Wafer fabrication is a complex, costly and lengthy process that involves hundreds of process steps with monitoring of the corresponding process parameters at the same time to enhance the yield. Large amount of data is automatically collected during these processes in wafer fabrication facility. Thus, potential useful information can be extracted from huge data sources to enhance decision quality and enhance operational effectiveness. This study aims to develop a framework to integrate FDC and MES data and then propose an approach based on data mining and time series techniques to investigate the data in order to enhance the overall usage effectiveness (OUE) for cost reduction. We validated this approach with an empirical study in a semiconductor company in Taiwan and the results demonstrated the practical viability of this approach. The extracted information and knowledge is helpful to engineers for identifying the major tools factors affecting indirect material usage effectiveness as well as for indentify periods of time when a specific tool is working using either low or high quantity of material.
HASH(0x1b8e15a0)