Publication Type

Journal Article

Publication Date

7-2017

Abstract

The last-mile problem concerns the provision of travel services from the nearest public transportation node to a passenger’s home or other destination. We study the operation of an emerging last-mile transportation system (LMTS) with batch demands that result from the arrival of groups of passengers who desire last-mile service at urban metro stations or bus stops. Routes and schedules are determined for a multivehicle fleet of delivery vehicles, with the objective of minimizing passenger waiting time and riding time. An exact mixed-integer programming (MIP) model for LMTS operations is presented first, which is difficult to solve optimally within acceptable computational times. Computationally feasible heuristic approaches are then developed: a myopic operating strategy that uses only demand information from trains that have already arrived, a metaheuristic approach based on a tabu search that employs demand information over the entire service horizon, and a two-stage method that solves the MIP model approximately over the entire service horizon. These approaches are implemented in a number of computational experiments to evaluate the system’s performance, and demonstrate that LMTS is notably preferable to a conventional service system under certain conditions.

Keywords

Last mile, batch demand, routing and scheduling, mixed integer programming, myopic strategy, tabu search

Discipline

Artificial Intelligence and Robotics | Operations Research, Systems Engineering and Industrial Engineering | Transportation

Research Areas

Intelligent Systems and Decision Analytics

Publication

Transportation Science

First Page

1

Last Page

17

ISSN

0041-1655

Identifier

10.1287/trsc.2017.0753

Publisher

INFORMS (Institute for Operations Research and Management Sciences)

Creative Commons License

Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.

Additional URL

http://doi.org/10.1287/trsc.2017.0753