Statistical Process Control (SPC) is a type of quality control practice which uses statistical methods to monitor and control a process. It allows manufacturing facilities to observe and manage behaviors, uncover bottlenecks, and discover solutions for production problems.
Much of the SPC we see today in modern facilities can be traced back to World War II, when products such as munitions called for stringent quality measures. SPC methods have seen a major resurgence in recent years, especially with initiatives such as Six Sigma, a method used to improve the capability of business processes. One hallmark characteristic of SPC is the control chart, which was first developed by Walter Shewart in the 1920s.
SPC Quality Charts
Also known as a Shewhart chart, an SPC control chart monitors how processes change over time. Data points are plotted based on time, and there is a central line to indicate the average, as well as an upper and lower line to depict control limits. Analyzing these data points allows manufacturers to determine whether process variation is consistent, in control, or unpredictable. When a process is deemed unpredictable, it means that it is out of control, and impacted by special causes of variation. For a quick video on how the math works, watch this introduction video about SPC: What is it, and why is it needed?
SPC Control Charts are Useful for:
- Identifying problems as they occur for prompt correction
- Predicting the anticipated range of outcomes for a process
- Finding out if a particular process is considered stable
- Drilling down into process variation caused by special causes, including non-routine events versus common causes which occur naturally within the process
Ultimately, SPC quality control charts can help inform your quality improvement initiatives by helping you determine whether you should focus on preventing an isolated problem or rework a process entirely. The first thing to think about is that all processes have a variation in them, and that forms the very foundation for SPC charting.
The Two Types of Process Variations
As mentioned above, process variations can fall into one of two categories.
- Consistent: Common cause variations are ever-present and intrinsic to a process.
- Unpredictable: Special cause variations are a result of external conditions, and are thus out of the realm of statistical control.
Common cause variation, sometimes referred to as “noise variation,” impacts all outcomes of a process and everyone involved in it. To manage this type of variation, manufacturers must therefore focus on the process itself. Common cause variations can either be addressed by making a change or by management.
On the other hand, special cause variation (or “signal cause variation”) is not an inherent aspect of the process. To address these variations, manufacturers must identify and address the special cause. Special cause variations are often removed by quality assurance analysts.
Being able to identify the quality challenges in your facility is important. Through SPC, you can determine which aspects of your production need fine-tuning and where you should direct your efforts. While SPC charts are one useful tool for helping manufacturers better understand trends in their facilities, there are also several others to try.
SPC Quality Control Tools
In the 1970s, Dr. Kaoru Ishikawa, author of Guide to Quality Control, described seven quality control tools which manufacturers could use to improve processes. In addition to the control chart (described above), here are the remaining tools:
- Cause-and-effect diagram: Also known as the fishbone diagram, this chart maps out the many potential causes of an issue. For example, you could explore possible reasons why one shift is producing out-of-spec products.
- Check sheet: This simple form tracks an event and the causes behind it to help identify patterns. For instance, you might track stops on a particular line over one week, identify a series of reasons, and use tick marks to tally them up.
- Pareto chart: This bar graph is designed to identify the significance of different scenarios. The bar lengths indicate cost or frequency in order from longest to shortest.
- Histogram: A histogram is similar to a bar chart but different in that it shows frequency distributions. You might use a histogram to plot out quality defects, where the X-axis would represent defects per hour and the Y-axis would represent the frequency (how many times there have been ten defects in an hour, for instance).
- Scatter diagram: Also known as an X-Y graph, scatter diagrams are ideal for a pair of numerical data. One variable is set up on each axis to determine if there is a relationship between the two.
- Stratification: Typically used alongside other quality control tools, stratification sorts data into distinct groups. For example, you might look at data sources such as equipment, materials, shifts, or time of day, and plot them using a scatter diagram. You’d then use different colors to stratify the data and look at each subset of data in an attempt to uncover patterns.
These quality control tools are used when looking at SPC as well as statistical quality control (SQC).
What’s the Difference Between SPC & SQC?
Although SPC and SQC are sometimes used interchangeably, there is an important difference between these concepts: SPC controls process inputs, or independent variables, whereas SQC monitors process outputs, or dependent variables.
Why SPC Quality Matters Today
Clearly, a manufacturing tool which has stuck around for nearly a century must have some merit. Despite the fact that SPC originated so long ago, its principles are more effective today than they ever were. Thanks to sophisticated data collection tools, pinpointing problem areas and monitoring process behavior have become automated, freeing up the time and effort that would otherwise go into manually tracking quality metrics.
By helping you uncover the root causes of quality issues, SPC delivers a number of noteworthy benefits for manufacturing. It can allow your plant to:
- Improve conformance with specifications. Consistently meet customer requirements by efficiently determining and addressing the cause of any non-conformances.
- Minimize waste. With less non-conformances, you’ll have reduced scrap and re-work.
- Satisfy regulatory and other requirements. In food manufacturing, quality and safety often go hand-in-hand. Chances are SPC will help you uncover trends which not only affect quality, but could also affect safety. By catching issues early, instead of during audits, you can stay on top of regulatory requirements such as FSMA.
- Decrease costs. Better conformance with specifications, reduced waste, and minimized risk of safety issues can all help food manufacturers keep their costs down.
- Become more efficient. Since SPC can also uncover bottlenecks and inefficiencies, you can use it to further boost performance and ensure processes are running smoothly.
While there are many factors to consider when implementing SPC, it’s clearly a worthwhile practice for manufacturers. Read our post to see why many plants are turning to OEE and SPC to save time and cut costs. You can find out how SafetyChain’s plant management platform can simplify quality assurance and compliance with an automated, paperless system here.