A Protype System for Hybrid Taxi Rideharing Service

Taxi ridesharing is increasingly recognized as a cost-effective transportation solution and an effective method for alleviating traffic congestion and reducing environmental pollution in major metropolitan areas around the world. By enabling multiple riders to share a single ride, this service not only lowers travel expenses for individuals but also enhances drivers' income. Despite these benefits, a critical challenge remains: the scheduling problem. This research presets a novel approach to taxi ridesharing through a hybrid system. Initially, the system's architecture is designed to handle ridesharing in two distinct phases. The first phase involves forming regular rideshare groups offline based on riders' frequent paths. This approach ensures that regular riders are consistently served without needing to interact the device each time they use the service, thus saving them time and effort. In the second phase, the system supports dynamic insertion, allowing real time rider to join an ongoing shared trip without disrupting the schedules of current riders. To manage this integration efficiently, the paper introduces an innovative insertion method called biIns. This method is specifically designed to handle shared trip request swiftly, ensuring that new request are accommodated while respecting time constraints of existing passengers. To validate the proposed system, a prototype is developed, demonstrating the hybrid approach in action. In this prototype, rideshare groups with regular paths are established offline, while real time requests are integrated into the existing schedule in an online manner. This dual approach exemplifies the system's capability to optimize both regular and dynamic ridesharing processes, offering a promising solution to the challenges of scheduling in taxi ridesharing services.