Lab 5 - Vehicle Routing Problem

Lab 5 centered around a type of network analysis called the Vehicle Routing Problem. This type of network route problem is based on finding the optimal routes for a fleet of vehicles to serve multiple customer orders. The VRP solver in Network Analyst can incorporate constraint information, such as time windows and travel costs to determine solutions for complex routing problems.

In the lab, we were provided data and a scenario of using the VRP solver to create optimized routes for a day's worth of pickups for a trucking company in Southern Florida. The company had a fleet of 22 trucks with one driver per truck and route, one distribution center, and 128 orders to assign to routes.

To work through the analysis I added the orders, depot, routes, and route zones features to my network analysis layer in a map. I specified certain time, distance, and vehicle parameters and solved the VRP. In this first solution, 6 orders were left unassigned, 14 trucks were utilized (based on preassignment rules), and there were a handful of time violations. The total cost was $15,067.19 and total revenue was $32,000.00.

To try and improve customer service by assigning all orders and reducing time violations, I changed the assignment rules for two additional trucks and reran the solver. This time, all orders were assigned, and there was only one time violation. The total cost increased by $1,852.43 and revenues increased by $1125.00. This second solution was an improvement over the first by better meeting the needs of the customer.

VRP solution showing optimized routes for one day's worth of 128 customer pickup orders by trucking company utilizing 16 trucks and one distribution center. Routes provide best solution with all orders assigned and only one time violation.