Solving Multi-Objective Disassembly and Assembly Line Balancing Problem with Worker Fatigue Level Using Multi-Agent Proximal Policy Optimization Algorithm

The recycling process plays a pivotal role in disassembly and assembly systems, whether isolated or integrated. However, traditional line balancing methods optimize physical processes but lack human-oriented workload adjustments. In this work, we propose a novel approach to mitigate the worker fatigue by alternating between high-intensity and low-intensity tasks. A multi-object mixed integer programming model is developed to maximize profit while minimizing fatigue levels. To solve this problem, we adopt a reinforcement learning algorithm, which is Multi-Agent Proximal Policy Optimization (MAPPO). Comparative performance analysis with four other intelligent algorithms demonstrates the superior effectiveness of the proposed approach in addressing the model.