Otimização de rotas de coleta em armazéns com colaboração humano-robô: uma abordagem computacional

Abstract

The picking process is one of the most critical operations in Distribution Centers (DCs), representing a significant portion of operational costs due to the time spent on operator movement. The inclusion of mobile robots, such as AMRs and AGVs, has been developed as an option to enhance these operations. These robots have the potential to reduce unnecessary travel, when applied effectively, thereby increasing the efficiency and productivity of DCs. This study focuses on the development of a computational tool using optimization techniques to define more efficient routes for human operators in a picker-to-parts system assisted by robots. The methodology involves virtual modeling of DC storage areas and the application of a clustering strategy to consolidate multiple addresses into collection points, aiming to optimize the distance matrix and robot movements. Additionally, it employs exact and heuristic methods to solve routing problems through association with TSP and OTSP models. The study evaluates different strategies for executing collection routes, where operators remain in the picking area while robots return to the central depot. The results demonstrate that applying these strategies reduced operator movement by up to 33.61%, with the optimization strategy achieving 96.15% efficiency and a global gap of 3.85% to reach the total optimization of each route. The tool was validated through experiments in a real-world scenario, enabling a comparison between the optimized routes and the actual routes currently used by the company, showing improved performance in the collection process. [resumo publicado pelo autor]