GPS data in urban online ride-hailing: A simulation method to evaluate impact of user scale on emission performance of system

With the spread of the ride-hailing service over the world, many scholars still argue whether ride-hailing is an effective travel mode for emission reduction. Since rider-hailing is a crowdsourcing system, the user scale can have a great impact on the emission performance of the system. A clearer pattern of the impact of user scale on emission performance of the ride-hailing system should be provided for better local market development and policymaking. In this study, based on massive Didi GPS records, we proposed a cross simulation model to evaluate the impact of user scale on the emission performance of the ride-hailing system and adapted the Gibbs sampling for a comprehensive computation. The result shows a strong impact of user scale on the emission performance. The mean of void cruising distance proportion varies from 2.12% to 44.58% under all situation simulation. Moreover, according to the simulation results under different day conditions, the relationship between the user scale and emission performance is not concerned with the day condition but the local regular travel pattern. Based on this relationship, we provided approximate user scales under expected thresholds of the emission and efficiency performance of ride-hailing. This work can be a foundation and guideline for future decision making on ride-hailing.