Determining safety stocks for production planning in uncertain manufacturing

When computing production plans and available-to-promise statements, the randomness of flow times and yield rates are the two major sources of uncertainty. Considering both of these uncertainties, this study employs a simple formula for computing the standard deviation for cumulative good product output of a particular product type at a particular time. Using this standard deviation, the necessary safety stock for that product type at that time can be obtained. This study was motivated by semiconductor wafer manufacturing, which requires many process steps performed by highly unreliable equipment. Such an environment causes the means and the variances in flow times to be generally larger than those for other industries. It also makes yield rates uncertain. Thus, when making order quotations and calculating production plans, safety stocks must be used to guard against the uncertainties of the manufacturing processes to ensure on-time-delivery ratios and to provide better customer service.