What Is Statistical Process Control?
Simply put, SPC uses statistical methods and sampling programs to help plant and operations managers understand and control variability in their manufacturing processes. Process variations are displayed in real-time charts and trigger alerts as processes begin to trend out of expected, standard limits. The main objective of statistical process control (SPC) is to better manage operations and reduce waste and the need for rework. Plant and Ops managers like SPC because it gives them the concrete information they need to stop a line and fix problems before they spin out of control.
SPC relies on several methods for improving manufacturing yield and profitability. For one, SPC uses control limits. These limits keep process runs well within specifications. Should a process surpass these limits, alerts can be triggered even before specification limits are approached. SPC also relies on run rules and the interpretation of histograms and process capability measures (CPKs & PPKs).
While SPC does enforce control, it also adds another layer of value: by understanding variability, you can better understand your processes and targets to see if any improvements can be made for greater efficiency.
Here’s a closer look at some common SPC tools.
SPC Control Charts
Control charts are real-time monitors that illustrate how a process changes. In these charts, periodic samples are plotted over time. Types of control charts include:
- Xbar (mean chart): This chart calculates the average values of a sample set.
- Xbar-R (range chart): This looks at a sample size of two to nine data points.
- Xbar-S (sigma chart): This chart is used for more than ten data points.
When these charts and control limits are utilized, it allows manufacturers to see whether they’re operating within limits. It also illustrates trends of variability. In some instances, increasing variability may be revealed through an Xbar-R or Xbar-S chart, even though the mean chart is still on track. This finding could prompt you to investigate further to look for the root cause of variability before the issue escalates.
SPC Control Limits & Run Rules
Control limits can be used in tandem with control charts. They leverage the natural variability of statistics to show a process’s expected upper and lower boundaries, illustrated by three sigmas on either side of the target (known as the six-sigma range).
Run rules can then be used to determine when a process has drifted and is no longer stable, based on the upper control limit, mean, and lower control limit. With SPC monitoring tools, you can set up alerts to be notified when you’ve violated a run rule. You can then assign different responses to different run rules. For instance, violation of one run rule may call for a temperature adjustment, while another one may require you to recalibrate machinery. Implementing run rules with control charts and limits allows you to stay on top of operations, keep processes under control, and minimize waste.
Histograms and Their Role in SPC
Histograms are an often overlooked yet valuable tool for SPC. These charts focus on control; for instance, you might use them to ensure you’re staying well within control of customer specifications. Sometimes, customers may even request histograms as a tool to make sure their suppliers are doing their due diligence. Histograms can also be useful for reviews and post-op analyses.
A histogram looks at process variation or the idea that a process has an inherent tendency to deviate. Process variation exists in all processes, both manmade and natural. You can account for process variation by using tools to evaluate how much a process varies, taking into consideration the bell curve and determining if the process is in control. Of course, the idea is to have a narrow curve with shorter tails, which would illustrate more values within specification limits. If outliers are identified, you can drill down further to seek solutions for controlling the variation.
Process Capability Measures: CPK & PPK
Process capability measures allow you to understand the extent to which your process is able to conform to specifications. In some processes, there will be some aspects of the process which will inevitably fall out of spec. CPK allows you to look at the variability of sample populations within standard deviations, while PPK measures normal process capability using the overall standard deviation.
If a CPK value is less than 1, it means that the process cannot meet specifications. Yet, it shouldn’t be too high, either, as this could also indicate issues related to the target or variability. The ideal value for PPK is 1.33, which means 99.99% of all materials are within specification.
SPC is useful in many types of manufacturing, but it’s an especially powerful approach for streamlining operations in consumer package goods (CPG). These manufacturers face unique challenges, including the need to meet rigorous customer specifications while controlling costs. Here’s a look at how SPC can be applied in the CPG industry.
Optimizing yield is a top-of-the-mind concern for many plant managers. Of course, it starts at the operator level, at the point of data capture. For instance, operators might capture internal product temperatures. A temperature that’s well over the limit could violate safety and regulatory requirements, but it might also result in high costs as yield and quality are affected. The goal is to create a safe product by achieving the right temperature but not overshooting it.
Temperature variability can have a tremendous impact on product quality in subsectors of the food and beverage industry, such as baked goods. For instance, if the temperature for bread becomes too high, the product can be baked unevenly and become misshapen, too low, and there’s a risk for raw product or increased denseness. With cookies, elevated temperatures can lead to overspreading, whereas temperatures that are too low will lead to pale, under-baked goods.
When an SPC tool such as SafetyChain is introduced, an operator will enter temperature data, and if there is a specification violation, it will trigger an alert. The operator can see on the appropriate control chart, which run rule has been violated, as it will appear in red, and the operator will be prompted to enter a comment. The records can then be accessed by management teams and will refresh in real-time, and users can even review and sign off right from their dashboard.
SPC Helps to Minimize Giveaway
Reducing giveaway is one of the greatest areas of opportunity for SPC, especially in the CPG industry. It’s easy for manufacturers to directly assign dollar values to package waste. SPC software can identify weight values without operators having to enter the data manually. With real-time control limits, operators can make sure they’re meeting label requirements in terms of weight without giving away product. Should they violate an upper control, they can enter a comment related to the filler, packer, or another factor.
The data analysis can then be rolled up to management, which is useful for identifying the distribution of waste. For instance, if the data spread indicates you have some room to improve, you might consider re-targeting to reduce giveaway.
Perfecting Package Weight in CPG Products
Consumer packaged goods such as shea butter lotion present notorious packaging challenges. If the formula is too light or fluffy, more volume will be required to hit the package’s target weight. On the other hand, a product that’s too dense could lead to overfill or rework. Since there are many attributes to look at to ensure proper package weights, this scenario would call for multi-attribute control. For instance, you could measure pH viscosity, among other parameters, to ensure you’re meeting both ends of the requirement.
As another example, a CPG company might have fruit fillings that would need to maintain the proper pH, temperature, and viscosity. Hot fill must maintain the proper temperature to sterilize the packaging, but this and other factors will also affect the product’s quality. Different kinds of charts can be used to track these critical metrics in real-time to ensure the right consistency and temperature, among other factors.
Building an SPC Program in Your CPG or F&B Plant
Here are the steps you can use to implement SPC in your plant:
1. Identify Attributes
- How can we impact performance?
- What are the data points that will allow us to do so?
Then, set an improvement goal.
2. Establish a Sampling Program
Next, determine where you’ll find the data, along with how to deploy and monitor the program. Who will be responsible for carrying out the program, and which tool setup and maintenance steps will be involved?
3. Monitoring & Control
Decide which control charts you’ll use and perform active monitoring. Review with your teams to make sure you’re on track.
When it comes to analyzing your data, visibility and share-ability is key. Return to your goals and determine whether you’ve hit them or missed them, and revisit any approaches as needed. Troubleshoot any quality issues you identify.
As you begin your journey into SPC, you might also consider asking around your plant to see if anyone has familiarity with the method. You might discover some helpful insights, and employees with previous experience might also be willing to take some degree of ownership over the initiative.
To watch the webinar replay of Putting SPC to Work in Your Manufacturing Plant click here.