Planning Transport Systems
Geetam Tiwari
Transportation Research and Injury Prevention Center
Indian Institute of Technology Delhi(IITD)
New Delhi, India
How can urban planning influence
traffic safety?
Interaction at three levels:
❑Landuse planning(travel mode choice &trip distances)
❑Urban design( built environment-blocks, density-travel
mode choice)
❑Transport infrastructure(roads, PT, mode choice, risk
to captive users)
How can urban planning influence
traffic safety?
How can urban planning influence
traffic safety?
Low density single use development
Sprawl
Long trips
Higher
share of car
trips
Higher
share of
motorised
trips
How can urban planning influence
traffic safety?
Table 1
Traffic risk and urban sprawl
Authors Main Findings
(Lovegrove and Related number of crashes to amount of travel within
Sayed 2007) geographic unit
(Galster et al. 2001) Accounted for multifaceted nature of design and
density and their relationship to sprawl.
(Ewing et al. 2003)Created a sprawl index and examined the relationship
between this index and traffic crashes. The main
findings include sprawling areas are associated with
more traffic and pedestrian fatalities.
(Trowbridge et al. Constructed sprawl indices to show that sprawl is
2009;Trowbridge associated with more teen driving and longer
and McDonald 2008) ambulance arrival times. In both papers the authors
conclude that sprawl can lead to more traffic fatalities.
(Lambert and Meyer Used another index of sprawl and found that sprawl is
2006;Lucy 2003) associated with more crashes.
Landuse policy can influence the following
dimensions to influence urban traffic safety
• Density : policy that increases or maintains the
population density
• Structure : policies that favour the concentration of
employment and retail activity
• Diversity: traditional separation of landuses
• Local Design: Cities can encourage short car trips
by good design of local facilities for nonmotorized
transport
What is Transport Planning?
• Spatial and temporal spread of activities;
where we live, work, how long we travel
• Travel demand estimation and design of
facilities to meet the present and future
demand; how do we travel
• Long range plans for multi-modal
transportation systems; roads, metro, bus
system, bicycle and pedestrian facilities
Existing Landuse (2001)
PLOT MATRIX: MF35: TWBSFH
MATRIX BY O/D PAIRS
PLOT MATRIX:
mf35: twbsfh
LINKS:
all
CONSTRAINT:
mf35: twbsfh
LOWER: 0
UPPER: 5
EXCLUDE
SCALE: 1000
Volume Scale (PCU)
2000
4000 DESIRE LINES FOR
6000 BASE YEAR: TWO
WHEELER TRIPS
8000
(PCU)
10000 WINDOW:
541.18/5346.11
4675.2/8446.62
Infrastructure Improvement
8 Bridges Across River Sabarmati (3
Proposed)
8 Underpass/ flyovers (11 proposed)
Problem Analysis
Solution??
Transportation Model as a
system model
• Activity System: To estimate travel
demand
• Transportation System: Supply or service
system
• Equilibrium between supply(service) and
demand
The Urban transport System Model:
Short term impact and long term impact
Transportation System
Infrastructure
Technology
outcome Flows
Congestion
Pollution
Accidents
City systems
Landuse plans
Transport Market & Latent Demand(induced
demand)
• Changing demand
s
er
vi T1
c T2
e
Volume/ Demand V1 V2
Investment in car infrastructure leads to higher
speeds(short term), increased accidents, more cars,
congestion!!!
Data – Behaviour - Structures
Data
Behavior
STRUCTURES
20.01.2025 H. Knoflacher 16
HOW TRAFFIC PROBLEMS OF TODAY ARE PRODUCED
Urban Transport
Data
Car-data
Problems
Car user
behaviour
behaviour
STRUCTURES
STRUCTURES
For Car Traffic
Conventional solution and
promise.....
Traditional „Problem solving measure“ :the Car
infrastructure
20.01.2025 H. Knoflacher 19
Seoul
Restoration of Cheonggyecheon
Decrease of car-traffic volume : 125,000 veh/day
Before After(Sep. 2005)
Lessons from International Experience
Part I
What does not work !
URBAN HIGHWAYS 1939
• 1939 Magic Motorways (freeways) will
make congestion a thing of the past in
cities of 1960s : Designer of GM Futurama
exhibit New York world fair
• If we build enough highways “we can lick
congestion” Robert Moses 1960
Traditional „Problem solving measure“ :the Car
infrastructure
20.01.2025 H. Knoflacher 23
Example:Schlachthausgasse und A 23
..real beahviour!
