The Data and Decision Sciences Lab is constantly working with local community partners on several data related projects that have a local and global impact.
Ant Colony Optimization for the Vehicle Routing Problem with Time Windows and Its Application to a Health Insurance Company’s In-Home Visit Scheduler
Community Partner: Medica
Team Members Involved
Many health insurance companies employ a staff of nurses who conduct in-home preventative exams and wellness checks for certain segments of their subscribers. In order to use its nursing staff in the most efficient manner, the company must optimize the overall process of determining which patients each nurse should visit as well as the order in which those visits should occur so that time windows specified by patients are honored. This can be modeled as a Vehicle Routing Problem with Time Windows (VRPTW). Ant Colony Optimization (ACO) is a heuristic method that mimics the natural behavior of ants by using artificial agents who deposit “pheromones” on shorter, better routes and over time are able to find a near optimal solution. This project developed, implemented, and tested a real-time appointment scheduler based on an ACO-VRPTW algorithm. The implementation was done in Python and included a graphical interface to assist with performance tuning and parameter selection. In order to remain HIPAA compliant, data for the scheduler was engineered through the use of Census Data and Google Maps API. Results of testing show that the ACO approach significantly increases the number of appointments that can be scheduled over the current ad hoc approach.