Approximately 48 million people every year get sick from foodborne illnesses, according to the U.S. Centers for Disease Control and Prevention (CDC). The CDC says the most common causes of food poisoning are foodborne pathogens like Salmonella, Listeria and E. coli. While some incidents are the result of poor food preparation, hygiene, and handling at home, many cases are caused by kinks in the food system itself. To find improvements, the food industry is turning toward artificial intelligence (AI) and edge computing.
Edge computing and new AI food safety applications help manage these events by leveraging big data in food safety practices across the food system, from the farms that produce food and the trucks that transport it to the restaurants that serve it and the retail stores that sell it.
How Edge Computing and AI Help Farmers
Farmers and food manufacturers can use edge computing near food production areas. Rather than sending collected data to a centralized network, this real-time data processing in agriculture can improve efficiency and quality control. As both edge and AI grow in usage, the two pair up nicely. For example, a farm or restaurant could combine surveillance cameras with processing capabilities on the edge with AI capabilities like facial recognition to quickly identify potential issues.
Dr. Nadia Sabeh, president and founder of Dr. Greenhouse, works with Forever Feed Technologies to develop automated sprouted grain (ASG) to feed more than 5,000 cows. Instead of in a field, which would require more horizontal space and a larger footprint, the ASG is grown on shelves. The organizations use sensors in the feed mill to identify which specific plant groups need water, light or cooling.
Though AI is beneficial for tracking, researchers must account for multiple variables. If a model is trained on what happened in February, for example, and March traditionally has different temperatures and precipitation levels, can the model adjust efficiently?
“What’s most important to me is the consistency,” Sabeh said. “I’d be interested in a computer model that said, ‘This is what you need to do not just to increase your yield, but to have the most consistent yield.’”
Visual models benefit production, too. Ben Miller, COO and EVP of scientific and regulatory affairs at The Acheson Group, said crop researchers can use microscopy to check for pathogens. These visual models could help prevent foodborne illnesses, such as recognizing Salmonella or E. coli on a lettuce leaf.
“You can pull that leaf off, put it under a microscope and run your visually trained model to say, ‘I see something. How does that compare to what I’ve been trained on?’” Miller said.
Organic Valley manages a network of more than 2,000 family farmers. The company uses Nutanix software to simplify operations and improve performance, offering ERP solutions, warehouse management systems and supply chain analyses. With the extra support, it can work on other initiatives.
“Our network, storage and compute teams were all working weekends and doing updates because there was a fear of breaking something in a complex operation,” said Nicholas Korte, director of business transformation at Organic Valley.
“When we went to Nutanix, it was like, ‘Hey, we have an opportunity to simplify and take that time back for ourselves.’”
AI in the Food Supply Chain
Even after items depart farms and factories, food safety measures remain critical. Traceability can reduce recall impacts and proactively get ahead of illnesses. For example, a Food Logistics report found that 98% of cases of people who fell ill in the U.S. in 2024 could be traced to 13 outbreaks. Tools like Index Biosystems’ BioTags connect product data to trace the entire supply chain.
Tyler Williams, CEO of ASI, said that visibility is essential for maintaining proper safety standards.
“If I have milk, for example, I know the temperature when it’s sitting in storage, when it’s in transit and when it’s at the store,” Williams said. “If there are any variances, we’re able to catch those things faster across hundreds of thousands of different SKUs.”
En-route traceability is one way to use AI for foodborne illness prevention. Shelf-life risk for perishable items is another. Much like retailers utilize the edge to enhance customer and employee experience, manufacturers are adopting edge computing in food safety to identify potential issues.
“Larger manufacturers use edge computing to monitor and predict recalls,” Williams continued. “If four people got sick, they can see where they purchased from and be more accurate in identifying the culprit.”
Forecasting in Restaurants and Retail
Recent years have seen an influx of cloud concepts for restaurants, from data sharing to ghost kitchens. Restaurants and grocers can implement AI to help with restaurant inventory software management, customer service and automation, such as robots assisting with prep work or packaging.
Other restaurants are using AI tools for recruiting and onboarding via AI chatbots and virtual assistants or to simplify scheduling, resulting in more efficient back-end operations.
Nathan Jarvis, a former chef and founder of Jarvis Hospitality and Advising, said AI has strategic value in predictive forecasting.
“In retail grocery settings, automatic sensors provide minute-by-minute data and establish trends,” Jarvis said. “They can predict when a refrigeration unit will go down and catch it before the refrigerator dies.”
Using AI to Scale Food Safety Checks
AI technology also improves food safety for individual items. For example, a human might only periodically check a display of rotisserie chickens. With AI sensors in place, a retailer can more accurately read its products.
“They can know that each chicken hits the right temperature each time, instead of one thermometer hitting the right chicken,” Jarvis said.
Food processing requires critical control points where companies collect data at certain frequencies. Predictive analytics could help identify a faulty piece of equipment or other malady.
“More immediate opportunities are in engineering failure point analysis, so you can predict when a part of the process might go off the rails from a food safety standpoint,” Miller explained.
“Anywhere you have a well-trained model looking at a particular problem and relatively high-quality data, something can be done.”
Implementing AI Agents and GenAI
Deloitte’s State of AI in Restaurants report found that 82% of restaurant leaders expect to increase their investment in AI next year, with more than a third actively using or testing Generative AI (GenAI). Common uses of GenAI include personalizing menus or sending marketing updates autonomously.
Some restaurants have also begun experimenting with AI agents. Yum! Brands, which includes KFC and Taco Bell, deploys AI agents to help call centers during demand surges. The agents can digest natural speech, process complex menu orders, and recommend additional items to users. The agents improve customer satisfaction and order accuracy, and can alleviate some of the pressure on busier locations.
In June 2025, the U.S. Food and Drug Administration (FDA) launched Elsa, a GenAI tool designed to help its employees work more efficiently. Among other tasks, Elsa can “accelerate clinical protocol reviews, shorten the time needed for scientific evaluations, and identify high-priority inspection targets.”
The FDA’s focus on the power of AI will likely result in similar capabilities extending to restaurants, retailers and other companies in the food space.
The Future of Food Safety is Sustainable
Food safety and sustainability go hand-in-hand. While the horizon is bright, the industry must maintain realistic expectations.
“People make these big claims,” Sabeh said. “We’re going to reduce water use by 94%, land use by 90%, blah blah blah…”
“Reducing water usage by 25 or 50% is a big number, but those lofty goals are a disservice to us.”
Emerging categories in food science are also worth watching. Jarvis pointed to mushroom mycelium, which modifies proteins to remove graininess and flavor changes. Fungi-based options require up to 99% less energy, land and water than compared to raising cattle.
“It performs better in a product, tastes better and takes away some of the negative aspects of plant-based proteins,” Jarvis said. “We’re going to find all sorts of interesting new uses and possibilities.”
There’s still plenty of research and development to be done around AI food safety practices. Williams stressed the importance of sound data collection and using AI as a complement to human experience and reasoning.
“The information coming out is only as good as what we’re putting into it,” he said. “AI shouldn’t be your whole program. Tailor it to your processes and operations.”
This is an updated version of the article originally published on April 2, 2024.
Editor’s note: Learn more about Nutanix’s AI platform, GPT-in-a-Box, and hybrid multicloud.
Joey Held is a writer and podcaster based in Austin, Texas, and the founder of Fun Fact Friyay and Good People, Cool Things. Connect with him on Twitter or LinkedIn.
© 2025 Nutanix, Inc. All rights reserved. For additional information and important legal disclaimers, please go here.