A Skill-Based MILP Model in Cellular Manufacturing Systems with Human-Robot Collaboration

• Published in: IFAC-PapersOnLine Vol. 55; no. 10; pp. 1728 - 1733
• Main Authors: Yetkin, Busra Nur, Ulutas, Berna Haktanirlar
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 2022
• Subjects: automation; cellular manufacturing systems; human-robot collaboration; mathematical programming; robot assignment; worker assignment; worker assignment; automation; human-robot collaboration; robot assignment; cellular manufacturing systems; mathematical programming
• ISSN: 2405-8963; 2405-8963
• DOI: 10.1016/j.ifacol.2022.09.647

Worker assignment in cellular manufacturing systems (CMS) has become more important for efficiency and flexibility under qualified workforce restrictions. Researchers primarily focus on assigning skilled workers to operations to optimize capacity and related costs. On the other hand, the advantages of automation in human-operated manufacturing cells are well known. However, some manufacturing operations require high customization and consequently depends on skilled workers. Therefore, complete automation is not desirable and a manufacturing system, with both manual and automated resources, can be the best configuration. To support this design, cobots can present suitable solutions to work side by side with humans. In this study, a stochastic mixed-integer linear programming (MILP) is introduced to minimize the total cost of robots’ operating cost and worker employment cost in CMS when assigning resources to cells which also considers technical skills and non-technical skills. The operation times of workers are considered as stochastic, while the operation times of robots are deterministic. The demands of the parts are assumed uniformly distributed and stochastic. The result for the optimization problem is presented on the small-sized test data. The proposed MILP model is solved in the GAMS modeling environment. The calculation of collaboration level for each operation is novel in this study and enables the best possible resource assignment based on skill-level. A small-sized example is used to test the impact of constraints. Results confirmed the success of the model as a sufficient tool to simultaneously assign workers and robots with skill-related constraints.