CART–BPN approach for estimating cycle time in wafer fabrication

Abstract

Cycle-time management plays a crucial role in improving the performance of a wafer-fabrication factory, beginning with the estimation of the cycle time of each job. Although this topic has been widely investigated, several problems still need to be addressed, such as how to classify jobs suitable for the same estimation mechanism into the same group. Most existing methods classify jobs by their attributes; however, the differences between the attributes of various jobs may not be reflected in their cycle times. The biobjective nature of a classification and regression tree (CART) makes it particularly suitable for resolving this problem. However, in a CART, the cycle times of jobs of a branch are estimated with the same value, which is inexact. Hence, this study proposes a joint use of a CART and back propagation network (BPN), in which the BPN is constructed to estimate the cycle times of jobs of a branch. A real case was used to evaluate the effectiveness of the proposed methodology. The experimental results supported the superiority of the proposed methodology over existing methods. In addition, the managerial implications of the proposed methodology are also discussed.