A MATHEMATICAL MODEL AND SOLUTION APPROACH FOR A HETEROGENEOUS FLEET OPEN VEHICLE ROUTING PROBLEM

Abstract

This research proposes a new mathematical model and a solution method for a heterogeneous fleet open vehicle routing problem (HFOVRP). HFOVRP focuses on determining the set of customers and delivery routes for a mixed fleet of vehicles that does not return to the depot after servicing the last customer on a route. This situation usually occurs in manufacturing companies or warehouses that use outsource delivery companies, as well as in delivery companies that use third party logistics companies. In this research, HFOVRP is applied at a government enterprise that hires outsource delivery companies to distribute parts from the depot to 49 regional warehouses. At present, the company uses a point-to-point delivery system. Therefore, the company tries to collect orders from each destination warehouse until there is a full truckload or the lead time to reduce the transportation cost is met. However, a full truckload may not be achieved when the destination warehouse has to wait for parts for a very long time. This research proposes a new delivery system that collects the orders from every destination warehouse and creates a cluster of destination warehouses and optimal vehicle routes, which is considered as a HFOVRP. The mathematical model is formulated by employing a modified heterogeneous fleet vehicle routing problem. The solution algorithm starts with clustering the warehouses, dealing with overload demand, and solving problems by using a branch-and-cut algorithm. After performing the numerical experiments, the results show that the transportation cost was reduced by 15.02%.