A systematic review and prospective outlook on bicycle-sharing rebalancing research

Effective bicycle rebalancing is critical to resolving the spatiotemporal supply-demand imbalance that limits the efficiency and user satisfaction of bicycle-sharing systems. This paper delves into the issue of bicycle rebalancing, categorising existing studies into four non-parallel strands based on the dimensions of rebalancing agents, demand characteristics, and temporal dynamics: static rebalancing under deterministic demand, dynamic rebalancing under deterministic demand, rebalancing under demand uncertainty, and user-based rebalancing. Each strand is systematically reviewed to synthesise its objectives, assumptions, and methods, and to summarise the canonical solution routines commonly used in practice. Building on this synthesis, we discuss the major limitations observed across the four strands and outline a set of prospective research directions, including integrated and feedback-enabled rebalancing, adaptive temporal granularity, balancing robustness and efficiency, behaviour-aware incentive mechanisms with response uncertainty and time delays, and hybrid rebalancing governance within an integrated decision framework.