We all learned to use Statistical Process Control (SPC) methods to monitor processes, look for trends, and be forewarned if our processes begin to perform in ways we don’t want them to. However, if we are practicing SPC we can also use out-of-control data points to show us new opportunities and ways we can improve, if we can replicate the conditions that drove an out-of-control event in our favor.
The Rest of the Story:
Many businesses, particularly those with sensitive manufacturing and fabrication practices, use Statistical Process Control (SPC) methods to monitor and manage those processes. SPC is a powerful tool for preventing and eliminating waste, and for minimizing or managing process variation.
SPC shows us increases or changes in variation in our processes before they become problems or create defects. A disciplined practice helps us plan maintenance, make timely adjustments to process controls, and prevents us from panicking or over-controlling our processes. Historical process data helps us design new products that are optimized for manufacturability.
There is something that SPC can also identify, but not everyone recognizes. Anomalies, or “out-of-control” events can sometimes show us the potential performance of our processes, referred to as “entitlement.” If you don’t already use your SPC discipline to identify your process entitlement, then herein is a simple explanation.
First, for those who are not already expert in SPC, a brief definition is needed. If we collect data on the key output of a process, such as yield or a critical dimension, after several data points are available, we can use statistical mathematics to predict a range within which the next data point should fall. We call upper and lower edges of this range “control limits.” If the next data point does indeed fall within this range, we conclude that the process is performing as expected.
As we collect more data, we can refine our range. If a data point falls outside of that range, then we conclude that something has happened that is unexpected, we investigate, and address it appropriately. We say a data point outside the limits to be “out of control” indicating we may have lost control of our process. There is more to the science of SPC, but the above explanation covers the root methodology.
Now, let’s discuss entitlement. Normally, when we find a data point that is outside of our control limits we investigate the cause and determine if the cause is persistent and we should make adjustments to get the process back into control. However, what if the metric we are tracking is yield, for example, and that out-of-control point shows an unusually high yield? A high yield is a good thing.
This out-of-control data point in a beneficial direction is showing us what can be produced by our process, if we can figure out how it happened. That’s entitlement. Of course, we should investigate that data point as the SPC methodology recommends, but what we want to find out is how to make it happen again.
If we can figure out the magical combination of process settings, or the noise element (unintended influence) that inadvertently played a role, and if we can manage or control the combination or noise element, we might get our process to produce the unusually high yield repeatedly.
If you practice statistical process control, but haven’t made it a practice of seeking those entitlement opportunities, go back to your recent historical data and charts. If you retained your findings on your out-of-control point investigations, you may be able to use your history to begin finding and achieving your process entitlement.
Make it a practice of looking for entitlement in your processes. There is power to be had in understanding and capitalizing on accidents. Don’t forget to look for other forms of out-of-control indicators such as trends and shifts.
If you don’t practice statistical process control, give it some serious thought. There is a great deal of benefit to be had from predictable process outputs, and more to be had from being able to see your process’s potential and know how to achieve it consistently.
When we develop and learn methodologies, it’s easy to limit our focus on its primary purpose. In the case of SPC, we limit our focus on maintaining process stability and control. It’s easy to forget that the same methodology can show us opportunity and potential. Don’t let your team forget. Take advantage of the possibilities of both your tools and your processes.
Stay wise, friends.