On-Demand Rideshare Service with Walkable Pickup and Drop-Off Points Reduced Ride Rejections by Over 80 Percent Using Manhattan Taxi Trips.
Trip Assignment Analysis Revealed Improved Mobility When Short Walkable Links Were Integrated with an On-Demand Shared Taxi Service Using Simulated Trip Data From Manhattan.
Manhattan, New York, United States
On-demand ridesharing with optimized pick-up and drop-off walking locations
Summary Information
Shared on-demand transportation systems, where passengers with similar routes travel together, have shown potential in reducing urban congestion. In this study, an on-demand service model was designed that encourages users to walk to and from nearby pick-up and drop-off (PUDO) points to enhance system efficiency. The model was tested using a real-world dataset of Manhattan taxi trips from 2013, including nearly 10,000 hourly trip requests.
METHODOLOGY
This study considered trip assignment under the shortest time constraint by vehicle and walking between network nodes, aiming to provide the best level of service throughout the whole period of operation of the Shared Mobility on-Demand (SMoD). The assignment problem minimized the number of rejected trip requests and the expected waiting, walking, and travel times for the accepted ones, with low Vehicle Hours Traveled (VHT). Feasible assignments between groups of users and vehicles were identified, considering all the possible PUDO points for each user in the group, and determining which PUDO points are feasible. This was done by including additional constraints such as bounds on the maximum walking time per user and ensuring that users have enough time to walk towards the pick-up point. For the simulated Manhattan application of the assignment system, three scenarios were considered: (i) base- 3000 vehicles of passenger capacity six, (ii) 2000 vehicles of capacity nine, and (iii) Same fleet as in base scenario but considering users’ costs only.
FINDINGS
- Results showed that the on-demand rideshare service with walkable PUDO points could reduce ride rejections by over 80 percent, based on Manhattan taxi trips data.
- Results also showed that average walks of about one minute can reduce the VHT by more than 10 percent.
