Integrating On-demand Ride-sharing with Mass Transit at-Scale

TL;DR

The approach extends a state-of-the-art trip-vehicle assignment model to the multi-modal setting, where it develops a new integer-linear programming formulation to solve the problem efficiently and shows that the hybrid system provides significant improvements in comparison to a purely on-demand model by exploiting the efficiencies of the mass transit system.

Abstract

We are in the midst of a technology-driven transformation of the urban mobility landscape. However, unfortunately these new innovations are still dominated by car-centric personal mobility, which leads to concerns such as environmental sustainability, congestion, and equity. On the other hand, mass transit provides a means to move large amounts of travelers very efficiently, but is not very versatile and depends on an adequate concentration of demand. In this context, our overarching goal is to explore opportunities for new technologies such as ride-sharing to integrate with mass transit and provide a better service. More specifically, we envision a hybrid system that uses on-demand shuttles in conjunction with mass transit to move passengers efficiently, and provide an algorithmic framework for operational optimization. Our approach extends a state-of-the-art trip-vehicle assignment model to the multi-modal setting, where we develop a new integer-linear programming formulation to solve the problem efficiently. A comprehensive study covering five major cities in the United States based on real-world data is carried out to verify the advantages of such a system and the effectiveness of our algorithms. We show that our hybrid system provides significant improvements in comparison to a purely on-demand model by exploiting the efficiencies of the mass transit system.

Figures (7)

• Figure 1: An illustrative example of the integrated system
• Figure 2: The decomposition of multi-modal options in a simple system with two requests and two bus lines, where both bus lines have only one bus. Request 1 has two options, but request 2 has only one. Note that $F_{r_2,b_{21}}$ is an empty travel segment representing that passenger of the request 2 can walk to the bus stop without taking a ride-sharing vehicle
• Figure 3: The Transit-Integrated RTV graph of a small system with two requests, nine potential travel options, and three vehicles. Request $1$ rides on vehicle $1$ to catch the bus $b_{21}$ and rides on vehicle $2$ for the last mile. Vehicle $3$ covers the last mile of the request $2$. Note that the first mile of request 2 is matched to the dummy vehicle because it is an empty (walking) travel segment. Vehicle $2$ serves as both the last-mile vehicle of the request $2$ and the last-mile vehicle of the request $1$. The other vehicles only support one travel segment each.
• Figure 4: Average computation time of the proposed framework (per iteration with 100 requests) for the city of Chicago: Comparison with varying vehicle capacities (1 and 4) and fleet sizes (2.5, 5, 10, 20, 40 per 1000 requests)
• Figure 5: Comparison of different metrics for varying vehicle capacities (1 and 4) and fleet sizes (2.5, 5, 10, 20, 40 per 1000 requests). The rows represent results for different cities while the columns contain results for different metrics. (A) Comparison of service rate, (B) Comparison of the composition of service options, and (C) Comparison of the distribution of types of multi-modal trips.