A passenger-to-driver matching model for commuter carpooling: Case study and sensitivity analysis

• Published in: Transportation research. Part C, Emerging technologies Vol. 117; p. 102702
• Main Authors: Liu, Xiaobing, Titheridge, Helena, Yan, Xuedong, Wang, Rui, Tan, Weimin, Chen, Deqi, Zhang, Jiechao
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.08.2020
• Subjects: Carpooling; Commuter; Generalized trip cost; Matching model; Promotion instruments; Sensitivity; Commuter; Matching model; Sensitivity; Promotion instruments; Carpooling; Generalized trip cost
• ISSN: 0968-090X; 1879-2359
• DOI: 10.1016/j.trc.2020.102702

•A passenger-to-driver matching model for commuter carpooling was initially proposed.•38.3% of trips can successfully form carpooling trips and save fuel significantly.•The participant flexibilities can help matching performance, especially when participant rate in a lower level.•The optimal fee-sharing ratio is not the traditional half to half.•The disincentives seem to have a greater effect on carpooling choice than the incentives. For the transport sector, promoting carpooling to private car users could be an effective strategy over reducing vehicle kilometers traveled. Theoretical studies have verified that carpooling is not only beneficial to drivers and passengers but also to the environment. Nevertheless, despite carpooling having a huge potential market in car commuters, it is not widely used in practice worldwide. In this paper, we develop a passenger-to-driver matching model based on the characteristics of a private-car based carpooling service, and propose an estimation method for time-based costs as well as the psychological costs of carpooling trips, taking into account the potential motivations and preferences of potential carpoolers. We test the model using commuting data for the Greater London from the UK Census 2011 and travel-time data from Uber. We investigate the service sensitivity to varying carpooling participant rates and fee-sharing ratios with the aim of improving matching performance at least cost. Finally, to illustrate how our matching model might be used, we test some practical carpooling promotion instruments. We found that higher participant role flexibility in the system can improve matching performance significantly. Encouraging commuters to walk helps form more carpooling trips and further reduces carbon emissions. Different fee-sharing ratios can influence matching performance, hence determination of optimal pricing should be based on the specific matching model and its cost parameters. Disincentives like parking charges and congestion charges seem to have a greater effect on carpooling choice than incentives like preferential parking and subsidies. The proposed model and associated findings provide valuable insights for designing an effective matching system and incentive scheme for carpooling services in practice.