Identifying the Collaborative Areas of Ride-Hailing and Traditional Taxi Services Based on Vehicle Trajectory Data

Ride-hailing services (RHS) and traditional taxi services (TTS) play important roles in satisfying people’s daily travel demands. In recent years, RHS have experienced rapid growth, making daily travel more convenient and fundamentally changing the vehicle-hailing service market once dominated by traditional taxis. However, the rapid growth of RHS has also raised the general number of empty ride-hailing vehicles, potentially resulting in increasing traffic congestion and pollutant emissions. Finding the collaborative areas of RHS and TTS can effectively improve the efficiency of travel services. Accordingly, we propose a method to identify the collaborative areas of RHS and TTS based on sparsely sampled trajectory data. We first recover sparse trajectories based on a shortest path algorithm. Then, the indicators describing empty vehicles and the travel demand are defined and calculated. Finally, to identify these collaborative areas, rules are adopted to finding the areas with high unloaded vehicles of one vehicle type with a low travel demand to supplement the other vehicle type in high demand. A data set of Xiamen Island is used to validate the effectiveness of the proposed method. We find that (1) the proposed method can improve the usability of sparsely sampled trajectory data and effectively identify the collaborative areas of RHS and TTS. (2) At facilities such as office buildings and Software Park, ride-hailing vehicles are dispatched to supplement taxis on weekday mornings, while the reverse relationship holds on weekend mornings. The collaborative areas where RHS are dispatched to assist TTS are concentrated near large shopping malls on weekdays and weekends at 18:00; additionally, residential areas near scenic spots demand supplementary ride-hailing vehicles at 18:00 on holidays, while supplemental taxis are dispatched at 09:00