Statistical process control - Using control cards

A process transforms inputs, such as raw materials, into outputs. Outputs may be physical products or services. The resulting product, whether physical or not, has characteristics, some of which are important to the customer or user. This transformation may differ depending on the setting of various adjustable process parameters, or on the level of other factors that may be experienced (e.g. outside temperature).

Process control means being able to predict product characteristics given the level of all influential process parameters. It also means being able to adjust adjustable process parameters according to the level of influential factors experienced, so that product characteristics meet expectations.

Among the factors acting on the product's characteristics, some have a very strong influence, others a moderate one, and still others a very weak one. Just as when we define a product, we admit a variability of the product characteristics (tolerance interval), we admit that certain factors with little influence on the product characteristics are not identified and generate, due to their unknown fluctuation, a variability of the product characteristics, as long as the product characteristics remain largely acceptable.

To guarantee that a product conforms to specifications, the process owner has two options:

• inspect all products if possible (in the case of destructive testing, this is not possible);

• to have identified all the influencing factors, to have made the necessary adjustments to ensure that product characteristics meet expectations, and to have verified by checking samples that no other previously unknown factors have come to light, i.e. that the result is consistent with the cause-effect relationships already identified.

As a result of these definitions, two cases of application of statistical process control (SPC) can be distinguished:

• influential factors are not fully identified: we need to observe process fluctuations, determine when unidentified influential parameters appear, identify these factors and find ways of counteracting their influence. Technical investigations are necessary;

• the influencing factors have already been fully identified: all that needs to be done is to check that no new ones emerge. A monitoring technique with simple reactivity rules based on observations, without necessarily using control charts, may prove sufficient.