in transportation systems analysis.
38
8/9/2019 fa04_lex5
39/39
Key Points -- Summary (6)26. Stochasticity -- in supply and demand -- is characteristic of
transportation systems.
27. The relationship among transportation, economic development,and location of activities -- the transportation/land-use
connection -- is fundamental.
28. Performance measures shape transportation operations and
investment.29. Balancing centralized control with decisions made by managers
of system components (e.g., terminals) is an important operating
challenge.
30. The integrality of vehicle/infrastructure/ control systemsinvestment, design and operating decisions is basic to
transportation systems design.
39
Recommended