Group sequencing a PCB assembly system via an expected sequence dependent setup heuristic

In circuit board manufacturing, the production sequencing problem with sequence dependent setups is complicated by the fact that the amount of setup required depends on not only the setup directly before the current setup, but can depend on all of the preceding setups. We present a case study that examines the use of a heuristic for estimating the expected number of setups from the sequence dependent setups. Our method is based on estimating the expected number of setups which may occur given a board-feeder setup configuration. Once estimates for the sequence dependent setups are established, they can be used to measure the similarity between boards in clustering algorithms and in nearest neighbor heuristics for group sequencing. The method is tested on an actual printed circuit board assembly system. Then, using a simulation of the assembly system, we compare the sequences generated using the expected number of setups distance measure and the Hamming distance measure to optimal sequences. Our results indicate that grouped sequences generated by using the expected number of setups had significantly better makespan performance when compared to sequences based on the more traditional Hamming distance for the particular system under study. The significant gains in makespan resulted in only moderate increases in work in process and slight increases in manual station utilization.