Leading ERP software for growing manufacturers

Categories: Process Improvement
Understanding and Applying Statistical Process Control in Manufacturing, Part 2

Understanding and Applying Statistical Process Control in Manufacturing, Part 2

By Bob Sproull

Review of Understanding and Applying Statistical Process Control in Manufacturing, Part 1

Part 1 was the introduction to a new series on statistical process control (SPC). In this post, I provided a brief historical perspective of SPC, defined its parameters, and discussed the reasons for using this method of quality control.  I finished with a discussion of two different types of variation, common cause and special cause, and then briefly introduced SPC’s primary tool, the control chart.

In this post, I will explain how the control chart is used to provide statistical control of your processes, as well as the basics of how to create one. 

How to apply SPC to processes

The implementation of statistical process control (SPC) includes four main phases, as follows:

  1. Select and develop a measurable control characteristic that directly relates to product quality.
  2. Understand the process when comparing it to the specification limits of your control characteristic. That is, we want the natural spread of the process to be inside the upper and lower specification limits of the selected control characteristic.
  3. Identify and eliminate all special cause variation to provide stable processes. What we want is a process that contains only common cause or natural variation and that is free of special cause variation.
  4. Observe and monitor the production process through the use of control charts. These enable us to detect significant shifts in the sample process average (i.e., process mean or X-bar) or variation (i.e., +/- 3 x standard deviation).

Study your process with a control chart

Control charts are statistically based graphs which are used to study the stability of your process as a function of time (i.e., sample data point averages are plotted in time order). A control chart has a central line representing the process average for the selected control characteristic (i.e., X-bar) and upper and lower lines (i.e., control limits) based upon the inherent variation of the data collected. By comparing your real-time sample data to these three lines, you are then able to draw conclusions about whether the process variation is consistent (i.e., in control) or inconsistent (i.e., out of control).

If your sample average remains between the upper and lower control limits and does not contain any abnormal patterns, then your process is said to be in a state of statistical control with only common cause variation present. On the other hand, if your sample average falls outside your control limits or exhibits unstable patterns, then your process contains special cause variation and is said to be out of control. 

So how do you determine the X-bar and control limits?  You do so by running a process potential and capability study.

 

When you put these components together, the control chart looks like the image shown above.  The center line represents the average value of your control characteristic as a function of time, while the upper and lower control limits (represented by dashed lines) are the control limits which correspond to +/- 3 times the standard deviation (i.e., σ) of your sampling data, with standard deviation being a statistical measure of variation.

The combination of these two charts is referred to as the X-bar/R chart. The moving range chart (bottom of figure above) is used to track the within-sample variation. Also listed in the figure above are the basic rules of action which define whether a process is in control or out of control.

Basic steps for implementing SPC

The first step in implementing SPC is to select what is referred to as a control characteristic. The selection should be based upon two primary questions:

  1. Is the control characteristic critical to product quality and is it measurable?
  2. Is an acceptable measurement system available?

When answering the first question, it’s important to understand that there is a direct correlation between the measurement being made and an important quality characteristic of your product. For example, if product width is critical to the functionality of your product, then width would be an acceptable control characteristic. When answering the second question, it’s important that your measurement system has minimal variation associated with it.

Coming in the next post

In my next post, I will discuss how to perform process potential and process capability studies and then move on to the measurement system evaluation. If you would like to learn more about this topic, check out my recent series, A Practical Guide for Manufacturing Process Improvement.

Until next time,

Bob Sproull

Bob Sproull

About the author

Bob Sproull has helped businesses across the manufacturing spectrum improve their operations for more than 40 years.

facebook-icon facebook-icon linkedin-icon linkedin-icon twitter-icon twitter-icon blog-icon blog-icon youtube-icon youtube-icon instagram-icon instagram-icon Bookmark this page Google +