台北公共自行車租賃系統使用型態之分析

Better understanding the Taipei public bikesharing system: Exploring activity patterns

鄭群彥(Chun-Yen Cheng)

指導教授 ： 汪進財

國立交通大學/管理學院/運輸與物流管理學系/碩士(2014年)

https://doi.org/10.6842/NCTU.2014.00621

摘要

City-wide urban infrastructure are increasing reliant on information and communication technology (ICT) to improve and expand the service. Therefore, by taking advantage of passive sensors and Open Data policy, large scale of human behaviour data can be sensed and become more easily accessed independently. Data mining plays a vital role in helping to discover the patterns of human behaviour. In this study, we focus on the emerging urban transport infrastructure: public bikesharing system of Taipei City: YouBike. YouBike is launched from August 2012, operated by Giant, recognised as the world’s largest bicycle manufacturer. Besides, it was the first large-scale public bike sharing system to be implemented in Taiwan. Currently, there are 166 bike station is operation. This study has shown that how bikesharing usage data which mainly focuses on the changes of the number of available bicycles across all stations not only reveals the station activity patterns but also explores the underlying temporal and spatial dynamics of a city. The clustering results indicate that station activity patterns could be categorised into three groups: which are daytime origins nighttime destinations, daytime destinations nighttime origins, and combined origins and destinations. Each cluster groups reveal the different activity patterns throughout the day. We believe that the visualisation of average temporal activity patterns and the clustered results could easily lead to a better understanding of the bicycle availability information. In addition, it is expected to improve the Taipei YouBike service itself, avoiding a future empty or full station through an improved redistribution of bicycles via rebalancing trucks. As a result, it would help to improve user satisfaction with the enhanced service and it is possible to attract more people to use YouBike as an enhanced green transport system.

關鍵字

公共自行車租賃系統、使用型態、資料探勘、集群分析

並列摘要

City-wide urban infrastructure are increasing reliant on information and communication technology (ICT) to improve and expand the service. Therefore, by taking advantage of passive sensors and Open Data policy, large scale of human behaviour data can be sensed and become more easily accessed independently. Data mining plays a vital role in helping to discover the patterns of human behaviour. In this study, we focus on the emerging urban transport infrastructure: public bikesharing system of Taipei City: YouBike. YouBike is launched from August 2012, operated by Giant, recognised as the world’s largest bicycle manufacturer. Besides, it was the first large-scale public bike sharing system to be implemented in Taiwan. Currently, there are 166 bike station is operation. This study has shown that how bikesharing usage data which mainly focuses on the changes of the number of available bicycles across all stations not only reveals the station activity patterns but also explores the underlying temporal and spatial dynamics of a city. The clustering results indicate that station activity patterns could be categorised into three groups: which are daytime origins nighttime destinations, daytime destinations nighttime origins, and combined origins and destinations. Each cluster groups reveal the different activity patterns throughout the day. We believe that the visualisation of average temporal activity patterns and the clustered results could easily lead to a better understanding of the bicycle availability information. In addition, it is expected to improve the Taipei YouBike service itself, avoiding a future empty or full station through an improved redistribution of bicycles via rebalancing trucks. As a result, it would help to improve user satisfaction with the enhanced service and it is possible to attract more people to use YouBike as an enhanced green transport system.

並列關鍵字

public bikesharing system, station activity pattern, data mining, cluster analysis

參考文獻

Bachand-Marleau, J. et al. 2012. Better Understanding of Factors Influencing Likelihood of Using Shared Bicycle Systems and Frequency of Use. Transportation Research Record: Journal of the Transportation Research Board. 2314(-1), pp.66-71.

Bakıcı, T. et al. 2013. A Smart City Initiative: the Case of Barcelona. Journal of the Knowledge Economy. 4(2), pp.135-148.

Baptista, P.C. et al. 2012. ICT Solutions in Transportation Systems: Estimating the Benefits and Environmental Impacts in the Lisbon. Procedia - Social and Behavioral Sciences. 54(0), pp.716-725.

Black, W.R. 1996. Sustainable transportation: a US perspective. Journal of Transport Geography. 4(3), pp.151-159.

Bordagaray, M. et al. 2012. Modeling User Perception of Public Bicycle Services. Procedia - Social and Behavioral Sciences. 54(0), pp.1308-1316.