The Vehicle Routing Problem with Urgent Stochastic Customers
Zoom: https://eur-nl.zoom.us/j/91208040970?pwd=MzRQL1JsMlNTN1F2R2drdy9RbjZjQT09, Meeting ID: 912 0804 0970, Passcode: 677082
Several industries deal with a combination of routine, plannable tasks and incident response. For vehicle routing, such problems arise in parcel pickup and delivery, maintenance, surveillance and security. Our research is inspired on hospital cleaning. In the Jeroen Bosch Ziekenhuis, routine tasks and incident response are divided over two teams. Travel distance as well as the response time are important performance measures. Intuitively, it might be beneficial for response time to merge the teams and sacrifice some a priori travel distance in order to spatially distribute the cleaners. The main research questions are how to route and schedule the routine tasks while optimizing for both travel time and response time to incidents and whether it is beneficial to merge the teams or not.
We have developed a mixed integer program that solves the A priori Vehicle Routing Problem with Urgent Stochastic Customers (a priori VRPuSC). We simulated incident response with separate teams (current practice) and with merged teams. For the merged teams we compared a VRP with route balancing to the a priori VRPuSC.