Exact method for the vehicle routing problem with mixed linehaul and backhaul customers, heterogeneous fleet, time window and manufacturing capacity

Due to concerns over food safety, quality, and transparency, the food industry has been challenged to deal with an increased complexity in its logistics and production planning. This complexity is mostly associated to the task of optimally exploring a highly constrained solution space. This paper addresses the problem of designing a set of vehicle routes satisfying the delivery and the collection requirements of a set of geographically scattered linehaul and backhaul customers. Each customer has a known demand for delivery and/or pickup, and it must be serviced within a predefined time window, representing the earliest and the latest times service time, by a vehicle of a heterogeneous fleet of depot-returning capacitated vehicles. Additionally the proposed model, aiming at minimizing the logistics costs, also takes manufacturing capacities and driving hour's regulations into consideration. The proposed algorithm was embedded in a native application. This digital tool, enabling the support of a holistic production and logistics planning, was implemented in a food company.