%0 Journal Article %A Sheng-Shu Cheng and Fong-Jung Yu %D 2013 %J International Journal of Industrial and Manufacturing Engineering %B World Academy of Science, Engineering and Technology %I Open Science Index 78, 2013 %T A CUSUM Control Chart to Monitor Wafer Quality %U https://publications.waset.org/pdf/418 %V 78 %X C-control chart assumes that process nonconformities follow a Poisson distribution. In actuality, however, this Poisson distribution does not always occur. A process control for semiconductor based on a Poisson distribution always underestimates the true average amount of nonconformities and the process variance. Quality is described more accurately if a compound Poisson process is used for process control at this time. A cumulative sum (CUSUM) control chart is much better than a C control chart when a small shift will be detected. This study calculates one-sided CUSUM ARLs using a Markov chain approach to construct a CUSUM control chart with an underlying Poisson-Gamma compound distribution for the failure mechanism. Moreover, an actual data set from a wafer plant is used to demonstrate the operation of the proposed model. The results show that a CUSUM control chart realizes significantly better performance than EWMA. %P 1183 - 1188