Engaging in sufficient physical activity is known to decrease the risk of poor health and premature
mortality, due to its impact on several noncommunicable diseases (e.g. heart disease). Urban
interventions in the physical environment (e.g. improved sidewalks) to promote active transport
are powerful tools to increase physical activity levels. However, urban interventions are not under
the control of the evaluating researchers, making the evaluations very challenging to design.
Complementary to evaluations of urban interventions, simulations of such interventions can
provide crucial information for transport planning and public health.
The main objective of this study is to evaluate the impact of simulated urban interventions on
transport-related physical activity with agent-based simulation models. Innovative machine
learning prediction models based on self-collected GPS and accelerometer data will allow us to
detect transport behavior and related physical activity levels in four data sets (n = 4443) previously
collected in Ghent (Belgium). With the combined data, we will run agent-based simulation models
of urban interventions. The simulated urban interventions will be evaluated on their impact on
physical activity levels, through changing transport behavior. Future research will be able to use
the models to evaluate the potential impact of urban interventions on other transport-related
exposures; for example, the impact of traffic calming measures on air pollution.