An integrated mixed-integer programming and metaheuristic approach for ergonomic cobot-assisted assembly line balancing

Following Industry 4.0, human–robot collaboration has increasingly been integrated into assembly systems to improve flexibility and productivity, while Industry 5.0 emphasizes human-centric manufacturing, worker well-being, and sustainability. To the best of our knowledge, this is the first study in human–robot collaborative assembly line balancing to incorporate the Occupational Repetitive Actions (OCRA) Index and solve a cobotic assembly line balancing problem (CALBP) considering ergonomic performance. A sequential multi-criteria solution approach is developed to optimize both the number of workstations and ergonomic risk. First, a mixed-integer linear programming (MILP) model minimizes the number of workstations under human–cobot assignment and scheduling constraints. The resulting configurations are then evaluated using the OCRA Index and explored through an iterative MOSSA procedure combined with a Compromise Programming metric. For larger instances, a hybrid metaheuristic framework integrating MILP-based feasibility control with Simulated Annealing, Particle Swarm Optimization, Jaya, and Genetic Algorithm is proposed. Computational results show that the proposed approaches improve ergonomic performance while maintaining competitive workstation counts, demonstrating the potential of cobot-assisted assembly line balancing to jointly enhance operational efficiency and worker well-being.