A two-phase algorithm for the Sequential Vehicle Routing Problem with Cross Docking (SVRPCD)

Abstract

Today’s customer-driven economic environment shifts from “supply chain” to “demand chain”. To successfully compete, companies are faced with the challenge of minimizing inventory and reducing warehouse costs whenever possible while keeping retailers’ shelves full. Cross Docking is a strategy whose ability to reduce lead times, control inventory carrying costs and deliver just-in-time service with a minimal investment makes it an attractive option for many companies responding to economic pressures. In a cross docking strategy, warehouses and distribution centers serve as transfer points for inventory, but no inventory is held at these transfer points. One of the keys to keeping a Cross Docking facility synchronized is to design a system and schedule that are obtainable, taking into account travel times and expectations for how much freight will actually fit on a truck; therefore, Vehicle Routing should be considered. This research examines a Vehicle Routing Problem with Cross Docking (VRPCD) considering cross docking operational characteristics such as assignment restrictions, deadlines and door environment. A sequential operational VRPCD model is presented. Its objective is to determine the best route that minimizes operational costs of shipment, and the minimum operational cost within the cross docking facility. A two-phase algorithm using Variable Neighborhood Search (VNS) has been proposed. The experimental results show that the VNS algorithm is capable of finding good solutions in a much shorter computation time when compared with Gurobi Optimizer solutions. Good quality of solutions is obtained in all the experiments developed.