60.000-
80.000
22 000
26.000 +
expected
7.000
1977 1978 1988
before ...........after A23 ......10 years afterA23
US CONGESTION GROWTH TREND
Source: The 2007 Urban Mobility Report IIT Delhi 2008
1991-1996 infrastructure investment $10billion
Average speed 15Km/h !!
Lessons from International
Experience
Part II
What works!
Central lanes(6.5m) reserved for buses and trams
Gothenburg
Cars not permitted, Bus and pedestrians only in
narrow streets
Vienna
Eisenstadt Before 10 000 cars, 6000 pedestrians per day
Eisenstadt after: 30 000 pedestrians per day plus .....
Basic Premises
• Transportation system must be viewed as a
single multi modal system( NMV, informal
transport, IPT, PT, personal motorized transport
etc.)
• Consideration of TS cannot be separated from
the consideration of the social, economic and
political system of the region.
Activity Shifts/Long term changes
Change in demand can occur
independent of transport
system(population increase,
income increase, - as a result
of exogenous events)
The implementation of TS
may cause shifts in demands(
landuse changes, land value
changes)
Systems of Models
To predict significant
impacts 5 models are
required:
1. Service Models
2. Resource models
3. Demand models
4. Equilibrium
Models
5. Activity Shift
Models
Simple equilibrium example
S= J(T, V)
or t = m+nV
where m= 10 mins,
n= .01 mins/veh/hour)
Demand function V= D(A, S)
V= a+bt (a= 500 veh/hour, b= -100 veh/hour
per min)
Equilibrium flow will satisfy service and
demand functions
Travel demand prediction
(passenger and freight)
• Life style aspirations
• Desired activity patterns
• Locational choices
• Travel choices
Hypothesis: It is possible to separate the long
run shifts in the location and scale of
economic activity from the short run behavior
of the market
Important concepts
Travel as a derived demand
Indifference curves
Behaviour at individual level
• Travel choices(destination, route, mode)
can be described in terms of service
attributes (time, distance, cost, comfort,
safety etc) S= (S1, S2, …..)
• Most often used attribute is Time, which
has many components(access time,
egress time, line haul time)
3 km car
bicycl BR metr wal
e T k
trip o
3
2.5
2
Distance, km
1.5
1
Metro Walking
0.5 Bicycling BRT
2-Wheeler/car
0
0 5 10 15 20 25 30 35
Time, minutes
IIT Delhi 2006
BR
6 km Trip T
metr
o
car
4 bicycl
Distance, km
e
3
2 Metro
Bicycling
1 BRT
2-Wheeler/car
0
0 5 10 15 20 25 30 35
Tim e, m inutes
IIT Delhi 2006
12 km Trip car metr
o
12
10
8 BR
Distance, km
T
6 S’pore average metro
trip 12 km
4 Metro
BRT
2
2-Wheeler/car
0
0 10 20 30 40 50 60
Time, minutes
IIT Delhi 2006
Decision Process
• Consumer can formulate preferences,
identify alternatives, and chooses that
maximizes his/her utility.
U= f(S,θ) ( attributes and parameters)
U= α t + βc (utility function)
Mode choice based on a simple utility function
Passenger example :
For home to work trip, a commuter has two choices, public transport(PT) or car.
Each mode is characterised by three service attributes: in-vehicle-travel-time t
in minutes; out of vehicle time x ( for car walking and parking time and for public
transport, walking and waiting time) in minutes; out of pocket cost c in Rs/100.
The utility function is given by:
U = wtt + wxx + wcc
The relative weights are : wt= -1.0, wx = -2.5, wc = -k/y; y is annual
income in Rs. and k= 5000 Rs-min/Rs-yr.
Public transport Car
In -vehicle time 30 min 40 min
Out of pocket cost Rs 8 Rs. 25
Out of vehicle 20 min 4 min
time
Passenger example cont.
a.Consider a commuter with an annual income of Rs10,000.
By how much a parking lot operator have to reduce the parking
charge of Rs.25 to attract this consumer back to car from PT
after the PT frequency has been doubled?
b.If the annual income were Rs 20,000, by how much would
PT waiting time have to be decreased to attract him from car to
PT?
c.Consider the initial conditions. At what income level will the
consumer shift from PT to car?
Freight Example
RAIL TRUCK
Cost Rs 4/kg Rs 5/kg
In vehicle time 2.5 days 2.0 days
Access time 1. 0 days 0.3 days
Shipper Relative weights
wt wx wc
A -2 -4 -1
B -3 -8 -1
C -2 -3 -4
D -2 -3 -8
Which mode will be chosen by shippers A, B, C, D respectively?
Which service attribute you could change and how to
cause a modal shift in each one of the
The four step model
I. Trip Generation
II. Trip Distribution
III. Mode Choice
IV. Network Assignment
