Jingqiu Guo

1
PATREC Research Forum 2 Oct 2008
Modelling User Response to Innovative Traveller
Information Systems
Jingqiu Guo PhD Candidate UWA Business School,
University of Western Australia
2
Outline
Introduction Evolution of Intelligent
Transport System (ITS) and Advanced Travel
Information System (ATIS) Model design
Stated-preference simulator design Further
research
3
Introduction
Providing travellers with relevant traffic
information is generally acknowledged as having
the potential to change their behaviour in ways
that are beneficial to the efficiency of the
transport system . The work described here is
part of my PhD research. It focuses on the
behavioural aspects with respect to traffic
information service based on innovative
technologies. Utilising an interactive
stated-preference simulator, discrete choice
models are developed to investigate factors that
influence both pre-trip and en-route switching to
alternate departure time, route and driving
style. The simulator has the potential to provide
a wealth of data collected relatively
inexpensively under controlled conditions.
4
Introduction

• Recent rapid developments in mobile
  communication technologies underly a vision among
  telecommunication operators, transport agencies,
  governments, and academia of a technological
  revolution in Advanced Traveller Information
  Systems (ATIS), which is called next generation
  ATIS.
• This study will explore the following
  questions
• What will be the user behavioural determinants
  of the use of next generation ATIS, and how will
  the relationships among them be defined?
• What will be the effects of traffic information
  provided by the innovative technologies on user
  choices?

5
Evolution of ITS and ATIS
Intelligent Transport System (ITS) Key
Development Fields

• Navigation Systems
• Traffic Management
• Public Transport
• Emergency Vehicle Operation
• Electronic Toll Collection
• Safe Driving Pedestrians
• Commercial Vehicle Operation
• Pedestrians

6
Evolution of ITS and ATIS
Traffic Information for Drivers
7
Evolution of ITS and ATIS
New generation in-car information system
-Road traffic info. ( Congestion, Travel time,
etc.) -Weather Info. (Fog, Snow, Icy road,
etc.) -Parking Info. Opening hours,
Availability -Reservation (Hotel, Restaurant,
etc.) -Help (Accident, Illness, Theft,
etc.) -E-mail, Internet, Music download,
Transaction
8
Model design
Technologies have provided great potential for
improving travel environment for the community.
The eventuation of the potential, to a large
extent, depends on how traveller react to the
information provided through these technologies,
which is still an unknown important factor.
Case Study The span of WAs rural roads is much
larger than that of urban roads, and flood
driving on rural roads presents a significant
safety risk and contributes to the higher
incident rate for rural driving. ITS has
significant potential for improving floodway
prediction and detection on rural areas. AusLink
2 funding approval of 700,000 is now requested
for works covering the research, development and
construction of road closure monitoring and
traveller information infrastructure on the Great
Northern Highway in the Kimberley Region.
9
Model design

• System Requirements
• Inform regional officers and TOC once flood
  predicted or detected
• Timely information to travellers when a flood is
  predicted or the road is flooded
• Alerts to travellers once a flood is detected

10
Model design
Information Mediums Comparison
In order to provide travellers effective and
timely information, we need to define what type
of information is required and make sure that
stable information sources are available in view
of cost and user preference considerations.
11
Model design
Theoretical model An integrative analytical
model is proposed for studying the evolution of
the next generation ATIS use and effects.
User-centred design
Social and political considerations
Information acquisition
Use
Traffic Information
Telecom Technology
Users
12
Model design
Model specification
Travel decision tree with the provision of
information
No
13
Model design
According to the decision tree, the selected
model specification is a standard linear-in-the
parameters specification. The utilities of the
standard Multi-nominal Logit (MNL) Model specific
to individual n, n 1, . . . , N are
No change
Change departure time

Change route
Change driving pattern
Change route and driving pattern
14
Stated-preference simulator design
To estimate the coefficients in the model
specified, it is needed to collect information
with respect to travellers behaviour on travel
departure time, route choice and driving
style. In this study the physical design of the
survey is chosen to be computer assisted personal
interviewing (CAPI) ,which is the current
state-of-the-art in stated preference surveys.
The survey instrument is designed considering
the different types of personalised traffic
information services. The respondents are to be
presented with various travel scenarios,
hypothetical, in order to generate complete
travel profile with traffic information. It is
specified as an information service that takes
into account users personal preferences
concerning route, departure time and driving
style. The use of a simulated environment
programmed for this survey can enable the feature
of generating dynamic stated-preference games.
From a statistical perspective, more variability
in the options presented to the respondent is an
advantage.
.
15
Stated-preference simulator design
A simulated trip from Oxford Street to the new
business school of UWA is examined. Although
there are many potential routes between these two
locations, only a subset of links and nodes are
chosen.
During travel simulation, users may encounter one
of four types of incidents. The incidents vary in
location and severity as following
.
A. traffic congestion, lane blocked B. technical
problem, bridge gates locked C. road work, one
lane closed D. vehicle accident, two lanes closed
16
Stated-preference simulator design
Departure time change and route change
Travellers repetitive travel choice behaviour
may be described as an iterative process. At
each stage the traveller chooses an action and
receives a payoff that depends on his/her action,
the new outcome, and the previous outcomes.
17
Stated-preference simulator design
Driving style change
In view of the significant financial and
environmental costs associated with different
route choices and departure time, there exists a
rationale and an opportunity for us to
investigate the effect of providing travellers
with fuel consumption information for travel
choice decision making. It is expected that for
low income travellers the fuel consumption is the
most important attribute that influences trip
choice decisions. Such travellers are assumed to
minimise their expected fuel use via changing
departure time, switching route or changing
driving style.
18
Further research
The work described in this paper is part of my
PhD research. The objective is to study the
behavioural responses of travellers under
innovative travel information. To achieve this, a
stated-preference survey needs to be conducted in
Perth after the instrument has been designed. In
the recruitment process, social characteristics
need to be considered, and the sampling process
will try to achieve a high level of
heterogeneity. However due to the limitations,
some potential sample biases can be expected. For
example, as a computer based experiment, it is
reasonable to expect that the sample will contain
a higher fraction of computer literates than in
the whole population of Perth. After survey
data is collected, the model estimation will
determine the values of parameters. According to
the time plan, the whole task will be finished in
June 2010.
.
19
Thank you