Connecting Machine Data Into Actionable Information for Food Safety

When Novolyze was founded in 2012, the industry did not have a good option for validating low-moisture foods like cocoa or spices, so the company created and patented a surrogate bacteria that could be used in low-moisture environments. Since then, the company has gained more than a decade’s worth of experience validating hundreds of types of foods for many of the world’s largest food companies, primarily in Europe.

Novolyze relaunched about a year ago with new headquarters in the U.S. and a new platform of services. While validation studies continue, the company’s core platform now is around leveraging Internet of Things (IoT), machine learning, and other technologies to digitalize food safety and quality processes.

Robert Wallace, NovolyzeNovolyzeThe company is working to solve the connectivity issue created by a predominance of legacy hardware in food and beverage manufacturing. “So many of the ovens, the systems, the dryers—the bits and pieces on the plant floor—they may have data they capture,” says Robert Wallace, vice president of product management. “But that hardware doesn’t necessarily talk to each other—it doesn’t talk to a mothership, it doesn’t talk to a dashboard, it doesn’t talk to anything. If you’re lucky, you can get reports off some of it. Some of those trends are changing, but there’s such a huge amount of legacy hardware out there.”

With a gateway connected to plant floor equipment and to servers, Novolyze can monitor more closely various control parameters. “You get access to a SaaS offering with a series of dashboards that feed from one or more plants into our dashboard so that you can begin to monitor that long tail,” Wallace explains. “You can see what’s going on within your plant operations—what’s working, what’s not working, how long it’s been running. We do a lot with statistical process control, other KPIs, etc., so that we can help you tune your process.”

This is not the sort of capability that many food and beverage producers have in their plants today, according to Wallace. “Typical for the last 20 years is the clipboard—what’s the temperature at 5 minutes, what’s the temperature at 10 minutes. You’re using your pen and paper to record the performance of the device,” he says. “That’s where we are today. If you were to look at the food manufacturing business in general, a good 60 to 80% of it is, frankly, in this pre-digital stage.”

Reams of data

“We’re helping companies deal with the labor shortages, but we also help them with efficiency and help them with reaching sustainability targets because a lot of these companies have zero emissions sustainability goals they’re trying to hit around water and power. So we’re taking all those paper manual processes and we’re making them digital,” notes Karim-Franck Khinouche, founder and CEO of Novolyze. “There is an unmet market opportunity to really get in there and help these companies—some of them with decades of experience in processes that are kind of still stuck in the 1980s.”

Karim-Franck Khinouche, NovolyzeNovolyzeSomething we’ve heard over and over again is that manufacturers are gathering tons of data, but they’re not able to do much with that data. “We see these big companies with reams of paper data, but they’re not able to tell a story with it; they’re not able to extract anything useful from this data,” Khinouche says. “Why are you logging all this data—all these positive and negative test results—if you can’t see patterns, if you can’t help yourself understand where in the plant floor there might be a higher risk of some contamination in terms of environmental monitoring, or when you’re seeing a batch problem?”

Instead, manufacturers have the opportunity to use that data to analyze trends, learn something for the future. Down the line, that analysis might help you put actions in place to avoid product holds or recalls.

The Novolyze platform is well suited to providing information across an entire manufacturing operation—bringing data from all the equipment to one view to improve throughput, efficiency, and sustainability.

“Today, the food manufacturers have so many different vendors coming in, and each vendor has its own little piece of software—the lab has a piece of software, the equipment has a piece of software, the oven, the extruders, they all have these pieces of software, but they don’t talk to one another. So they’re very limited in their scope of what they can oversee,” Khinouche says. “If you’re a food manufacturer, maybe you have 20 different vendors who are covering everything from supply chain logistics to testing to sanitation to equipment, and you want to be able to put that all into one pane of glass over the plant. And that’s what this software does.”

Protein’s sweet spot

Novolyze is finding that its process control platform is particularly well suited to the protein sector, which has seen the biggest gains from its capabilities. “We’ve really excelled in the thermal processes specifically with proteins and being able to really optimize the process so that they’re cooking the product just enough that it maintains its quality, that it maintains its safety, but not overcooking it,” Khinouche says, explaining that this really comes into play on the sustainability side, where manufacturers have an opportunity to optimize energy, water, and chemical use. “We’re trying to help food manufacturers validate their processes, whether it’s the kill step process or it’s the testing process or it’s the sanitation process.”

Whether it’s in sous vide equipment, ovens, or other processes, protein producers are dealing with a range of bacteria such as listeria or salmonella. “So the opportunity for them is to say, ‘I’ve cooked it well enough, I’ve cooked it long enough, I’ve retained the flavor, I’ve retained the nutrition, and I’m not wasting as much as I used to waste,’” Wallace says. “If you look at the data, something like 30% of our manufactured food is wasted. I’d sure like to be able to put a bigger dent in that.”

Proteins here mean not only ready-to-eat meats, Khinouche notes, but also plant-based proteins. “It’s very difficult because you want to kill the pathogens, but you also have non-pathogenic bacteria that is surviving. You want to make sure it’s safe, but you don’t want to make it so that it’s inedible,” he says. “There’s a lot of optimization that we see that can happen for the protein manufacturers.”

With each new customer, Novolyze visits the plant and studies the process to show, through modeling, how throughput efficiencies and cook times will improve. “Every case that we’ve been through so far with protein has shown that using our system is truly going to save that customer money and time,” Khinouche says. “We’re able to help them with managing labor shortfalls, and we’re able to make sure that everything is traceable and compliant with regulations.”

Novolyze sees its technology on a continuum of progress for the industry—moving from paper records and Excel spreadsheets to data digitalization and aggregation and on to offerings like this platform that provides cloud services and trend analytics. “Next is we’re going to get into the advanced analytics, because now we have cloud services, both predictive as well as prescriptive analytics—here’s what we think is going to happen and here’s what we think you should do,” Wallace says. “And then we get into the AI and machine learning element of this. For instance, we’re talking with a very large manufacturer that’s saying, ‘We want to release our stuff sooner. Can you help us do that?’”