SPC for predictive quality control

Getting into SPC is really about assessing quality, using quality metrics and not just focusing on measuring one part. Photo courtesy of Hexagon.

The story of quality really started to prevail right after Henry Ford introduced mass production into manufacturing. This created the situation whereby if the components and sub-assemblies didn’t fit together, the line had to stop, and somebody had to fix it. And this gave way to quality control departments. Fast-forward to the ‘30s with Walter Shewhart, the “grandfather of total quality management,” who developed statistical quality control and process control charts.

It wasn’t until the 1980s that another significant jump in quality management occurred with the diffusion of the ISO 9000 standards for quality assurance. In tandem, large multinational companies in the automotive sector like Ford, Chrysler, and General Motors developed the QS9000 standard, which was developed from the Ford Q 101 Quality System Standard. However, after these standards began to take effect, the rise of the quality control department started to taper off and decline.

Now Industry 4.0 and the digitization of metrology is starting a new wave in quality management. Statistical process control software has enhanced capabilities to leverage the immense data available and has an updated interface for effective data analysis.

STATISTICAL PROCESS CONTROL

SPC is not new to manufacturing quality departments. The principles of SPC were developed in the 1930s when similar quality systems were emerging. This was long before computers came along, but the principles themselves remain similar today, even with the added benefit of automation and digital inspection.

“You don’t need a big investment in hardware, and you don’t need to be a statistical genius to do data collection for SPC,” said George Schuetz, director of precision gages, Mahr, Providence, R.I. “All you need, really, is an indicating gauge, a pad, and a pencil to start.”

It’s important for shop owners to understand what exactly they hope to achieve by integrating SPC software and data analysis on the shop floor. For the most part, shops already have a basis for data collection, usually in the form of pen and paper.

“SPC today really merges and manages the various smart technologies, which can optimize the interactions between people, standards, process, and technology to achieve desired results, better customer satisfaction, and higher profitability for the manufacturing company,” said Tom Stewart, president, Q-DAS, a subsidiary of Hexagon Manufacturing Intelligence, North Kingstown, R.I.

For Stewart, one of the advantages of incorporating SPC on the shop floor goes with what he calls the best-kept secrets: When measuring diameter or geometry on the part for quality metrics, the thing you can control is the process or the process parameters. By assessing the process parameters and capturing the data, you then can correlate them with the quality measurement. Then you can start to monitor them and control them before you measure a piece part as a means to predict the quality instead of measuring the quality after the fact, which translates into predictive quality.

SPC software should do two things: collect data from metrology devices or tools and chart it out for data analysis.

“With SPC we can then look at trends over time,” explained Erin Farmer, business development specialist, Mitutoyo Canada, Mississauga, Ont. “We then can take pieces of information for specific metrics and say, ‘Hey, you might want to watch this, what happened here, or we should run this process in control.’ Or maybe the data shows that it’s a little out of control. Even if you are making good parts, it will help determine if and where you can tighten up aspects of how the parts are actually made to really get the best out of the process. It’s really understanding what you’re looking for. And the software puts the data into charts while making it very customizable the way a user wants to see it.”

Mitutoyo’s MeasurLink SPC software centralizes quality data and allows for analysis of data in real time. Photo courtesy of Mitutoyo.

Many SPC features and interfaces have been updated to provide better support and make it more user-friendly, which should make the idea of integrating it a little less daunting.

“I am continually surprised at how often shop owners tell me they would like to do data collection for SPC, but they’re ‘just too small,’ or they ‘can’t afford all that fancy equipment,’” said Schuetz. “There seems to be a mistaken perception out there: that SPC is a lot more complicated than it is and that you need expensive hardware and a sophisticated computerized system online just to get into it.”

Getting into SPC is really about assessing quality, using quality metrics and not just focusing on measuring one part. It’s great that that one part is correct, but all parts need to be in tolerance. SPC helps determine optimal process parameters, such as ambient temperature, bearing temperature, vibration, feed force, and so forth, so that shops can adjust the process and predict quality outcomes.

DIGITAL METROLOGY

As machine shops continue their foray into Industry 4.0 technology, inspection and metrology also are being adapted to fit this new and exciting endeavor. Schuetz pointed out that today, the pad and pencil can be eliminated because data collection can easily be automated and integrated into existing inspection processes. Digital data collection helps remove operator and copying errors often associated with manual collection.

“Digital tooling is what we look at first when talking to customers about upgrading their current SPC techniques,” said Farmer. “If all the customer has is a manual caliber with an analog scale, they won’t be able to automatically collect the data. They still can enter that specific measurement data into the software. However, having digital tools like micrometers, height gauges, calipers, and even machine tools that can connect, either wired or wirelessly, to pull the data out of the shop floor and get it into that data analysis software really makes a difference.”

Connect technology makes collecting data even easier, and the interoperability between measuring devices and the software is essential for large data collection. Farmer pointed out that you can’t just measure one or two parts and then expect the graphs to provide any real information. The most important feature of SPC is that it collects enough data and then analyzes the trends from it so a shop can make improvements over time.

When you’re measuring a lot of variables, it’s important to ensure a proper naming structure to filter through and best use the data.

“When measuring components, subassemblies, machine characteristics, tools, fixtures, and all the process parameters related to them, everything needs a proper name,” explained Stewart. “That’s one of the mistakes that [shops] make: putting together systems that are too dissimilar in the naming strategy of the data. If you just name them right to begin with, then they will flow into the database in a logical, structured way.”

GETTING STARTED

Schuetz explained that the first step is to look at the part you’re going to measure and determine what its important dimensions are and what part of the process controls it. It’s important for shops that are new to the process to start simple, with a single measurement, until they get the feel for it.

Choosing the best measurement tool for the job is essential in collecting accurate and relevant data. As previously mentioned, several flexible options are available for digital measurement tools to make the data collection process easy.

Q-DAS SPC software merges and manages various smart technologies, which can optimize the interactions between people, standards, process, and technology. Photo courtesy of Hexagon.

“You also can run studies to make sure your measurements are as accurate as possible,” Schuetz added. “Remember, whatever ‘analysis’ you do can only be as accurate as the measurements you start with.”

Farmer noted that it’s not just about collecting data; customers are now wanting to start to analyze their processes. For example, examining a machining process through SPC can help determine the reasons for certain results. The data can help account for adjustments that need to be made to reduce scrap or improve RPM of a machine, for example.

“All of the process parameters and measurement data that can come out of these machines and metrology devices will help a customer over time with SPC analysis,” Farmer added. “I think everyone who’s involved in manufacturing and machining already understands how important quality assurance, quality control, and measurement are.

Taking the data and doing something with it helps save money. Now, industry is also looking at in-process measurements/ real-time analysis and improvement, so they can fine-tune processes without collecting data after the fact.

The experts agree that SPC and digital technology are really helping to empower workers on the shop floor. The software enables them to take inspection or testing equipment and turn it into a quality assurance tool and ultimately a tool to perfect quality.

“Charting empowers the operator, and that, ultimately, is what SPC is all about,” said Schuetz.

Associate Editor Lindsay Luminoso can be reached at lluminoso@canadianmetalworking.com.

Mahr, www.mahr.com

Mitutoyo, www.mitutoyo.ca

Q-DAS, www.q-das.de