ControlCharts_04_ProductPerformanceProcessControlProcessCapability

Before delving into in the concepts of process control and capability, we must first understand the difference between them.

When Shewhart (1931) introduced his concept of statistical control, three concepts were highlighted:

• The fundamental focus is on the process
• The overarching objective is economic operation of the process
• Process behavior falls within predictable limits of variation.

With those three things in mind, we can now start to understand the notions of the product's conformance to specifications; and statistical control of the process. The comprehension of those ideas is vital to understanding the difference between process control and capability.

Conformance to specifications means that the characteristics of the product falls within specification limits, which implies that the customer expectations are being met. Process performance falls also within limits of variation, called control limits. To have a product within specifications and a process under control implies that the customer expectations are being met in an economic fashion.

Figure 1 Product Conformance and Process Control

We can take a look at figure 1 above to achieve a better understanding of those concepts. In Figure 1(a) we have a graph comparing characteristic measurements to the specifications, telling that those characteristics are within them. However, in Figure 1(b) we have Shewhart $\overline{X}$ and $R$ control charts representing the same characteristics, this time showing that we have an out of control process, in other words, this process is not running in an economic fashion.

Below we have key points to have in mind that help to understand the difference between process control and process capability:

• Statistical control examines the variability pattern in the data to determine the extent to which the process is driven only by the forces of common-cause variation.
• Process capability is an assessment of the ability of the process to meet the expectations of the customers in terms of the quality characteristic(s) of interest.  i.e., conformance to the specifications.
• Statistical control focuses on the study of the economic viability of the process.
• Process capability focuses on the conformance of the process to yield a product that meets the pre-specified needs of the customer, and, hence is an analytical study.
• A process can exhibit good statistical control, i.e., be stable, predictable, but fail to meet the customer expectations.
• Failure of a controlled process to meet the specifications results from a process whose common-cause variation pattern is too large, or one that is not properly centered.
• To make the process capable, sources of common cause variation must be identified and removed to reduce the spread of the process relative to the specifications.
• A process not in control cannot be capable because statistical control is an essential prerequisite to the assessment of process capability.
• Without control, statistical models used to assess capability are not valid, i.e., the assumptions upon which they are built are not satisfied.
• Hence, a process must be in statistical control before its process capability can be rationally assessed.