Unbalanced usage of free-floating bike sharing connecting with metro stations

Free-floating bike-sharing systems allow users to unlock nearby bikes and leave the bikes at any authorized parking area, acting as a time-saving approach for solving the first/last mile problem by connecting trip origins/destinations to metro stations (i.e., bike-and-ride). Inspired by a recent study uncovering the scaling-free phenomenon on the bike-sharing mobility network, this paper addresses the following question: does the power-law distribution also appear in bike-and-ride activities? How does the unbalanced usage pattern of bike-and-ride vary by time and space? To answer the questions, we examine one-week trip orders of free-floating bike sharing in Beijing. We find that the unbalance activities of bike-and-ride show a similar segmented scaling-free feature dis-playing a first sub-linear regime followed by another super-linear regime on weekdays and weekends. The unbalanced features on weekdays are more substantial than those on weekends and uncovered by the time-varying unbalance in different regimes. Moreover, the unbalanced state corresponding to each metro station obeying non-stationary and dramatically changing peak hours on weekdays suggests that commuting between the suburban and urban areas may be the primary source of the unbalance usage of bike -and-ride in Beijing. Our findings can improve the efficient redistribution of bike-sharing supply and promote the coupling efficiency among transportation modes.(c) 2022 Elsevier B.V. All rights reserved.

Automated summary

This paper examines the unbalanced usage pattern of bike-and-ride activities in free-floating bike sharing in Beijing. The study finds that the unbalance activities show a power-law distribution, with weekdays exhibiting more significant unbalanced features than weekends. The peak hours of unbalanced state at metro stations on weekdays suggest commuting between suburban and urban areas as the primary source of the unbalance usage of bike-and-ride in Beijing.