Statistical process control

To use of statistics in quality control - is the concept of variance.If we would like to summarize the entire field of statistical quality control (also called statistical process control or SPC) in one word, that word would be variance. Shewart, Deming, and others wanted to bring the statistical concept of variance down to the shop floor. If supervisors and production line workers could understand the existence of variance in the production process, then this awareness by itself could be used to help minimize the variance. Furthermore, the variation in the production process comes due to two types of cause : variation due to natural causes and variation due to assignable/special cause. Examples of assignable causes are fatigue of workers and breakdown of components. Variation due to assignable causes is especially undesirable because it is due to something being wrong with the production process, and may result in low quality of the production items. 

"A process is considered in statistical control when it has no assignable causes, only natural variation."

Actually , any kind of variance is undesirable in a production process. Even the natural variance of a process due to purely random causes rather than to assignable causes can be matter of loss to the production. The control chart, however,will detect only  assignable causes. but the use of control charts can go a long way toward improving quality, shown in the below diagram