Database Reference
In-Depth Information
opportunities to satisfy interaction needs, a poorly connected transport system
limits economic and social development (
Ort uzar and Willumsen
,
2002
). The
transport system thus allows individuals to trade time for space when moving to
(activity) locations (
Miller
,
2003
;
Rietveld
,
1994
).
Traffic Modeling Standard
Rising concerns over these increasingly intolerable externalities have generated
particular interest in how transport planning policies might at least moderate the
pressures in growth in personal mobility and support the principles of sustainable
development (
Barrett
,
1996
;
Salomon et al.
,
1993
). Originally, transport planning
policies focused on mastering the massive growth in car mobility. These policies
were adopted in an immediate response to the predicted growth in (car) mobility.
The estimation and forecasting of travel demand and behavior were handled
by a standard methodological approach commonly referred to as the four-step
modeling approach consisting of trip generation, trip distribution, mode choice,
and assignment of travel demand to highway and transit networks. In the trip
generation stage, the goal is to predict the total number of trips generated and
attracted to each zone of the study area. In the second stage, the question is how
to distribute trips among destinations. The result of this step is a 2D array of
cells (matrix) where rows and columns represent each of the zones in the study
area and the cells contain the number of trips that go from the origin zone (in
the rows) to the destination zone (in the columns). The latter is also known as an
origin-destination matrix
. Next, in the third step the transport mode is chosen.
The output of this step is typically an origin-destination matrix that represents
the number of trips that are carried out by the different transport modes. While
the previous three steps mainly deal with the demand side of travel, the last step
in the four-stage methodology is mainly related with the supply side. In this
step, the supply side of the transportation system - which is made up of a road
network and is represented by links and costs - is confronted with the demand
side of travel that has been estimated in the first three steps. The result of this
step is the amount traffic projected on the road network, typically represented
as number of vehicles on road segments.
Toward Data-Driven/Aware Models
In parallel with the traffic science evolution discussed above, there is growing
literature and research available, originating from the field of mobility data
mining (see Chapter
6
of this topic), which emerged only recently, during the
last decade. While the overall goal is the same, that is, to help policy makers
to deal with traffic-related questions, the techniques used and the processes
adopted are completely different. The main difference is the fact that most of
the techniques are fully data driven and therefore also less policy sensitive.
It can be very interesting to see how both domains could complement each
