Integrated model assignment and multi-line balancing in human-robot collaborative mixed-model assembly lines

Abstract

The adoption of collaborative robots in assembly lines is increasingly widespread due to their ability to work alongside human operators. This study introduces a simultaneous model-line assignment and robotic mixed-model multiple assembly line balancing (MLA-RMMALB) problem, formulated as a multi-objective mathematical model. The objectives are to (i) minimize production costs, (ii) reduce the energy consumption of cobots, and (iii) evenly distribute the physical workload among operators. A numerical example with 21 tasks is used to analyze the impact of these objectives. The example is first solved separately for each objective and then as a combined multi-objective function. A numerical example with 21 tasks is used to evaluate the effects of the objective functions, first solving each separately and then as a multi-objective function. The results indicate that the model provides high-quality solutions for small instances, with solution times directly influenced by the type of objectives and the problem size. Model assignments, the number of active lines and workstations, and the allocation of operators and cobots vary significantly depending on the objective function. Notably, all models can be produced on a single line with a slight increase in cycle time. To assess the model's performance, benchmark problems with increasing task numbers are analyzed. Findings reveal that due to the problem's complexity -especially under strict cycle times and limited CPU capacity- optimal solutions are not always attainable even for small instances, and no feasible solutions may emerge for medium-sized problems. The proposed approach offers valuable managerial insights by enabling decision-makers to simultaneously optimize cost, energy consumption, and workload distribution. Furthermore, the effective assignment and scheduling of heterogeneous operators and cobots enhance production flexibility and resource utilization. Consequently, this study contributes to the literature by integrating model-line assignment with mixed-model robotic assembly line balancing while considering resource heterogeneity, limited resource availability, and collaboration between cobots and operators.