Aplicação de métodos de roteirização de rotas para minimização de custos em uma empresa de distribuição de equipamentos de tecnologia no Rio Grande do Norte

Abstract

In an increasingly competitive and dynamic environment, operational efficiency is a key factor for the success of companies, especially in the technology sector. Logistics, and more specifically transportation, accounts for a significant portion of operational costs, directly impacting the competitiveness of organizations. In the context of distributing telecommunication equipment in the state of Rio Grande do Norte, companies face challenges related to delivery route inefficiencies, which increase costs and compromise service capacity. Within the framework of optimization in Operations Research, this study addresses the Vehicle Routing Problem (VRP) and aims to implement, compare, and analyze different methods to solve the VRP, seeking to optimize route planning and reduce operational costs, particularly in complex logistical scenarios, with the goal of improving operational efficiency in the distribution of telecommunication equipment in the state of Rio Grande do Norte. To achieve this goal, two distinct methods were implemented: the Nearest Neighbor heuristic and the Tabu Search metaheuristic, which were compared with the company's current method. The analysis of the methods was based on the following criteria: total route cost (expressed in meters), computational time, and the quality of the solutions obtained. The methodology used in this study included structured steps: data collection, data analysis and preprocessing, algorithm implementation, and result validation. Additionally, a literature review was conducted to provide theoretical support. To evaluate the performance of the methods across instances of varying sizes, different scenarios with varying dimensions (5, 10, 20, 25, 40, and 167 points) were tested. The results indicated that the empirically-based method currently used by the company provides good solutions for smaller instances but presents limitations in larger-scale problems. The Nearest Neighbor heuristic stood out for its fast execution time but generated lower-quality solutions. Tabu Search, on the other hand, proved to be more balanced, offering near-optimal quality solutions with moderate computational times, even in more complex scenarios.