Active Travel England and Alan Turing Institute Exhibition

Using historic geospatial data to predict change in walking and cycling mode share in response to infrastructure change

Safe, accessible active journeys enable healthier, more sustainable travel routines, but rely on appropriate infrastructure. New walking and cycling infrastructure can encourage a switch to active travel modes, and understanding the relationship between localised changes in active travel infrastructure and localised changes in active travel behaviour can inform effective and high-value-for-money infrastructure planning. Our work uses historic geospatial data representing the cycling and walking network infrastructure. We observe changes in these networks over time, such as new or improved cycleways, and we encode these observed network changes in a set of features for predictive modelling. We combine these features with publicly available demographic and journey type data, and train models predicting the change in the share of journeys walked and cycled. This allows us to make locally-specific predictions of the impact of proposed packages of walking and cycling infrastructure improvements: we will share this proof-of-concept tool through our exhibition, which includes an interactive map of infrastructure change features and model predictions. Our work explores the value of historic geospatial data for measuring changes to active travel infrastructure and provides a pipeline allowing planners to explore the potential for active travel mode shift from their proposed infrastructure changes.