Comparing the performance of different types of bike share systems. (May 2021)

• Record Type: Journal Article
• Authors: Kou, Zhaoyu; Cai, Hua
• Is Part Of: Transportation research. Volume 94(2021)
• Publication Date: 2021-05
• Subjects: Bike share -- Stochastic simulation -- Excess time -- Station-based -- Dockless; Transportation -- Research -- Periodicals; Transportation -- Environmental aspects -- Periodicals; 354.76
• DOI: 10.1016/j.trd.2021.102823
• Languages: English
• ISSNs: 1361-9209

Abstract

Highlights: Proposed a modeling framework to quantitatively compare different types of BSSs. Service level is more sensitive to system type change in high-usage-intensity cities. Dockless BSSs save 1.2–1.8 min' excess time per trip compared with station-based BSSs. Station-based BSSs with sparse network benefit more from changing to hybrid BSSs. Abstract: Bike share systems (BSSs) can be implemented in different types: station-based, dockless, and hybrid. Existing literature lacks models to quantitatively compare and assist the selection of system types. This study proposed a stochastic simulation framework to compare how different system types may impact user experience and system operations. The framework estimates actual origins-destinations of travel demands and integrates the user behavior model and rebalance optimization model. Applying the framework to three cities' BSSs, our simulation results found that, although station capacities restrict bike allocations, the forced rerouting trips (returning bikes to neighboring stations when target stations are full) indirectly rebalance station-based systems. This benefit is lost in dockless systems, partially canceling the benefit of bike allocation flexibility. However, dockless users can save trip time by 10%–15% with bike access/return. Overall, systems with high usage-intensity (e.g., those in Philadelphia and Chicago) can benefit from transitioning their station-based systems into hybrid systems.