System reliability evaluation of a labor-intensive flow shop network considering learning effect and human error rate

Abstract Labor-intensive flow shops are common in many manufacturing systems. In these environments, the efficiency of human workers can vary due to factors such as fatigue, injury, or psychological state, leading to multistate production capacities. Furthermore, the repetitive nature of tasks in a flow shop induces a learning effect, where workers' efficiency improves with experience, reducing the processing time per unit. In this paper, we model such a system as a multistate labor-intensive flow shop network (MLFSN) and use system reliability as the primary performance metric. System reliability is defined as the probability of meeting a specified demand within a given time constraint. The proposed approach first establishes a probability distribution for processing times and then dynamically transforms it into a capacity distribution using a learning curve model that accounts for the number of repetitions. The model is further extended to incorporate the Human Error Rate (HER), which also decreases as a function of the learning effect. Considering HER necessitates producing additional units to meet demand, which in turn impacts system reliability. The proposed algorithm reevaluates system reliability by integrating both the learning effect and HER, providing managers in the MLFSNs with a more accurate and realistic tool for decision-making and production planning.