Evaluation of Station Distribution Strategies for Next-Generation Bike-Sharing System

Bike-sharing systems have been widely used as an eco-friendly mode of urban micro-mobility in metropolitan areas. The next generation of bike-sharing may use autonomous bikes to solve the problem of station imbalance. Such an on-demand shared-use self-driving bikes service (OSABS) utilizes electricity-driven self-driving cargo bikes to offer the service at user demand locations. One can book the bike using the mobile app, and then the bike drives autonomously to the user. Unlike conventional bike-sharing systems, OSABS bikes can wait at the street or a station for the next user request after the former ended the trip. Autonomous cargo bikes require a charging station to refill the electric energy and waiting stations on the street level to avoid public disturbance while awaiting the next user. The scope of our paper is firstly to investigate suitable station distribution strategies taking both stations types into account, and secondly, to test them in a simulation model. Initially, this paper investigates various station distribution strategies for the conventional bike-sharing system and shared autonomous vehicles (SAV) based on a literature study. We derive station distribution strategies for the OSABS system. Consecutively, we present a detailed simulation model of the system. The station distribution strategies will be evaluated for charging and waiting stations based on key performance indicators (KPI) such as overall station cost, bike cost, service level, and demand distribution. Finally, we provide the first simulation-based evaluation of station distribution strategies of next-generation bike-sharing schemes.