In modern manufacturing environments, maintaining production continuity and cost efficiency critically depends on the seamless coordination of interconnected processes. However, bottlenecks disrupt this coordination, leading to capacity losses, workflow interruptions, and increased operational costs. This study addresses a critical bottleneck observed in the slitting stage of a paper bag manufacturing system, where sequence-dependent setup times significantly affect operational efficiency. The problem is formulated as a single-machine scheduling problem with sequence-dependent setup times, explicitly incorporating both learning and deterioration effects to capture human and system dynamics. While the learning effect reflects performance improvements due to operator experience, the deterioration effect represents performance degradation caused by fatigue and interruptions. The integration of these opposing dynamics results in a more realistic yet computationally challenging NP-hard problem structure. To effectively solve this problem, a Genetic Algorithm (GA)-based solution framework is developed. Given the sensitivity of GA performance to parameter settings, a full factorial experimental design is conducted to systematically analyze parameter interactions and identify the most effective configuration. The proposed approach is validated using real-world production data consisting of 66 jobs derived from Pareto analysis. The results demonstrate that the proposed model not only captures the complexity of real production environments but also provides a robust and practical decision-support tool capable of significantly improving production flow efficiency.

Keywords

Single machine scheduling, Sequence-dependent setup times, Learning and deterioration effects, Genetic Algorithm.