Abastecimento de linhas de produção enxutas: uma abordagem de roteamento de veículos para o sistema milk run

• Ano de defesa: 2024
• Autor(a) principal: Falleiros, João Paulo Barbosa
• Orientador(a): Toso, Eli Angela Vitor
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Universidade Federal de São Carlos; Câmpus Sorocaba
• Programa de Pós-Graduação: Programa de Pós-Graduação em Engenharia de Produção - PPGEP-So
• Palavras-chave em Português: Problema de roteamento de veículos; Indústria automotiva
• Palavras-chave em Inglês: Lean logistics; Milk run; Vehicle routing problem; Automotive manufacturer; Logistica lean
• Área do conhecimento CNPq: ENGENHARIAS::ENGENHARIA DE PRODUCAO::PESQUISA OPERACIONAL::PROGRAMACAO LINEAR, NAO-LINEAR, MISTA E DINAMICA
• Link de acesso: https://hdl.handle.net/20.500.14289/21495

The milk run system is inserted in the context of Lean Logistics, the logistics dimension of the Lean Production System. The combined pick-up between different suppliers characterizes it, optimizing the material’s transportation and storage costs. Milk-run can be treated as a vehicle routing problem (VRP), one of the most explored problems of Operations Research in supply chain management. In this study, we address the Milk Run Routing Problem with Progress Lane (MRPPL), which consists of a VRP variant that models a milk-run system in the context of Lean Logistics. This problem encompasses the trade-offs between the lean performance criteria, such as pick-up frequency, vehicle efficiency, and inventory level at the manufacturer’s plant. We present a literature review on the basic concepts of Lean Manufacturing applied to the logistics of the milk run pickup systems and the state-of-the-art VRP and its application to the automobile industry. Additionally, we develop a Systematic Literature Review (SLR) to identify the characteristics and peculiarities of the VRP application to lean logistics. To model and solve the addressed problem, we propose two mixed integer linear programming formulations, one based on a previous study from the literature, and the other originally proposed in this study, considering different types of decision variables and constraints. The results of computational experiments show that applying the second formulation considerably reduces the computation time to solve realistic instances. On average, the total computation time to solve an instance using the second formulation is 8% of the total time for the first formulation. We provide managerial insights considering several scenarios and different variations of parameters in the real problem presented in the case study. An optimal solution could be found for the real problem, indicating a low number of vehicles required and an increase in their efficiency.