Adaptive Tabu Search for Flexible Assembly Line Balancing

Abstract This research presents the development and evaluation of the Adaptive Tabu Search for Flexible Assembly Line Balancing (ATS-FALB), a metaheuristic algorithm designed to optimize task assignment in flexible automotive production lines. Assembly line balancing problems are computationally complex and classified as NP-hard, making optimal solutions difficult to obtain in large-scale manufacturing systems. To address this challenge, the proposed algorithm integrates adaptive mechanisms based on dynamic long-term memory structures and advanced penalty strategies that guide the search process and reduce the risk of premature convergence. Experimental validation was conducted using a benchmark dataset of 30 vehicle models representing realistic automotive production scenarios. The results indicate that the ATS-FALB algorithm consistently produces optimal or near-optimal solutions while maintaining stable convergence behavior. In particular, the method achieves an average reduction of 0.14 hours in cycle times and reaches high-quality solutions in fewer than 100 iterations. These findings demonstrate the robustness and efficiency of the proposed approach for improving operational performance in manufacturing environments.