At CPNET, we bring cutting-edge AI technology to the manufacturing industry, making the production process more efficient. We reduce waste by leveraging data and algorithms that are built on a modern computing infrastructure. In the past few years, we have had the chance to talk with many manufacturing practitioners and researchers. These conversations have taught us a lot. We hope to impart some knowledge just like these experts have with us. In today’s blog post we’ll be talking about Statistical Process Control (SPC), which was first introduced at Bell Telephone Laboratories in 1920 by Dr. Walter A. Shewhart. Additional references can be found at the end of this blog post, if you are interested in learning more about the topic.
What is SPC?
Manufacturing professionals who are already familiar with SPC can skip this section. For those who are new to SPC, here is a short crash course.
According to ASQ (American Society for Quality) SPC is “... the use of statistical techniques to control a process or production method. SPC tools and procedures can help you monitor process behavior, discover issues in internal systems, and find solutions for production issues (asq.org).” This came about as a way to lessen any variation that occurs during production, or before, if it had occurred naturally in the materials.
Since this is a way of monitoring the process, there are tools that can be used to aid in the organization of the data. A specific example of this is a control chart. “Control charts attempt to distinguish between two types of process variation: 1.common cause variation, which is intrinsic to the process and will always be present, 2. special cause variation, which stems from external sources and indicates that the process is out of statistical control (asq.org).” The data on these charts is usually displayed in a bell or curve shape.
There are several experts who have been instrumental in the study of this method. Dr. Walter Shewhart is often described as the “father” of SPC. He is known for creating Control Charts, which were previously mentioned, and bringing them to the forefront. According to Quality Magazine, Shewhart first created Statistical Process Control at Bell Laboratories in 1920. Although this method has grown and changed over time, he had a solid knowledge of manufacturing during that time period and what needed to be done to make it more efficient (QualityMag, A Brief History of Statistical Process Control).
Another person who is known in the study of SPC is W. Edwards Deming. He was a well-known figure in academia, and had published many papers on manufacturing-related topics, including variation. He made strides in popularizing the method during the 1950’s and 1960’s. He made changes to how management in this industry works – which has had a lasting impact – while he consulted for businesses. He created some principles that can help businesses improve, based on the leadership of their management team. These principles are called the 14 Points. They talk about taking on leadership roles, not relying on inspection, the importance of training, among many other things in hopes of overall improvement (The Deming Institute, Dr, Deming's 14 Points for Management). Not only is his work in the US notable, he is known to have done work in Japan also (The W. Edwards Deming Institute, Deming The Man). Some of his thoughts on SPC and variation include, “... all things vary, and that is why the statistical method is needed. People are not alike; they have different tastes, and do not all perform best under the same conditions (from A View on the Statistical Method by W. Edwards Deming).”
How a team manages their quality control may vary according to each factory. The following example shows quality control measures at a paper mill. Their document suggests training a variety of employees in SPC, ranging from people on the production line to higher up positions. They also suggest learning from experience and knowing what benefits the customer while taking part in the implementation project. Knowing what the customer wants or needs can help to get an idea of what quality standard needs to be met and adhered to using Statistical Process Control.
Another way of ensuring quality is the detection of any issues early on. SPC relies on continually monitoring and making changes in order to make room for improvement. Some companies may run into problems at the start of their journey in SPC. One issue that may occur is related to having a system that isn’t automated yet. This can make human error more prevalent, while automation can make production more meticulous and lessen the possibility of error. It could be a challenge to get used to a totally new routine, while learning about technology. Those who aren’t trained in machine learning may not know a lot about it and the things related to it such as control charts.(QualityMag,Three Common SPC Initiative Challenges).
Learn more about SPC
There is a lot of information out there on this topic; we could write many blogs related to it. Because of this, you should make sure that the information you are reading is from a reputable source. If you want to learn more about the topic, start with the experts and pioneers of SPC like Shewhart, Deming and Donald J.Wheeler. The latter puts out blog posts about SPC frequently while Shewhart is the one who started it all.
Sources
A Brief History of Statistical Process Control:
Optimizing Mill Effectiveness Through Statistical Process Control:
https://www.tappi.org/content/Events/19PEERS/19PEE26.pdf
What Is Statistical Process Control?:
Walter A. Shewhart:
https://asq.org/about-asq/honorary-members/shewhart
Comparative Study of Lean Manufacturing Tools in Manufacturing:
http://www.iaeng.org/publication/WCE2013/WCE2013_pp604-608.pdf
A Comparison of Manufacturing Technology Adoption:
http://simula.cesga.es/document/estudios/A_comparison_of_manufacturing_technology_adoption.pdf
Continuous Quality Improvement by Statistical Process Control: