SPC gives the full picture!
Brief background
In 1923 Walter Shewart invented a method for distinguishing between random and systematic causes of variations in a process. This method later came to be called statistical process control (SPC).
What is SPC?
Is quality sorted out before delivery to your customer? Does your final inspection department or your customer find parts that are out of specifications, even though the operators by the machine have not intentionally produced any wrong parts?
The reason might be that you measure single parts and consequently interpret each single part as representing the total outcome of the process. Because the process at any given time consists of a natural randomly distributed spread, there is a risk of assuming that the entire process is well centered within the tolerance range. This may result in scrap:
Michael Nielsen, originator of
SPC – in plain language
Conversely, you could misinterpret the one single part to mean that the entire process is not centered and therefore in need of adjustment. Adjusting the process could then result in scrap:
Using SPC is a way of thinking. It means, among other things, that instead of checking individual parts against tolerance limits, you keep check on the outcome and can thus control the process on the basis of reliable data. That is why it is cost-effective to use statistical process control (SPC) in manufacturing.