Abstract:

E-scooters have emerged in recent years as an innovative and user-friendly transportation mode in urban areas worldwide. These systems offer short-term rental of vehicles with various return locations, thereby providing mobility when public transport or vehicle ownership are not a sufficient option. However, their incorporation in urban road networks is complex, since legal, social, and technical considerations often arise with respect to their usage. As official regulatory frameworks are often lacking, micromobility operators typically devise ad hoc geofences to create virtual geographic boundaries, restricting parking or access for users. In this context, this paper proposes a methodological tool for assisting decision-making in regulating e-scooter usage in urban areas. A service network design model is introduced for that purpose, aiming to maximize the extent of geofences in an urban area for the sake of maximizing road safety, while considering travel time impacts for users. The problem is formulated as a bi-level bi-objective optimization model and solved using a Non-Dominated Sorting Genetic Algorithm, NSGA-II. The results show that a range of non-dominated solutions is generated, with significant discrepancies between operator-optimal and user-optimal solutions. In general, results highlight the complicated decision-making process for restricting e-scooter access and thus, the need to consider a set of solutions, versus a single candidate geofence plan.