1 Curve Monitoring for a Single-Crystal Ingot Growth Process

In conventional statistical process control, a conforming process is usually assumed to be stable with a constant mean value. However, dynamic data streams of variables over time are widely encountered in many complex manufacturing processes. In this paper, we study the ingot growth process, which exhibits a time-varying mean and variance even if the process is in-control. We propose different charting schemes for the statistical monitoring of the process. The performance of the new charts are compared via simulation. Implementation of the newly proposed chart to the real data shows that the chart is effective in detecting process failures in the growing curves. Key words: control chart, manufacturing process, quality control, statistical process control 1