# Process Capability and Process Control – how are they different?

In our Green Belt course I introduce the concepts of process capability and process control.   I find that many students have trouble understanding the difference between the two.  This is compounded by the fact that the output of a process may fall into of four possible combinations of capability and control.
Process capability looks at whether or not the output of the process is capable of satisfying the customer specification or requirement.   Process capability is usually measured using either the Capability Ratio (Cp) or the Capability Index (Cpk).  A value of 1.0 for either Cp or Cpk implies that the spread of the process output is equal to the specification or requirement.  Process spread in this case is defined as six standard deviations, with standard deviation being a statistic that measures variability.  A minimum value of 1.33 for Cp and Cpk is desired, and the larger the value the better.   The formulas for Cpk consider how well the output of the process is centered within the specification; the formula for Cp does not.
In order for us to calculate Cp or Cpk for a process, three conditions must be met.  One, the output of the process must be a continuous variable.  Two, the output of the process must be normally distributed.  Three, the process output must be in a state of statistical control.
Process control looks at whether or not the output of the process is in a state of statistical control.  There is variation in the output of all processes, and this variation may be due to common causes or special causes.  Common causes are inherent in the process, while special causes are assignable to a person or an event.  The variation that is present in the output of the process, whether it is common cause or special cause, may or may not result in defects.
If a process exhibits special cause variation, it is said to be out of statistical control.  This implies that the output is not stable or predictable.   If a process exhibits common cause variation, it is said to be in a state of statistical control.   This implies that the output is stable and predictable.   The fact that a process is in control does NOT mean that the output of the process is normally distributed and no such inference should ever be made.  We determine whether or not a process is in control by using either a run chart or control charts, with control charts being the preferred method.
Dr. W. Edwards Deming stated the following: “…the type of action required to reduce special causes of variation is totally different from the action required to reduce variation and faults from the system itself…” in his book Out of the Crisis, p. 309.  In order to make the right decision about how to improve a process, we must separate common cause variation from special cause variation.  Special cause variation is reduced by eliminating the special causes.  Common cause variation is inherent in the process and can only be reduced by changing the process.
Two mistakes are therefore possible.  One is to tell people to find and eliminate special causes when none exist.  The other is to start making changes to the process when we should in fact be looking for special causes.
Once we understand whether or not a process is capable and whether or not it is in control, we must next look at the two together.  There are four possible combinations of capability and control, and the actions that are appropriate in each case are different.
The most desirable state is a process that is in control and capable, where the output is stable and the output is meeting the requirements of the customer.    Even if this is the case, further improvement is possible by centering the mean of the output of the process on the target value that provides the most satisfaction, and by reducing variation around the mean of the output so that the capability of the process is as high as possible.
Another possible combination is a process that is in control but not capable.  The first action should be to center the output of the process on the target value and then reevaluate to see if the output became capable.  If the process is still not capable, then the process needs to be changed in order to reduce common cause variation and improve capability.
The third possible combination is a process that is out of control but is capable.  The appropriate action here is to find and eliminate the special causes of variation in order to bring the process under control.   Further improvement may still be possible and desirable.
The fourth and final combination is the worst of the four, a process that is out of control and not capable.   In this case the process should be centered on the target and the special causes of variation eliminated.  At this point it should be reevaluated to see whether or not it is capable.  Further improvement may still be possible and desirable.