Workforce planning in mixed model assembly systems

Serial assembly systems are formed by arranging several production cells or stations in series. We study a popular class of serial assembly lines where all stations have the same production cycle. We address a workforce planning problem for such lines which finds applications in labor-intensive operations in automobile, fire engine, aircraft, and PC board assembly. The problem presented can be applied to lines that produce several variations of a basic stable design; i.e., mixed model transfer lines. Given a set of n jobs, we want to find a sequence that minimizes the maximum workforce requirements over all production cycles. An optimal polynomial algorithm for the two-station line is presented, and the three-station case is proved to be strongly NP-complete. Several heuristic algorithms that produce upper and lower bounds are developed for the general problem. Worst case behavior of the upper bounds, as well as average performance of lower and upper bounds, are reported. Computational results show that some of the heuristics produce near optimal solutions. As an extension of the basic model we exploit the tradeoff between cycle time and workforce level.