Adaptive worker assignment for the flow shop scheduling with human diverse behaviors

The rapid growth of the fast fashion industry has necessitated business models prioritizing agile responsiveness to shifting market trends. This demand drives the need for accelerated garment design, production, and distribution cycles. To this end, the Hanging Transportation (HT) system has emerged as a critical component of infrastructure in labor-intensive apparel manufacturing, facilitating efficient product flow and allocation processing. Crucially, scheduling within HT systems presents a distinct challenge compared to traditional flow shop scheduling: the optimal allocation of workers, rather than machine assignment, determines production efficiency. To address the Flow Shop Worker Assignment Problem (FSWAP) in HT systems, this paper proposes a novel aDaptive wOrker aSsignment (DOS) approach. DOS explicitly incorporates workers’ proficiency heterogeneity and dynamic availability while optimizing for completion time efficiency and solution stability. Furthermore, we introduce a Local Optimum Breaking Strategy (LOBS) integrated within DOS to mitigate the local optimum issue commonly associated with evolutionary algorithms. A comprehensive evaluation through theoretical analysis, simulations, and real-world case studies demonstrates that DOS significantly outperforms existing popular methods in terms of completion time, computational efficiency, and solution stability.