Modern SPC software is a crucial tool for packaging manufacturers, especially those that have plants scattered across the globe. Gaining visibility within and across plants helps quality managers maintain quality control, which keeps customers satisfied and the enterprise itself cost-competitive.   

Moreover, a cloud-based SPC software solution can bring all production facilities, including recent acquisitions, into one single, standardized system. The automated data collection and analysis capabilities of on-demand SPC software can drive proactive improvements early in the production process, and in turn improve customer relationships and the enterprise’s bottom line. These improvements include significant reductions in waste, scrap rates, process variability, recall time and costs, and customer complaints. 

Company example: Sonoco

Sonoco is a global provider of packaging and product services, headquartered in Hartsville, South Carolina. Jennifer Desormeau is a quality engineer at Sonoco and primarily supports the company’s plastics division, Sonoco Plastics.

Seven plants in this division currently use ProFicient on Demand, a software-as-a-service SPC platform from InfinityQS. Three of these plants began using the software in 2018, Desormeau says, and at least two more plants are slated to begin using it by the end of the year.

Quality asked Desmoreau about some of the improvements Sonoco has made since implementing the software.

Sonoco

Standardization and Customization across Plants

Many of the plants within Sonoco Plastics are acquisitions, Desormeau says, and came in with different software systems already in place, including different quality systems.

“We had one plant that was using a system that we no longer supported, so they had to do something about it—upgrade or replace it,” Desormeau explains. That’s when the decision was made, around 2014, to standardize the division’s quality system.

“We looked into different systems, including InfinityQS, and that’s the one we went with,” Desormeau says. “As a company, we’re not forcing any of our plants to switch to InfinityQS, but highly encouraging them that if they get to the point where they have to do an upgrade or replace it, then this is what they should be looking at.”

Part of Desormeau’s job is to keep an eye on the plants through the software, which is organized into projects.

“For projects that deal with data entry, we try to standardize them somewhat in terms of providing control charts that are specific to the people doing the data entry, and then creating other projects,” Desormeau says. “What I’ve tried to do, for the most part, is give specific reports that shows scrap by category, or one or two key factors, and then put in some generic reports—I call them templates—so someone at the plant can go into the software and say, ‘Here is my template for capability analysis, but I also want to see this part, and see this particular test, and I want this date range,’ and they can modify that template as needed.”

By providing “blank” versions of most of the charts, Desormeau says the user can then change the data selection, and perhaps the processing options as well, to see the data the way they want.

“These are used more for data analysis and will change depending on what you are reviewing any given day,” she says. “By creating the templates, it reminds the users of the different types of charts available.”

Quality Improvement through Data Collection and Analysis

Desormeau offers an example of a Sonoco plant that improved their quality process by combining the Sonoco Performance System (SPS)—a data-driven system focused on identifying and eliminating losses—with the data and tools of ProFicient on Demand.

This plant performs visual inspections where they keep track of the defect code and quantity of parts found with that code, and detect the top codes by using a Pareto chart.

“This particular problem was #2 on the chart and corresponded to a variable check which has a Cpk value of 1.17,” Desormeau says. “The actual amount scrapped for this code was much higher than the expected based on the Cpk. This lead to the realization that the defect code needed to be better defined and that the scrap could be reduced.”

In this instance, the plant was able to utilize the data collected by ProFicient on Demand to pinpoint the root cause of an issue, focus their improvement project, and use their current spec and process limits to develop a better gage for the check.

Desormeau also notes that some plants use ProFicient on Demand to track and classify their internal rejects, and that all plants using the software receive real-time email notifications whenever a product is out of spec.

Increased Speed and Accuracy

Direct gauge input, as opposed to people typing in the data, also has been helpful for improving the speed, accuracy, and quality of several Sonoco plants’ production processes.   

“There’s definitely an improvement in the quality of the data, and it definitely speeds up the data collection—to not only have the data going in somewhat automated, but to have the measurements somewhat automated,” Desormeau says. “I’d like to see that happen for other plants and for other characteristics. But knowing that’s possible, and that we were able to do that, was a big step.”

Direct gauge input is happening in all of the Sonoco Plastics plant using InfinityQS, Desormeau says, and all handheld gauges are linked directly, including scales, micrometers, calipers, digital indicators, and MagnaMikes. 

“The gauges either have a footswitch or a data send button that pushes the data to the software,” she says. “This has allowed us to push the measuring of the parts to the production floor in some plants, meaning that the people making the parts are the ones doing the measuring, rather than in a lab.”

Desormeau also is interested in doing more remote data collection, citing a recently implemented dryer monitoring system as an example

“One of the plants put in this new system to monitor dryers [for plastics], so we were able to automatically pull that data into the InfinityQS software and do a moisture test,” she says. “Part of it was making sure the dryers are working properly, but when we do those moisture checks, we can tie it to that dryer data that’s already in the system. Every hour we get the dryer data automatically; nobody has to do anything. We want to be able to do that with some of our other systems.”

Most of the equipment linked to ProFicient on Demand has some kind of monitoring on it, Desormeau says, but most of the monitoring is done in real-time, and not stored for later.

“If we can collect that data on some frequency, we could use that to do real SPC: statistical process control, not statistical product control,” she says. “Most of what we do in terms of process data right now is manual collection, so we’re trying to get more automated.”

A big part of that transition will likely be hardware-related, Desmoreau notes, with more automated measuring and in-line gauges from which to glean actionable intelligence.

“The ideal would be to use more vision systems, as opposed to handheld gauges,” she says. “We have made progress in terms of good data, but we still have a way to go, because there’s still measurement technique [to work on] that’s also part of the process.”

Giving Data a Second Life

Even with the most sophisticated software, data collection alone isn’t worth much—what matters is how you use it. That’s why Desmoreau gives purposefully generic templates to plants’ quality managers: to trigger them to ask questions about their specific reporting needs, and  to empower them to decide what they want to with the data they’re collecting. Too often, people take only the initial steps, she says, like primarily using data for control checks and only coming back to it if there is an issue or customer complaint.

“Some of the goal is trying to change that philosophy in the plants—to get them more data-driven and to actually look into the data, not just collect it,” Desmoreau says.

Her advice: Have the data available on a cloud platform so you can not only see it, but also go back to it, analyze it, and then mine it for a continuous improvement project.