Food Rescue and Delivery Heuristic Algorithm for Periodic Unpaired Pickup and Delivery Vehicle Routing Problem

This paper addresses a special case of periodic vehicle routing problem in which each node has a nonnegative supply or demand of a single product that is unpaired. The product collected from a pickup node can be delivered to any one node or multiple delivery nodes, and the demand of a delivery node can be met by the product collected from any one node or multiple pickup nodes. This periodic unpaired pickup and delivery vehicle-routing problem is a novel variant of the periodic vehicle routing problem. The objective of the problem was to design the pickup and delivery vehicle routes to meet required service levels of delivery nodes, minimizing the total transportation cost while satisfying certain operational constraints. This problem was driven by food relief operations in Sydney, Australia. The logistics aspect of the approach was to design and execute a vehicle routing problem for a food rescue and delivery network. The specific goals were to develop an integrated linear programming model for this new variant of the periodic vehicle routing problem and to propose an integer programming based heuristic solution approach to solve the problem introduced in the paper. The heuristic algorithm was tested with small instances created from Cordeau's benchmark instances, and the solution approach was validated against optimal solutions obtained through the exact method before implementation on a food rescue and delivery network. The heuristic approach was found to be comparable with the optimal solution and can solve the real-world scenarios with significantly fewer resources than are used in practice.