Restrict-and-Fix Heuristics for Large Robotic Assembly Line Balancing Problem.

The rapid advancements in robotics and automation have revolutionized the manufacturing industry, introducing new complexities in the design and balancing of robotic assembly lines. The Robotic Assembly Line Balancing Problem (RALBP) extends the classical assembly line balancing problem by incorporating equipment selection for each workstation. In other words, RALBP requires tasks to be grouped based on shared equipment needs. As RALBP is an NP-hard problem, efficient metaheuristics or heuristics must be developed to design realistic assembly lines because such lines can span several kilometers. This paper tackles these challenges by introducing a novel Restrict-and-Fix (RF) method, inspired by the Relax-and-Fix framework. The proposed heuristic decomposes the RALBP into smaller, manageable subproblems, solving them iteratively to find good quality solutions within reasonable computational times. Numerical results demonstrate the method’s strong performance and scalability. Furthermore, as the method relies on mixed-integer linear models, it can easily be extended to account for a wide range of practical constraints.