User Behavior of Bikesharing Systems Under Demand–Supply Imbalance

Bikesharing systems have been expanding rapidly. They are recognized as a strategic tool to integrate transit and to foster sustainable transportation in urban areas. A critical problem that arises in the planning and deployment of a bikesharing system lies in demand–supply imbalance as the result of the spatiotemporal dynamics of demand patterns. The problem manifests itself in situations in which no bikes are available to rent and no docks are available for bike return at certain popular stations. This study developed a framework for modeling user behavior in response to such situations. Particular emphasis was placed on user on-site choice behavior and the behavioral aspect related to wait time if a decision to wait for the next bike or dock was made. These behaviors were approached with the use of a discrete choice model and a duration model, respectively. Further, a discrete-continuous process was proposed to address the potential interrelation that existed between the choice to wait and the length of waiting time to be endured. Within the developed framework, the models were estimated on the basis of the data from a revealed preference–stated preference survey conducted on the YouBike system in Taipei, Taiwan. The estimation results provided insights into user familiarity and uncertainty toward the system, travel situation, or environment and the associated effects on user behavior of a bikesharing system under a demand–supply imbalance.