Optimizing Demand-Responsive Paratransit Operations: A Mixed Integer Programming Approach

The traditional demand-responsive paratransit system plays an important role in connecting people to health care, particularly those who are carless, low-income, over age 65, underinsured/uninsured, or who have a disability. However, the existing paratransit system usually has low service quality due to long waiting times, low operation frequencies, and high costs. In order to improve the service quality, we propose to design a new demand-responsive paratransit system that offers public, Uber-like options for non-emergency medical transportation. We developed mixed integer programming models to optimize the system operations with the objectives of minimizing waiting times for riders as well as operating costs for operators. The results produced in this paper will assist local departments of transportation and transit agencies as they consider operational strategies to meet non-emergency medical transport needs.