How shareable is your trip? A path-based analysis of ridesplitting trip shareability

Abstract

As an emerging sustainable mobility solution, ridesplitting services match passengers in a similar direction with a single vehicle to reduce fleet size, vehicle kilometers traveled and traffic emissions. However, these benefits can only be achieved with successful matching (sharing) between passengers, which emphasizes the importance of a comprehensive understanding of the matching success rate, i.e., shareability. Despite extensive research into the determinants of shareability, existing literature either relies on simulations and theoretical models with limited empirical validation, or focuses on system-level shareability for the whole market, overlooking the significant spatiotemporal variability of shareability across trips. This study aims to fill these gaps by proposing a path-based model that leverages real-world ridesplitting data to quantify the determinants of shareability at a finer spatiotemporal granularity. Utilizing data from New York City, our results show that: (1) shareability is spatiotemporally heterogeneous; (2) high demand intensity, especially the intensity of medium−/short-distance trips, contributes to greater shareability; (3) the positive contribution of demand intensity diminishes as it increases; (4) a higher road speed improves shareability; (5) excessive one-way street and over-dense street network are related to low shareability. These findings validate and enrich prior findings, which can be used to inform the future development of ridesplitting services.