Grouping Patients for Ridesharing in Non-emergency Medical Care Services*

In this project, we consider an application scenario of ridesharing for non-emergency medical and health care services. In this scenario, patients located at different places need to get to the designated hospital or health care service center by cars. To reduce the number of cars used to transport the patients, ridesharing is used in this scenario. Due to the diversified locations of patients, patients must be grouped properly based on proximity of locations. Patients in a group need to get to the pick-up point to get on the car and get to their destination (the designated hospital or health care service center) after getting off the car. Due to the distance constraint of patients, grouping patients must be take into account the distance constraints of patients. In this paper, we propose a clustering method to group patients. The proposed method is developed by modifying the K-means algorithm to take into account the distance constraints of patients. We apply the proposed modified K-means based algorithm to group patients for a test case to illustrate the practicality of the proposed method.