A Regression-Based Cycle Time Estimator for the PCB Placement Machine

The assembly of printed circuit boards is conducted by the sophisticated PCB placement machines. The cycle time for a machine to complete an assembly job is determined by solving two machine optimization problems, i.e., the feeder arrangement and placement sequencing.Rapid and accurate estimation of these cycle times is important for solving the higher level planning problems including the line balancing problem. In this paper, a regression-based estimator was established to estimate the cycle time. Statistical results on 100 PCB samples showed that the proposed estimator could estimate the cycle times accurately. At the same time, the computationally efficiency makes it suitableto be used in solving higher level planning problems.