Interpretation of analysis results

Data analysis itself has no value. To bring economic benefits to your process, every analysis you perform must lead to improvement actions.

You don't need to do anything special. All you need is a simple hypothesis and data to test it. Hints for hypotheses are often found within the process. For example,

• Example 1: "Machines 1 and 2 are used with the same settings, but for some reason, Machine 1 produces many defective products."
• Example 2: "The machining accuracy of the machine tool deteriorates in the afternoon."

Even if you know from experience that such hypotheses are correct, without data, it is only worth a" rumor."
You never know the cause and need evidence for improvements.

In a case like Example 1, you may be able to find the cause by collecting inspection data on a machine-by-machine basis.
In a case like Example 2, you might be able to find the cause by collecting data on temperature and tool wear, which change over time, along with product inspection data.

Unlike sales and marketing data, almost all process quality data comes from manufacturing and inspection equipment. Machines do not change their behavior based on their mood or whim. Therefore, if you find a particular trend within a process, it will likely be reproduced.

Suppose a highly reproducible trend is found in processes. That means you can correct or compensate with a high probability. This knowledge becomes a shareable property for all personnel engaged in production.

Find a trend (correlation)