Quality improvement

Good quality means having control of the important details!

statistical process control

In order to do quality improvement it is important to consider all factors that can be significant. Start by identifying everything that can affect your product. Even if you do not think it matters, include it! It is reflecting about everything that may affect product quality, which is important.

All processes are subject to variation, which may be classified as random or chance variation (common cause) or assignable cause variation (special cause). Common cause variation is found in all processes and exhibits random yet stable variation that is predictable within calculated process control limits. Common cause variation is synonymous with natural variation, expected variation, or random variation. Special cause variation, on the other hand, is unpredictable, sporadic, or unstable variation. It is the result of a specific assignable cause, for example Machine (Equipment), Man (Operator), Material, Method, Milieu (environment).

Humidity is an example of specific assignable cause (mileu) that can be controlled. The challenge of quality assurance is to identify the critical factors and then control the variation of these factors. Then you achienve quality control.

Six Sigma involves statistical tool to understand variation and identify critical factors. I have experienced that companies have used resources to control a factor that was not decisive for the quality. I have also experienced that businesses had to cancel the launch of a new product because they failed to identify the critical factor that gave variable product quality.

What factors do you think are critical for your production? And how do you control the variation? You can use Component of Variation - CoV to investigate how much different components contribute to the overall variation in product quality. When you have identified critical factors you can monitor them by using Statistical process control - SPC. It's a tool to separate between special and normal cause variation. Statistical process control can be used to monitore your process and react to special cause variation. You can use it to decide normal variation of different machines / processes and compare it to customer requirements. This way you can identify your process robustness. I was part of a team that implemented online statistical process control - SPC (norwegian article about the work) at Axis-Shield. I have also prepared a video with different applications of statistical process control - SPC. You can register and get if for free.