It’s 10:00pm; do you know where your control limits are?
You have spent a great deal of time and money implementing Statistical Process Control. You have the best SPC software, calibrated and certified measurement equipment, an excellent quality staff that follows the testing standards to the letter, but are you really getting any benefits from your control charts? If you aren’t seeing the improvements that you would like, maybe you should check to see where your control limits are.
If your control limits are based on anything other than actual process data then you probably aren’t seeing any benefits! Control limits must be set from a measure of the true process variation. Limits based on anything else or limits that are “automatically” calculated, will not give you the increase in process performance that a control chart was designed to give.
Control limits should be set from a measure or approximation of the process variation, such as standard deviation or sigma, to truly be able to monitor the process behavior. The limits should only be adjusted when the process behavior changes through a shift of the mean or through an increase or decrease in the variation of the process itself.
Anytime that the process shows a decrease in variation, we must recalculate the control limits so that the true process behavior can be monitored. Only then can we maintain the improvement and make the process even better!