关键词: SERU生产, 生产柔性, SERU系统构建, NSGA-Ⅱ, 工人配置

Abstract: Due to the fierce global market competition, the manufacturing environment shows characteristics like diversified customer demands, smaller batch sizes, and shorter product lifecycles and lead times. This drives manufacturing companies to transform their systems for improved flexibility and responsiveness. A SERU system consists of one or multiple Serus, which comprises multi-skilled workers and flexible resources such as a simple and movable equipment. Such a system can be rapidly and frequently constructed, reconfigured, dismantled, and reconstructed. The construction of SERU systems is a critical step in achieving flexibility and a focal point of research in SERU production. Existing research predominantly focuses on Task-Oriented SERU system formation (TOSF), which assumes specific production tasks are known. However, such strategies are difficult to avoid frequent reconstructions in dynamic and evolving market environments. To address this issue, some scholars have considered the dynamic nature of demand in SERU system formation. However, they have not focused on the flexibility performance, resulting in the SERU systems lacking flexibility. There is still a significant shortage of research on flexible SERU systems formation. Many related studies have shown that multi-skilled workers are an important factor in achieving flexibility. The higher the level of multi-skilling among workers within the system, the greater the system's flexibility. However, cross-training is both a time-consuming and costly activity. In practice, training for multi-skilled workers has become increasingly challenging. Therefore, the existing literature based on the assumption of full-skilled workers is difficult to apply in SERU systems. To solve this problem, many manufacturing companies have adopted some transferable full-skilled workers (TFSW) strategies, in which only a small number of workers are fully cross-trained and allowed to move between serus. In TFSW strategies, capacity can be dynamically shifted from any serus to any other by TFSWs within the serus. This makes them very robust to fluctuations in workloads (e.g., due to temporary shifts in product mix) or staffing levels (e.g., due to absenteeism). While some literature has examined the flexibility achieved through cross-seru worker transfers, their primary focus is on workforce scheduling within a given SERU system. Therefore, studying how to construct a flexible SERU system by considering transferable full-skilled workers is of practical significance.
This paper addresses the Flexible SERU System Formation Problem based on Transferable Full-Skilled Workers (FSFP-TFSW). A multi-objective model (FSFP-TFSW) is established to minimize the training costs for skills and the expected total completion time of orders. The complexity of the problem is analyzed and the NSGA-II algorithm is designed to solve it. The effectiveness and superiority of the FSFP-TFSW model is verified by comparing its performance with the non-transferable full-skilled workers strategy (NTS) and the fully flexible SERU system formation (FULL) strategy based on workers who are full-skilled under different demand scenarios. The results show that the proposed FSFP-TFSW model can better balance cost and responsiveness by configuring a reasonable number of transferable full-skilled workers, more suitable for SERU system formation in dynamic demand environments. The experimental results have shown that FSFP-TFSW can achieve a balance between cost and responsiveness.
As a flexible SERU system formation strategy, the solutions obtained by the FSFP-TFSW model achieve a better balance between training costs and responsiveness. However, this article does not consider the TFSWs scheduling or the internal configuration of serus. How to simultaneously consider SERU system formation, SERU system scheduling, and TFSWs scheduling is a problem worth further research in the future.

Key words: SERU system, flexibility, SERU system formation, NSGA-Ⅱ, worker allocation