Understanding operation patterns of urban online ride-hailing services: A case study of Xiamen

Online ride-hailing has gradually become a popular travel choice worldwide, while it also brought policy challenges to balance the traditional taxi industry and online ride-hailing services. Understanding the operation patterns of urban online ride-hailing services is essential for government policy-making. However, insufficient attention has been paid to the operating characteristics of online ride-hailing vehicles due to limited empirical data. This paper proposes a cluster analysis framework for the identification of different operation patterns of urban online ride-hailing. The customer order and GPS data of online ride-hailing vehicles and traditional taxis in Xiamen, China is used in this study. The k-means++ clustering algorithm is used based on the proposed intensity and stability indices of ride-hailing vehicle operating characteristics. The results show that there are three types of online ride-hailing operation patterns, namely full-time (which accounts for 52.801%), part-time (29.502%), and occasional (17.697%). The operation pattern of full-time ride-hailing vehicles is similar to that of traditional taxis, but with lower intensity and stability due to a reduced workload and flexible time schedule. Part-time ride hailing vehicles are operated unsteadily and irregularly in the drivers' spare time, and the working time periods are mainly concentrated in the morning and evening peak hours. Occasional ride-hailing vehicles provide very limited service. Finally, several policy suggestions for online ride-hailing from the perspective of government management, e.g., the number of licenses and operation places and time periods, are proposed based on the results.