A constraint relaxation-based intelligence algorithm for workshop layouts with automated guided vehicle systems

With the widespread adoption of automated guided vehicle (AGV) systems for material handling in manufacturing plants, it has become practical and crucial to delve into the layout problem associated with AGV systems. In this work, we focus on a unique layout problem encountered in a hybrid workshop where AGV systems are employed for transporting semiproducts along the manufacturing line. Several distinctive features in this system contribute to the challenge of the problem. Notably, manufacturing occurs in an uncertain environment, and certain manufacturing cells may produce semiproducts that do not meet quality standards, necessitating repair. Additionally, each AGV requires recharging in a designated area within the workshop. Given that the proposed layout problem is NP-hard, we present an intelligence variable neighborhood search heuristic integrated with a constraint relaxation strategy to address its complexity. The numerical results demonstrate the algorithm's ability to generate high-quality solutions within a reasonable timeframe, even for large-scale test instances. The layout solutions obtained through our algorithm outperform those produced by the CPLEX solver and the practical layouts devised by the company. This highlights the efficacy of our approach in tackling the unique challenges posed by the layout problem in a hybrid workshop with an AGV system.