Proposta de otimização de logística reversa de embalagens de óleos lubrificantes por meio do método Travelling Salesman Problem

Abstract

The national reverse logistics operator for lubricating oil packaging (Instituto Jogo Clean) has shown limitations as to its scope, prioritizing the collection of this material only at authorized stations and vehicle dealerships, leaving countless other generators on the margins of the collection system. In view of this, the present The objective of this study was to propose the optimization of the reverse flow of these enclosures, in municipality of Juazeiro do Norte - CE, through the application of the Traveling Salesman Method Problem (TSP). In this sense, a diagnosis of the current situation of destination of packaging and mapping of generating points of these materials. took place also the selection of a priority area for the implementation of a Voluntary Delivery Point (PEV) through the Analytic Hierarchy Process (AHP) multicriteria method. The area selected by AHP was defined as the starting point for the door-to-door collection system of packaging. In addition, three scenarios were defined for the simulations of the optimized routes (application of the TSP method) for itinerant collections: scenario (i) – single collection in all generating points; scenario (ii) – clustering of generators into two groups, dividing thus the municipality in two different collection regions; and, scenario (iii) - clustering of points mapped into an optimal number of clusters. The results demonstrated that approximately 51.0% of the projects carry out oil maintenance and/or replacement only in motorcycles, most of which are family businesses. On average, more than 5,000 packaging units are generated and discarded weekly. However, around 19.0% of these establishments still dispose of the wrappers as waste common. As for the choice of the area for the implementation of the PEV, a property with average distance to the generating points of approximately 1,890.3 m. The scenario (i) presented the smallest optimized route, with a length of 55,234.8 m, followed by scenario (ii) whose increase in total path length was 0.35%.