Meaning
Quality monitoring techniques are employed to identify a permanent change in the average value of a measured manufacturing parameter. Implementing process mean shift detection allows production engineers to identify drift in component placement or solder paste volume before the process produces defective boards. It does not apply to random, short-term variations that do not affect the long-term stability of the manufacturing line.
Statistical Monitoring
Control charts are utilized to plot the average value of critical parameters over time. When utilizing process mean shift detection, statistical algorithms analyze the patterns of consecutive data points to determine if the mean has moved from the target. A run of several consecutive points on one side of the center line is a typical indicator that a shift has occurred.
This statistical approach helps distinguish between natural variation and systemic tool wear.
Control Limit
Calculation of action limits is based on the historical standard deviation of the stable process. If a shift is detected, it indicates that the process is no longer in statistical control.
Corrective Action
Maintenance operations are scheduled immediately when a shift is detected to recalibrate the affected machines. This proactive response prevents the generation of scrap material and reduces rework costs.
