GPS Data and Car Drivers’ Parking Search Behavior in the City of Turnhout, Belgium - Geoinformatics for Intelligent Transportation - page 241

parking strategy evaluation procedure of the previous parking search models
delivers systematically too high search times .

The second group of related studies concerns studies in which parking search
models are developed (for a literature review see [ 3 , 7 , 8 ]. The studies focus on
extensions of network based assignment models [ 9 ] or agent-based models [ 6 , 8 ,
10 ]. The models aim to investigate:

'

The impact of cruising for parking on traf c congestion [ 9 ];

￿

The environmental costs of parking search process [ 10 ];

￿

The generation of distributions of key values like search time, walking distance,
and parking costs over different drivers groups [ 8 ].

￿

A variety of publications shows that the use of GPS tracking data becomes more
popular to collect travel related data [ 11

13 ]. The data is easy to collect, accurate,
and detailed [ 14 ]. Also in the context of parking search behavior, GPS tracking is a
useful means of data collection [ 3 , 4 ]. The experiences with the use of GPS tracking
to identify parking search behavior are still limited.

-

3 Research Approach and Data Collection

To get insight into car drivers parking search behavior the following research

approach has been adopted. First, a city was selected according to requirements
de

'

ned by Kaplan and Bekhor [ 3 ]. Next, participants were recruited who plan to
make several trips by car to the centre of the selected city. All participants received
a GPS logger and were asked to turn on the logger when driving with their car to
the city center of the selected city. In this study the

'
was used. Every 3 s, the recorder stores its (horizontal and vertical) position, the
day and time of day, and the number of satellites involved in the registration. For
technical details of this recorder see http://blinkgadget.blogspot.com/2010/10/kode-
produk-747a.html . The logger is easily to set and the data is easily to retrieve using
for example the software

' i-Blue 747A+GPS recorder

( http://www.datalog.de ). The spatial data are
stored, analyzed, and presented using the Geographic Information System Trans-
CAD 6.0 ( http://www.caliper.com ) .
In additions, the participants were invited to

' DataLog '

filled out a small questionnaire
consisting of some questions regarding personal characteristics (gender, age,
familiarity with the city, and possession of parking card) and for each trip some
speci

c characteristics (type of car used during the trip, orientation on speci c

parking facility before leaving home, and trip purpose).
The study was carried out in the city of Turnhout, a small city in the North of
Belgium (Fig. 1 ). The city meets the four basic requirements set by Kaplan and
Bekhor [ 3 ]: (i) area should be characterized by high activity generation and limited
parking supply; (ii) area should offer a selection of off-street and on-street parking
opportunities; (iii) area should attract both local and non-local visitors, and (iv) area
should be bounded by natural and/or street network boundaries.