Minimizing makespan and flowtime in a parallel multi-stage cellular manufacturing company

• Published in: Robotics and computer-integrated manufacturing Vol. 72; p. 102182
• Main Authors: Saraçoğlu, İlkay, Süer, Gürsel A., Gannon, Patrick
• Format: Journal Article
• Language: English
• Published: Oxford Elsevier Ltd 01.12.2021; Elsevier BV
• Subjects: Cell loading; Cellular manufacture; Cellular manufacturing; Finishing; Flow shop scheduling; Genetic algorithm; Genetic algorithms; Injection molding; Job shop scheduling; Job shops; Manpower; Manufacturing; Manufacturing cells; Mixed integer linear programming; Packaging; Cellular manufacturing; Mixed integer linear programming; Genetic algorithm; Flow shop scheduling; Cell loading
• ISSN: 0736-5845; 1879-2537
• DOI: 10.1016/j.rcim.2021.102182

•We formulate an approach for a parallel multi-stage cellular manufacturing system.•Makespan and total flowtime found in reasonable time for large problems.•Genetic algorithm is proposed to improve the efficiency of the cell loading process.•Approach verified by shoes manufacturing case study to yield.•More effective approach offered by using multi-stage solution ways. This study proposes a 3-phase solution approach for a multi-product parallel multi-stage cellular manufacturing company. The study focuses on a case study involving a shoe manufacturing plant in which products are produced according to their due dates. The investigated manufacturing process has three stages, namely lasting cells, rotary injection molding cells, finishing-packaging cells. System performance is measured based on total flowtime and makespan. We propose a 3-phase solution approach to tackle the problem; 1) the first phase of the proposed approach allocates manpower to operations in the lasting cells and finishing-packaging cells, independently. The objective is to maximize the production rates in these cells. 2) The second phase includes cell loading to determine product families based on a similarity coefficient using mathematical modeling and genetic algorithms (GA). The proposed GA algorithm for cell loading performs mutation prior to crossover, breaking from traditional genetic algorithm flow. The performance measures flow time and makespan are considered in this phase. 3) Flow shop scheduling is then performed to determine the product sequence in each (lasting, rotary injection molding, finishing-packaging) cell group. This 3-phase solution approached is repeated with alternative manpower level allocation to lasting and finishing-packaging cells where the total manpower level remains the same.