A comparison in travel patterns and determinants of user demand between docked and dockless bike-sharing systems using multi-sourced data

Abstract

The co-existence of traditional docked bike-sharing and emerging dockless systems presents new opportunities for sustainable transportation in cities all over the world, both serving door to door trips and accessing/egressing to/from public transport systems. However, most of the previous studies have separately examined the travel patterns of docked and dockless bike-sharing schemes, whereas the difference in travel patterns and the determinants of user demand for both systems have not been fully understood. To fill this gap, this study firstly compares the travel characteristics, including travel distance, travel time, usage frequency and spatio-temporal travel patterns by exploring the smart card data from a docked bike-sharing scheme and trip origin–destination (OD) data from a dockless bike-sharing scheme in the city of Nanjing, China over the same spatio-temporal dimension. Next, this study examines the influence of the bike-sharing fleets, socio-demographic factors and land use factors on user demand of both bike-sharing systems using multi-sourced data (e.g., trip OD information, smart card, survey, land use information, and housing prices data). To this end, geographically and temporally weighted regression (GTWR) models are built to examine the determinants of user demand over space and time. Comparative analysis shows that dockless bike-sharing systems have a shorter average travel distance and travel time, but a higher use frequency and hourly usage volume compared to docked bike-sharing systems. Trips of docked and dockless bike-sharing on workdays are more frequent than those on weekends, especially during the morning and evening rush hours. Significant differences in the spatial distribution between docked and dockless bike-sharing systems are observed in different city areas. The results of the GTWR models reveal that hourly docked bike-sharing trips and dockless bike-share trips influence each other throughout the week. The density of Entertainment points of interest (POIs) is positively correlated with the usage of dockless bike-sharing, but negatively correlated with docked bike-sharing usage. On the contrary, the proportion of the elderly has a positive association with the usage of docked bike-sharing, but a negative association with the usage of dockless bike-sharing. Finally, policy implications and suggestions are proposed to improve the performance of docked and dockless bike-sharing systems, such as increasing the flexibility of docked bike-sharing, designing and promoting mobile applications (APP) for docked bike-sharing, improving the quality of dockless shared bikes, and implementing dynamic time-based pricing strategies for dockless bike-sharing.