Encyclopedia  |   World Factbook  |   World Flags  |   Reference Tables  |   List of Lists     
   Academic Disciplines  |   Historical Timeline  |   Themed Timelines  |   Biographies  |   How-Tos     
Sponsor by The Tattoo Collection
Statistical process control
Main Page | See live article | Alphabetical index

Statistical process control

Statistical Process Control, or SPC is a method for achieving quality control in manufacturing processes. It was pioneered by Walter A. Shewhart and taken up by W. Edwards Deming with significant effect by the Americans during the World War II to improve aircraft production. Deming was also instrumental in introducing SPC techniques into Japanese industry after that war.

The technique hinges on the observation that any manufacturing process is subject to seemingly random variations, which are said to have common causes, and non-random variations, which are said to have special causes. A common cause might be air movement in the manufacturing environment, which causes variations that are outside the control of manufacturing operatives. A special cause might be the fact that the operative has a hang-over. Management can usually determine special causes for manufacturing defects by consulting the workforce, but dealing with common causes is a management responsibility.

SPC relies on measuring variation in manufacturing output and setting control limits based on observations of variations arising solely from common causes. A process that is "in control" is expected to generate output that is within the control limits. If the process produces an "out of control" point, one would not necessarily assume the process had moved to an "out of control" state but would try to locate the special cause(s) for this condition. Only if special causes could not be found would an assumption be made that there might be new common causes to be identified. One aspect of process quality improvement is achieved as these common causes are found and corrected - special causes have no bearing on the overall quality improvement process.

The main quality improvement process consists of the intentional varying the production process to achieve a smaller range of control limits (See, for example, design of experiments). It has been shown that manufacturing processes can achieve control limits which are a tenth of the specified manufacturing tolerance. Such a process can achieve zero defects - because even articles that are outside the control limits due special causes are still within the specified tolerances. The reduction in waste and inspection resources can make processes subject to SPC far more efficient, and the predictablility implied by processes that are in control allows further savings to be made by adopting just in time inventory control.

Processes may have outputs that can be measured as variables or as attributes. Variables are characteristics of a product that can be measured on a continuous scale. An example of a variable would be the length or width of a product or part. An attribute is an aspect or characteristic of a product that cannot be put on a linear scale. For example, a light bulb will either light or fail to light. "Good/bad" is an attribute, as is the number of defects.

There are several types of commonly used process control charts. Among them are X-Bar, R Chart; P Chart; NP Chart; C Chart; and U Chart. Each chart has a specific area of application.

See also:

This article is a stub. You can help Wikipedia by [ expanding it].