Understanding the spatiotemporal variation of ride-hailing orders under different travel distances

This paper aims to explore the spatio-temporal variation of ride-hailing demands under different travel distances, using operation data of ride-hailing in Chengdu, China. First, the characteristics of ride-hailing demand under different travel distances during the morning and evening rush hours are analyzed. Second, to capture the spatio-temporal heterogeneity, geographically weighted regression (GWR) models are established to discern influential factors of ride-hailing demand under different travel distances and time periods. The model results demonstrated that the fit of the geographically weighted regression model is better than that of the traditional ordinary least squares (OLS) model. The built environmental factors such as road density, population density, sports and leisure services, medical care services, and bus stops all have significant spatio-temporal heterogeneity on ride-hailing demand at different travel distances. The spatial analysis suggests that there is competition between short-distance ride-hailing service and public transportation in the urban core area, and long-distance ride-hailing service supplements public transportation gaps in the suburbs. Finally, based on the research results, several effective policies and suggestions were proposed to improve the service level of ride-hailing.