Mobile phone data in urban bicycle-sharing: Market-oriented sub-area division and spatial analysis on emission reduction potentials

Abstract

In recent years, Internet-supported non-dock bicycle-sharing service had emerged and gradually become warmly applauded by the public. In the non-dock bicycle-sharing system planning, the definition of coverage area and the division of sub service areas are highly essential and fundamental, which get little attention in the existing studies. In this paper, we propose a market-oriented methodology for non-dock bicycle-sharing system planning. Firstly, the potential demand of bicycle-sharing use is identified using mobile phone data. Then, a link network is constructed to represent the spatial distribution of potential demand. The algorithm of community detection is applied to the link network to discover the densely connected nodes in the area and define the division of sub service areas. Finally, the potential patterns of bicycle-sharing use, the spatial distribution of the potential demand, and the potential emission reduction are analysed based on the resulting division of sub service areas. The proposed method is tested using a data set of approximately 34 million GPS trajectories obtained in Tokyo. The area of Tokyo is subdivided into 21 sub service areas. Suggestions for bicycle management, infrastructure development, and the bicycle-sharing system planning are given regarding sub service areas with different properties. The potential emission reduction for the introduction of bicycle-sharing system is calculated in the area of Tokyo.