Application of Binary Integer Programming for the Balancing of an Automotive Production Process

Abstract

Line balancing allows activities to be distributed among workstations while considering time and precedence constraints. This study analyzes the balancing of an automotive assembly process through the sequential application of heuristic methods and binary integer programming. A reference case consisting of 33 activities was used, with a total work content of 89.28 min and a takt time of 18.018 min. Initially, the longest candidate rule produced a feasible six-workstation solution, which was then used to limit the number of workstations considered in the binary formulation and reduce the number of decision variables from 1,089 to 198. The model was implemented in R and also produced a six-workstation solution with the same task assignment. As a complementary analysis, the Helgeson-Birnie method also resulted in six workstations, although with partial differences in task allocation. The results show that different balancing strategies may converge to the same number of workstations and that an initial heuristic solution can be used to reduce the size of the binary formulation. From a practical perspective, this approach can support decision-making in automotive assembly processes by facilitating the evaluation of workstation configurations and the identification of opportunities for a more balanced use of production resources.