Usage and Temporal Patterns of Public Bicycle Systems: Comparison among Points of Interest

Abstract

The public bicycle system is an important component of “mobility as a service” and has become increasingly popular in recent years. To provide a better understanding of the station activity and driving mechanisms of public bicycle systems, the study mainly compares the usage and temporal characteristics of public bicycles in the vicinity of the most common commuting-related points of interest and land use. It applies the peak hour factor, distribution fitting, and K-means clustering analysis on station-based data and performs the public bicycles usage and operation comparison among different points of interest and land use. The following results are acquired: (1) the demand type for universities and hospitals in peaks is return-oriented when that of middle schools is hire-oriented; (2) bike hire and return at metro stations and hospitals are frequent, while only the rental at malls is; (3) compared to middle schools and subway stations with the shortest bike usage duration, malls have the longest duration, valued at 18.08 min; and (4) medical and transportation land, with the most obvious morning return peak and the most concentrated usage in a whole day, respectively, both present a lag relation between bike rental and return. In rental-return similarity, the commercial and office lands present the highest level.

Keywords

• public bicycle system
• bicycle usage
• temporal changes
• <i>K</i>-means clustering
• distribution model
• pattern comparison