Abstract
This paper presents the insights of imbalanced public bicycle distributions, i.e. unable to pick up/return bikes due to empty/full stations through the analysis of spatio-temporal activity patterns of bike stations. YouBike, the public bikesharing system of Taipei City was examined. Taking advantage of Open Data policy, the changes of the number of available bikes across all stations were collected to identify station activity patterns. The relationship between spatial characteristics and station activity patterns were explored. The clustering results indicate that station activity patterns could be categorised into three groups and each reveals different activity patterns throughout the day. The visualisation of average temporal activity patterns and clustered groups are illustrated as well. Such results could provide better understanding of bikesharing usage and the underlying temporal and spatial dynamics of a city.
