Master SPC to Improve Quality and Stability

«Statistical Process Control – Ensuring Quality and Preventing Deviations» Lean Tech

Statistical Process Control (SPC) is used to monitor measurement results and ensure that quality stays within customer requirements.

SPC is not just a tool for reacting to errors – it helps you work proactively by detecting changes in the process early, allowing you to act before deviations occur.

How does SPC work?

SPC uses control charts – a Run Chart (measurements over time) with control limits. These limits distinguish between common variation and special variation, which is essential for responding appropriately to measurement results.

Incorrect adjustments can create bigger problems:

⚠ Overreacting to common variation? This can lead to unnecessary adjustments that actually increase variation and create more issues.

Responding Correctly to Variation

Common and special variation require different approaches:

Special variation: Identify and eliminate the specific cause that has disrupted the process.
Common variation: This is inherent in the process. To reduce it, you need to improve the process's robustness – not just make random machine adjustments.

Example from production

A machine has too much common variation and produces defective products. All machines have a certain degree of natural variation, but repeated adjustments can actually make the problem worse.

💡 Solution:

Is the machine inaccurate? Adjustments may help.
Is precision the issue? Adjustments are rarely effective. You may need to invest in more precise equipment or calculate the cost of defects into your product pricing.

Understanding the difference between common and special variation is the key to taking the right actions! Learn more in the video below:

Common variation and process control

Common variation is the sum of all random factors in production. It can be reduced with a good understanding of the process:

Identify critical variables – what has the greatest impact on the process?
Control the variation of these variables to ensure stability.

Here is an article from the magazine "Automation" about online process control, from a production line at Abbott where Sissel from Lean Tech implemented online SPC.

Examples of SPC applications

Determining robustness (capability): What is the normal variation of the process, and how much margin do we have for customer requirements?
Predictive maintenance: Wear and tear can be detected in a control chart as special variation, such as increased variation or trends.
Effectiveness of improvements: Special variation can be an early signal that the measures you've implemented are working!
Early warnings: Work proactively by detecting tendencies for deviations early and taking action before quality issues arise.
Ensuring stable processes: A process with only common variation is predictable. If it also meets customer requirements, you have a robust and stable process.