Desenvolvimento de um sistema de roteirização de veículos baseado em uma meta-heurística de algoritmo genético: um caso envolvendo frotas heterogêneas e janelas de tempo

Abstract

Optimizing the logistical operations of organizations is of great importance to keep the service level and costs balanced. Therefore, vehicle routing systems become extremely efficient tools to reach such levels. This means having an agile planning that minimizes costs within the operational conditions while meeting customer requirements. This work presents a complex routing problem, with heterogeneous fleet and time windows. Therefore, an optimization model based on a Genetic Algorithm was proposed. This algorithm is characterized by evaluating population groups of solutions following an evolutionary iterative stochastic process. Such an algorithm can show both local and global optimal optimization behaviors. Data on geographic locations and routes between customers were obtained through the use of free APIs provided by Google Maps©. The company studied is a producer and supplier of frozen food products that serves more than 200 customers in the region of the Minas Gerais and southern Goiás triangle. Within the study proposal, the optimizations achieved a reduction of 1,900 km per month, which corresponds to a reduction of mileage driven of approximately 13.50%. These values represent approximately R$ 18,300.00 in annual costs. Therefore, the main advantages for the company were the reduction in fuel costs and shorter labor time for drivers.