Maintaining Process Control with SPC Methods

Steve Wise
By Steve Wise | December 9, 2020
Vice President of Statistical Methods

Fact checked by Stephen O'Reilly

If you’ve followed along in this blog series, you may recall that we discussed what I called “breakthrough” performance in manufacturing in a previous article. You may recall that I used the example of a bicycle manufacturer to describe the difference between continuous and breakthrough improvements. To encapsulate that thought (which will, of course, lead to others), I offer this: as a producer of bicycles, I am always looking to reduce weight.
 
That is my continuous improvement goal. My breakthrough improvement was when someone invented aluminum welding, thus leading to reduced weight. Further breakthrough came when carbon fiber was introduced. The issue I’d like to discuss here is maintaining your improvement after the breakthrough has come. You’ve moved on to aluminum welding, or whatever technological breakthrough works for an apt example in your industry, and so…now what? Statistical process control (SPC) methods can help you maintain that manufacturing momentum and keep you on the path of continuous improvement. Let’s dive in…
Continuous Improvement with SPC

Applying SPC Methods

So, continuing on with our example, if you’ve had your breakthrough improvement, and you’ve learned a better way to set things up on your production line—there’s a different combination of input parameter set points, or maybe different types of inputs—then how do you ensure that things will stay at this new level? You want to make sure that your inputs remain stable.
 
And the way in which we do that is we create control charts on the inputs (just as we do with the outputs) with SPC. And I would add: especially the ones that we’ve discovered have the greatest effect on the output. Essentially, you’re taking one step back—upstream—and applying SPC to those inputs.
 

SPC for Upstream Monitoring and Prototypes

When you apply SPC to your inputs, you are watching the central tendency (mean) and variation (standard deviation) of these inputs. You then correlate that with the expected outputs, making sure that everything behaves the way you want it to. One way to achieve this is with prototype runs.
 
So, you have new settings based on what you discovered from your breakthrough improvement. And you’ve adjusted your expected results to these new settings. Sure, you achieve the results you want in the lab. Now, onto the factory floor with this. The best way to verify is with small prototype runs. Keeping in mind, of course, that when you go to full-scale production, things don’t always translate. But you have to start somewhere! So why not start with small, prototype runs? They're inexpensive, yet they can verify that you’re set up right and getting the results you want. You’re continuously improving!
Prototype Runs for Verification

Every Day Statistical Process Control

SPC is used for maintaining process control in the day-in, day-out rigors of making sure that the processes we use in our manufacturing are doing what they need to be doing. And they all require constant monitoring. Why do they always need to be monitored? Because they’re machines. You set up the machine. It should just do its thing, right? Not so fast!
 
You have a car, right? You go to the grocery store, run errands, drop the kids off here or there, maybe take a trip every once in a while. It doesn’t always behave like it should, does it? And any number of things might cause something to misbehave. Sometimes unwanted behavior can be predicted, but sometimes things go haywire without any notice. But at least you’re keeping an eye on things. Same with manufacturing. You’ve got to keep an eye on your machines. As my colleague Doug Fair, InfinityQS COO, discusses in his blog, Don’t Drive Blindfolded—The Importance of Real-Time Data in Modern Manufacturing, manufacturing machines have “personalities” all their own.
 
No two machines are alike. Even “identical” machines purchased from the same manufacturer. “We found that under the same test conditions, and for the same material types, each machine produced a little different quality from the others.” So, we constantly monitor them. Because, well, you just never know. Early warning detection is everything…in automobiles, in healthcare (don’t forget your annual check-up), and in manufacturing. And SPC is all about early warning detection.
Early Warning Detection with SPC

SPC Terms: Mean and Sigma

In SPC, early warning detection is keeping an eye on those means and sigmas. As I mention in the second blog in this series, The Top Advantages of SPC, “The two statistics that drive SPC are mean and sigma. That’s where we derive all the knowledge from SPC.”
 
When you’re looking to maximize the efficiency of your machines, matching the right machine to the right job is paramount. You can save wear and tear, money, and headaches by always “job matching” correctly. “If machine A produces 10% rework and no scrap, versus machine B’s 1% scrap and zero rework, I might want to rethink how I’m using the machines. This is valuable information. Is it more profitable for me to scrap 1% and not have to spend any money on rework? Or for me to rework 10%?”
 

SPC Tools to Explore the Unknown

If you've got these known, predictable means and sigmas for a particular machine or process, then we utilize SPC on those just to make sure that machine or line remains at the assumed level of variability. Whether we like it or not. It doesn't have to be good; it just has to be predictable, and then we can work with it. You utilize SPC on those things to make sure the assumptions we're making can still be relied upon. Make sense?
Monitoring Everything with SPC 
And then we approach the unknown… The worst sort of problem or issue in manufacturing is the one that you had no way of predicting. You just didn’t see it coming. And then it jumped up and bit you on the _____. Anyway, a tool breaks, the feed on a machine malfunctions, or the resistance of some little capacitor on a control board misbehaves—it just causes a blip, or causes a mean shift, or something in the process. Maybe it doesn't produce bad parts yet, but something has definitely changed.
 
SPC will point out when that changes, and you need to go in there and correct it immediately. Sure, these types of things are out of the norm, but they happen enough. And you’ve got to be prepared. SPC helps you be prepared for anything, because it’s looking for anything outside the norm.
 

Now Is the Time for Statistical Process Control

In closing, I would add this: Don’t wait until customer complaints start rolling in before employing SPC. Your processes can tell you all you need to know to produce high quality products all the time—if you keep an eye on them. Focus on the data points that define your processes and set up your SPC software to detect any variations. Make sure you’re measuring the right things in your processes in order to get the desired results.
 
Sometimes it’s subtle, easy to overlook. SPC software won’t miss it. SPC will help you maintain consistent quality across your entire organization. And, before you know it, you’ll be talking about that breakthrough improvement that vaulted your organization into the stratosphere. Happy hunting and manufacturing!
 
 
To read other blogs in this series:  
Take advantage of the technology at your fingertips today: contact one of our account managers (1.800.772.7978 or via our website) for more information.
 

 
 

InfinityQS Fact Checking Standards

InfinityQS is committed to delivering content that adheres to the highest editorial standards for objective analysis, accuracy, and sourcing.

  • We have a zero-tolerance policy regarding any level of plagiarism or malicious intent from our writers and contributors.
  • All referenced articles, research, and studies must be from reputable publications, relevant organizations, or government agencies.
  • Where possible, studies, quotes, and statistics used in a blog article contain a reference to the original source. The article must also clearly indicate why any statistics presented are relevant.
  • We confirm the accuracy of all original insights, whether our opinion, a source’s comment, or a third-party source so as not to perpetuate myth or false statements.


 

SUBSCRIBE TO OUR BLOG

Never miss a post. Sign up to receive a weekly roundup of the latest Quality Check blogs.

Archives