[Paper Review] An algorithm for integrating peer-to-peer ridesharing and schedule-based transit system for first mile/last mile access

Pramesh Kumar, Alireza Khani|arXiv (Cornell University)|Jul 15, 2020

Transportation and Mobility Innovations45 references64 citations

TL;DR

This paper proposes a transit-based ridesharing algorithm that integrates peer-to-peer ridesharing with schedule-based transit systems to solve the first mile/last mile (FMLM) problem. Using a schedule-based shortest path and space-time prism (STP) constraints, the method identifies feasible rider-driver matches and optimizes for either maximum matches or vehicle-hour savings, reducing system-wide vehicle-hours by up to 46% in simulations with 1% participation.

ABSTRACT

Due to limited transit network coverage and infrequent service, suburban commuters often face the transit first mile/last mile (FMLM) problem. To deal with this, they either drive to a park-and-ride location to take transit, use carpooling, or drive directly to their destination to avoid inconvenience. Ridesharing, an emerging mode of transportation, can solve the transit first mile/last mile problem. In this setup, a driver can drive a ride-seeker to a transit station, from where the rider can take transit to her respective destination. The problem requires solving a ridesharing matching problem with the routing of riders in a multimodal transportation network. We develop a transit-based ridesharing matching algorithm to solve this problem. The method leverages the schedule-based transit shortest path to generate feasible matches and then solves a matching optimization program to find an optimal match between riders and drivers. The proposed method not only assigns an optimal driver to the rider but also assigns an optimal transit stop and a transit vehicle trip departing from that stop for the rest of the rider's itinerary. We also introduce the application of space-time prism (STP) (the geographical area which can be reached by a traveler given the time constraints) in the context of ridesharing to reduce the computational time by reducing the network search. An algorithm to solve this problem dynamically using a rolling horizon approach is also presented. We use simulated data obtained from the activity-based travel demand model of Twin Cities, MN to show that the transit-based ridesharing can solve the FMLM problem and save a significant number of vehicle-hours spent in the system.

Motivation & Objective

• To address the first mile/last mile (FMLM) problem in suburban transit networks where coverage and frequency are limited.
• To develop a dynamic, real-time matching system that integrates ridesharing with fixed-schedule transit services.
• To reduce vehicle-miles traveled (VMT) and vehicle-hours spent in the system through optimized ridesharing integration.
• To improve matching efficiency using space-time prism (STP) constraints and a rolling horizon approach.
• To evaluate the method’s performance on a large-scale, real-world transit network using simulated data from the Twin Cities, MN.

Proposed method

• Uses a schedule-based transit shortest path algorithm to compute feasible transit itineraries for riders, including access, waiting, in-vehicle, transfer, and egress times.
• Applies space-time prism (STP) constraints to limit the search space for feasible rider and driver matches based on time and location availability.
• Develops a matching optimization model that selects optimal driver-rider pairs, transit stops, and vehicle trips to minimize detours and maximize system efficiency.
• Employs a rolling horizon approach to solve the dynamic ridesharing problem in real-time, processing requests in 5-minute intervals.
• Implements two optimization objectives: maximizing the number of matches and maximizing vehicle-hour savings.
• Uses activity-based travel demand simulation data from the Twin Cities to calibrate and validate the algorithm.

Experimental results

Research questions

• RQ1 Can a transit-based ridesharing system effectively reduce the first mile/last mile problem in low-density suburban areas?
• RQ2 How does integrating ridesharing with schedule-based transit improve system-wide vehicle-hour savings and matching rates?
• RQ3 To what extent do time flexibility and participation rates affect the computational performance and matching success of the algorithm?
• RQ4 How effective is the space-time prism (STP) technique in reducing computational time without sacrificing solution quality?
• RQ5 Can a rolling horizon algorithm efficiently handle dynamic, real-time ridesharing matching at scale?

Key findings

• The algorithm successfully reduced vehicle-hours in the system by up to 46% when maximizing vehicle-hour savings, significantly outperforming the objective of maximizing match count.
• With a 1% participation rate, the algorithm completed each 5-minute time step in under 4.23 minutes, demonstrating feasibility on standard desktop hardware.
• Approximately 54% of riders and drivers found a match during the simulation, with the majority of unmatched participants failing to find a feasible partner due to time constraints or low request volume.
• The driver detour time averaged 9.14 minutes, and rider ride time averaged 20.43 minutes, indicating that the first-mile journey is a major component of both rider and driver trip time.
• Increasing time flexibility and participation rates improved matching rates and computational load, with equal numbers of riders and drivers yielding the highest matching rate.
• The integration of stand-alone ridesharing with transit-based ridesharing improved both matching rates and vehicle-hour savings, suggesting synergistic benefits.