The vehicle routing problem of Masoutis S.A.

Abstract

Optimizing the vehicle routing system, is one of the main challenges a company has to face. A vast number of models have been developed to solve almost every variation of a Vehicle Routing Problem (VRP). In a real life case such a problem consists of way too many constraints and variables. The time needed to find a solution increases exponentially as the problem becoming more complex. As a result, models should provide a good solution in as short time as possible. In this thesis, a Vehicle Routing Problem with Time Windows (VRPTW) is considered. The problem consists of a list of stores that must be serviced by a number of heterogeneous vehicles during predefined time windows. It focuses on a routing problem faced by Masoutis S.A. Two models are developed to solve the problem. A heuristic model trying to find the nearest possible to the exact solution and an exact model trying to find the exact solution. The aim is to compare the time needed for each model to find a solution and which model has found the best solution.