Mandjo Béa BoréData analyst - Developer

In this project, I worked on a medical transportation issue aimed at people who do not have access to a means of travel to get to their medical appointments. The goal was to automatically plan the work of a small fleet of vans in order to organize, in a single operation, the pickup of patients at home and their drop-off at the assigned hospital, while respecting several operational constraints.

To achieve this, I used a Vehicle Routing Problem (VRP) analysis in ArcGIS Pro. Patient and hospital addresses were first geocoded to create usable points on the road network. Each patient was linked to their medical facility using the concept of order pairs, which forces the planning process to include both a pickup and drop-off step for each person.

There were numerous constraints: ensuring respect for appointment time windows so that no one is late, limiting time spent in the vehicle to ensure acceptable comfort, and managing the reduced capacity of each van (maximum six people). In addition, some patients required a vehicle equipped with wheelchair access, so I integrated specialties to ensure that only compatible vehicles could serve those individuals. Finally, each vehicle was allowed to operate only within its geographic zone, which led me to define route zones limiting their movements.

Once these parameters were configured, the VRP solver calculated optimized routes: each van begins its circuit at its origin depot, picks up patients according to the defined time constraints, drives them to their designated hospital, and then returns to the depot. The result provides a route plan that can be used immediately by drivers, with the possibility of publishing it for navigation in the field.

This work demonstrates how route optimization can concretely support access to healthcare by making transportation more efficient, more punctual, and better suited to the specific needs of users.