Understanding and Applying Statistical Process Control in Manufacturing, Part 1
In a recent post, I briefly mentioned statistical process control (SPC) and why it is such a valuable tool for industry today. In this post, I will begin a new series on SPC, laying out a more detailed look at this very important tool for process improvement and why companies need it more than ever in today’s marketplace. It is not my intent to make you an expert on SPC, but rather to explain the framework, applications, and other considerations for its use.
A historical perspective on the origin of statistical process control
SPC was initiated by Walter A. Shewhart of Bell Laboratories and dates back to the early 1920s. Back then, Shewhart developed a tool he referred to as a control chart, as well as the concept of a state of statistical control. Dr. W. Edwards Deming invited Shewhart to speak at the Graduate School of the U.S. Department of Agriculture and served as the editor of Shewhart's book, Statistical Methods from the Viewpoint of Quality Control.
Deming was the primary developer of quality control short courses that were used to train American industry professionals in these new techniques during WWII. Some of the graduates of these courses were credited with creating the American Society for Quality (ASQ) of which Deming was then elected as the very first president. Deming subsequently traveled to Japan where he introduced SPC methods to Japanese industry and the rest, shall we say, is history!
What is statistical process control?
SPC begins with a measurable product characteristic such as the diameter of a hole drilled into a metal part. Sampling data is collected and sample averages and ranges are calculated to determine whether or not the process is in control. Control, in its most basic form, is the process of establishing and meeting the characteristic’s predetermined standards and then operating within a predictable range of variation.
SPC is a method of control which employs statistical approaches to monitor and control a process. SPC helps to make sure that the process will operate more efficiently and effectively, with the result being more product conforming to specification limits. All this is accomplished while significantly reducing the amount of rework and scrap. Some of the key tools associated with SPC include items like run charts, control charts, and design of experiments which are all focused on continuous improvement.
One advantage of using SPC over other methods of quality control, such as inspection, is that SPC emphasizes early detection and prevention of problems, rather than the identification and correction of problems after they have occurred. This is a key function. SPC also results in a reduction in the time required to produce the manufactured product, simply because it results in much less finished product requiring rework or scrap, which are both the result of excessive variation.
An overview of the fundamentals of variability
Variability exists in all production systems and can have an enormous impact on performance. For this reason, it is critical that companies are able to measure, understand, and manage the variability of their manufacturing processes. So just what is variability? In its most basic form, variability refers to the degree to which a group of data is spread out. In other words, variability measures how much your control characteristic measurements differ from one another.
In any process, there are primarily two types of variation present that can create disorder, if they aren’t identified and acted upon. These are referred to as common cause and special cause variation. Common cause variation is the normal, quantifiable, and predictable type of variation that occurs within any process or system. Special cause variation, on the other hand, is the type of variation which must be identified and completely removed from any process or system if the process is to be seen as predictable.
A process that is operating with only common cause variation present is said to be “in a state of statistical control,” while a process with special cause variation is referred to as being “out of control.” The fundamental challenge of SPC is to separate special cause from common cause variation. Because we only observe the control characteristic, and not the cause of the variation, SPC helps to determine the presence of out-of-control conditions.
It should be apparent that no two products or characteristics are ever exactly the same simply because any process contains both of these sources of variability. In any type of manufacturing, the quality of a finished product is many times safeguarded by inspecting the entire lot of finished goods. Each unit is either accepted or rejected based upon whether it meets its design specifications or not. SPC, on the other hand, uses statistically based tools to determine how well a production process is performing, while simultaneously looking for variation before defective product is produced in large production runs.
When both types of variation are either removed (i.e., special cause variation) or reduced (i.e., common cause variation), the process is said to become stable and predictable. That is, when a process is stable and predictable, its variation will remain within a known set of limits referred to as control limits, at least until another special cause source of variation rears its ugly head! The use of control charts will limit the number of times special cause interrupts a process to produce defective product.
Coming in the next post
In my next post, I will discuss SPC’s tool used for controlling processes, the control charts. Meanwhile, if you’d like more information on process improvement, please watch this video.
Until next time,
- Walter A. Shewhart, Statistical Methods from the Viewpoint of Quality Control, Graduate School, Department of Agriculture, Washington, D.C., 1939; Dover, 1986
Don't miss out!
Stay on top of the latest business acumen by subscribing to the Manufacturing Breakthrough blog.