Welcome


Welcome to my blog for all things related to business quality (processes, systems and ways of working), products and product quality, manufacturing and operations management.

This blog is a mixture of real-world experience, ideas, comments and observations that I hope you'll find interesting.

Pages

September 2009
M T W T F S S
« Aug   Oct »
 123456
78910111213
14151617181920
21222324252627
282930  

How does Statistical Process Control work?

Compared with many quality management techniques, Statistical Process Control is self-evident. It’s all about controlling processes. Using statistics. “Simples!” as the meerkat in the advert says…

Well, no, statistics are rarely that simple. Underpinning SPC’s apparent straightforwardness is mathematical detail that can be quite difficult to get your head around.

SPC is one of the key tools at the heart of Six Sigma but also has huge value and validity as a quality improvement tool in its own right. In an blog of this length I can’t go into a lot of detail, but here’s a summary:

SPC monitors processes as they are being applied in order to control quality in something close to real-time rather than testing or inspecting at a later stage. This allows operators to immediately bring processes under better control and reduce variations, which in turn leads to reduced waste or improved quality.

At the heart of SPC techniques are Process Charts. These are histograms of data, produced from manufacturing or testing or inspection processes on a regular basis, that show the variation in specific parameters from sample to sample or batch to batch. SPC analyses these variations and enables you to get the process optimised and fully under control.

SPC200

There are many types of control chart that are used for different types of measurement with varying degrees of mathematical sophistication. The most common are the Mean (‘X-Bar’), showing the variation of measured values – see the diagram above for a simple example – and the Range (‘R’) chart, showing the range of values within each set of measurements.

Most control charts consist of:

  • Performance data, or values derived from analysis of groups of data points, plotted over time or over a series of samples
  • A centre line – often the mathematical average of the performance data or its analysis (‘Avg’)
  • Control Limits (upper – ‘UCL’ and lower – ‘LCL’) that define the range of common cause variations, often set at three sigma (three standard deviations) from the centre line although six sigma is used for highly demanding applications.

Visual inspection or mathematical analysis and manipulation of the control charts help the user to decide which type of process variations are being experienced:

  • Common causes; these are natural variations that will always occur (in the green zone of the diagram)
  • Special (or ‘assignable’) causes; these are external events or stimuli (change of materials, change of personnel, tool performance, environmental effects, etc) – their presence indicates that the process is out of control.

Although it is beneficial to reduce the effect of common causes, special causes are the main focus of SPC as they indicate extraordinary reasons for poor results that can often be easily rectified.

Special causes can be spotted on the control chart by using tests such as:

  • 1 data point falling outside the control limits (in the red zone of the diagram)
  • 5 points in a row increasing or decreasing (indicates a trend)
  • 7 points in a row on one side of the centreline (indicates a shift in the mean)
  • …and a number of other established rules and more sophisticated statistical analyses that are optimised for different applications and types of data.

If you detect special causes, which means the process is not under stable statistical control, then you need to do something about it. You can use root cause analysis, experimentation with different materials, environmental conditions, equipment or process settings, and problem-solving techniques such as 5-Whys or 8D or FMEA or TRIZ or Pareto Analysis or Cause and Effect (Ishikawa) Diagrams or… etc. Sorry, but a full explanation is too long for this blog piece although it does gives me an idea for some future blogs!

The drive to achieve Continuous Improvement would suggest that you should also improve the basic process, i.e. reduce the variations due to common causes. You should certainly continue to monitor the process to ensure the improvements are sustained (and improved further if appropriate). As you would imagine, control charts and their statistical analysis are used in all these process improvement activities.

This is by no means the whole of Statistical Process Control. To be frank, it’s not even the tip of the iceberg but I hope it does give you a small insight into whether it might be relevant to your business.

Share

2 comments to How does Statistical Process Control work?

Leave a Reply

 

 

 

You can use these HTML tags

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>