Hydrokube: A case study of demand forecasting and Vehicle Routing Problem modelling

Details

Supervisors

Catanzaro, Daniele

Faculty

Louvain School of Management

Degree label

Master [120] : ingénieur de gestion, à finalité spécialisée

Abstract

Water hardness is a cause of several problems in everyday life. Hard water reduces indeed the life of household appliances, increases energy consumption and can affect skin health. A solution to these issues is provided by the Belgian company Hydrokube, which offers CO2 water softeners to individuals and companies. The company is growing rapidly, but faces currently two main challenges. First, the global organization of the company needs to be improved by forecasting the future demand of Hydrokube's products. Second, the way CO2 water softeners are delivered to customers needs to be optimized to reduce distribution costs. Hydrokube manages indeed itself the delivering of its products. This report addresses both challenges. The first part of the report addresses the temporal and spatial forecasting of the demand. While the temporal forecasting of the demand is done using a linear regression model, the spatial forecasting is done using the existing customers' locations and the map of water hardness in Belgium. The results found in this part show that Hydrokube can expect 243 new customers between February 2022 and January 2023, located for most of them in and around Brussels. In the second part of the report, the optimization of the distribution is studied by applying the Vehicle Routing Problem (VRP). With Hydrokube having only one depot and one plumber, the VRP degenerates into a Travelling Salesman Problem (TSP). Since the TSP is NP-hard, various heuristics are explored to approximate the optimal solution. The results found in this part show that applying the 3-Opt algorithm to the Nearest Neighbour (NN) initial solutions provides, in the case of Hydrokube, the best results.