Optimization approaches to identify cyclists’ route choice criteria

When choosing a path to follow along a cycling network, cyclists are not merely interested in minimizing the distance to travel. Other features also play a role, such as safety, practicability, and environmental conditions. Different users may assign different weights to each one of these features. Subdividing users into groups that assign similar weights to each feature, evaluating the numerical consistency of each of these groups, and reporting the weights assigned by each group to different features of the cycling network are quite relevant from a practical point of view, in particular, to allocate in the best possible way a given budget for infrastructural interventions. Indeed, decision makers need to know the impact of their decisions, and this is strictly related to the way users perceive different features of the cycling network. In this paper, we propose two different optimization problems and related solution algorithms to detect the weights assigned to different road features by various user groups. The two problems are based on different available data. In one problem we assume that flows along some arcs of the cycling network (collected, e.g., through cameras) are available. In the other problem we assume that the true paths followed by some users (collected, e.g., through bike-sharing services) are known.