A Rich Vehicle Routing Problem for a City Logistics Problem

In this work, a Rich Vehicle Routing Problem (RVRP) is faced for solving city logistic problems. In particular, we deal with the problem of a logistic company that has to define the best distribution strategy for obtaining an efficient usage of vehicles and for reducing transportation costs while serving customers with different priority demands during a given planning horizon. Thus, we deal with a multi-period vehicle routing problem with a heterogeneous fleet of vehicles, with customers' requirements and company restrictions to satisfy, in which the fleet composition has to be daily defined. In fact, the company has a fleet of owned vehicles and the possibility to select, day by day, a certain number of vehicles from the fleet of a third-party company. Routing costs must be minimized together with the number of vehicles used. A mixed integer programming model is proposed, and an experimental campaign is presented for validating it. Tests have been used for evaluating the quality of the solutions in terms of both model behavior and service level to grant to the customers. Moreover, the benefits that can be obtained by postponing deliveries are evaluated. Results are discussed, and some conclusions are highlighted, including the possibility of formulating this problem in such a way as to use the general solver proposed in the recent literature. This seems to be the most interesting challenge to permit companies to improve the distribution activities.

Automated summary

This work addresses a Rich Vehicle Routing Problem for city logistic problems, aiming to find the best distribution strategy for a logistic company to achieve efficient usage of vehicles and transportation cost reduction while serving customers with different priority demands. The problem involves a multi-period vehicle routing problem with a heterogeneous fleet of vehicles, where routing costs and the number of vehicles used need to be minimized. A mixed integer programming model is proposed and an experimental campaign is conducted to validate its effectiveness. The study also evaluates the benefits of delaying deliveries and discusses the possibility of applying a general solver for this problem as suggested in recent literature.