Traditional manufacturing relies upon production to make the product and the quality department to inspect it. After-the-event inspection is expensive and wasteful because:

• The product has already been made
• Costly re-work is not always possible

It is much more cost effective to avoid waste by monitoring and analysing the process during manufacture - this is the basis of Statistical Process Control.

Controlling Your Process

For a product to be made without scrap, it must be manufactured within specified limits. But factors can prevent this from happening:

Natural Variation: inherent in the machining process and cannot be changed without using a different process or machine

Assignable Variation: outside influences that are controllable - temperature, sharpness of the blade, speed of manufacturing, skill of machinist etc.

An Example of Variation

A machine cutting straws to length will give an error from straw-to-straw. This is because of the inherent tolerances of the machine - Natural Variation. BUT this error is less significant than an individual cutting straws to length using a ruler - Assignable Variation

This raises the question - Is my manufacturing process able to manufacture within specification?

Testing your Process Capability

Continuing our example, cut a number of straws to the required length (usually 50). Accurately measure the straw lengths. Plot the lengths on a graph to identify the variation.

Histogram and Capability charts can be used for this purpose. Once you have determined that you are process capable, you can now monitor your process over time.

Monitoring Your Process Over Time

In an ideal world, you would measure each and every product that is being made...… In the real world, there is not enough time or resource to do this so we measure a sample group of product on a regular basis. These groups are known as subgroups.

The subgroups of data are plotted onto a graph - in chronological order

The average value of each subgroup is then used to generate the Process Control Chart - building up your actual manufacturing process over time - this is known as the Xbar Chart

Setting Control Limits

In order to prevent scrap, a set of “early warning limits” known as Control Limits are established. These warning limits are set inside the upper and lower specification limits.

Datastat+ provides the means to display your production process and highlights when your process violates these control limits. Datastat+ can also predict future violations.