Growing interest in sustainable transportation systems has driven decision-makers toward policies and investments aimed at promoting cycling, but little to no effort has been made toward incorporating bicycle transport in transport planning models. This study contributes toward this direction by estimating a bicycle route choice model in value-of-distance space from a large sample of 3384 cycling trips that were traced with GPS devices in the Copenhagen Region. The novelty of this study lies in (i) observing cyclists’ behavior in a cycling-oriented country, (ii) exploiting rich data about the cycling environment, (iii) estimating the model in value-of-distance rather than preference space, and (iv) not focusing only on preferences for traditional variables (e.g., distance, turns, hilliness, intersections, motorized road characteristics), but also on perceptions and preferences for bicycle facilities (e.g., bicycle lanes, bicycle paths, bicycle traces) and land-use designations (e.g., residential, industrial, sports, scenic areas). The findings from the model show that: (i) cyclists exhibit heterogeneous preferences for avoiding right and left turns, cycling the wrong way, using roundabouts and bridges, and cycling alongside residential and scenic areas; (ii) cyclists dislike cycling on unpaved and hilly surfaces and alongside larger roads; (iii) cyclists have clear perceptions about different types of bicycle facilities, with a preference for bicycle lanes and segregated paths; (iv) cyclists have clear perceptions about land-use designations, with a preference for cycling alongside sports and scenic areas; (v) time-of-day and air temperature contribute to the perceptions of cyclists and their preferences for bicycle facilities and land-use designations.
|Journal||International Journal of Sustainable Transportation|
|Number of pages||12|
|Publication status||Published - 2018|
- Bicycle infrastructure
- Doubly stochastic generation function
- Generalised mixed path size logit
- Route choice