Can task clustering and intelligent allocation enhance the trade-off between ergonomic and temporal targets in...

Achieving both temporal efficiency and optimal human factors concurrently in U-shaped production lines poses a significant, multifaceted challenge. Proposing a bi-objective framework, this article concurrently aims to minimize ergonomic risks and optimize production cycle time. The proposed method offers a flexible and adaptable model for real-world scenarios by employing activity clustering into work families, an ergonomic penalty coefficient for disproportionate assignments, and workload balancing across workstations. To solve this model, two complementary methods were developed: first, an exact solution using a branch-and-bound algorithm implemented in Lingo; and second, an innovative heuristic algorithm called Ergonomic-Aware Clustered Prioritization with Deep Learning (EACP-DL). EACP-DL combines clustering, an ergonomics-time scoring system and deep learning for automatic adjustment of workstation allocation coefficients. Results demonstrate that the proposed algorithm reduces ergonomic risks by 12% compared to the branch-and-bound method, with a time increase of less than 6%, exhibiting rapid execution and scalability.