Statistical Process Control Defined
Although SPC is normally thought of in industrial applications, it can be applied to virtually any process. Everything done in the workplace is a process. All processes are affected by multiple factors. For example, in the workplace a process can be affected by the environment and the machines employed, the materials used, the methods (work instructions) provided, the measurements taken, and the manpower (people) who operate the processthe Five Ms. If these are the only factors that can affect the process output, and if all of these are perfectmeaning the work environment facilitates quality work; there are no misadjustments in the machines; there are no flaws in the materials; and there are totally accurate and precisely followed work instructions, accurate and repeatable measurements, and people who work with extreme care, following the work instructions perfectly and concentrating fully on their workand if all of these factors come into congruence, then the process will be in statistical control. This means that there are no special causes adversely affecting the processs output. Special causes are (for the time being, anyway) eliminated. Does that mean that 100% of the output will be perfect? No, it does not. Natural variation is inherent in any process, and it will affect the output. Natural variation is expected to account for roughly 2,700 out-of-limits parts in every 1 million produced by a three-sigma process (±3? variation), 63 out-of-limits parts in every 1 million produced by a four-sigma process, and so on. Natural variation, if all else remains stable, will account for two out-of-limits parts per billion produced by a true six-sigma process.
SPC does not eliminate all variation in the processes, but it does something that is absolutely essential if the process is to be consistent and if the process is to be improved. SPC allows workers to separate the special causes of variation (e.g., environment and the Five Ms) from the natural variation found in all processes. After the special causes have been identified and eliminated, leaving only natural variation, the process is said to be in statistical control (or simply in control). When that state is achieved, the process is stable, and in a three-sigma process, 99.73% of the output can be counted on to be within the statistical control limits. More important, improvement can begin. From this, we can develop a definition of statistical process control:
Statistical process control (SPC) is a statistical method of separating variation resulting from special causes from variation resulting from natural causes in order to eliminate the special causes and to establish and maintain consistency in the process, enabling process improvement.
Review the section Statistical Process Control Defined. Explain how environment and the Five Ms can affect processes used in the following:
A hardware store
An accounting firm
A new-car dealership
Explain the relationship that exists between the histogram and the control chart.
Contrast the histograms characteristic of representing a snapshot of a process with a control chart.
Defend the statement that the operator of the process should be the owner and data plotter of the control chart, as opposed to a person from quality assurance or engineering, for example.
Comment on the significance of this statement: Control chart parameters must be statistically derived and cannot simply be specifications or some arbitrary values that are based on production expectations.