How AI Helps Cut Food Waste

By Mark Hamstra

About 60 million tons of food in the U.S., or 29% of the total food produced, went uneaten or unsold in 2024, according to ReFED, a nonprofit organization dedicated to reducing food waste. Food surplus within all food industry sectors was worth about $230 billion, the organization said.

Artificial intelligence (AI) is now playing an important role in reducing food waste in retail and foodservice environments, primarily through inventory optimization and predictive modeling.

“The biggest thing that AI is doing is helping to remove the guesswork from anything leading up to the actual distribution of food,” said Angel Veza, director of innovation at ReFED. “The purchasing, the prep, the inventory decisions—AI can help make smarter decisions around those processes so that you are meeting demand and not generating surplus.”

Proof of Concept

Wawa is among a number of retail companies that have deployed data analytics technologies, including AI, to cut food waste. The retailer partnered with Relex Solutions to roll out a forecasting and replenishment solution aimed at reducing fresh food waste. The system uses algorithms drawing on multiple data streams to automate demand forecasting at stores, according to a case study published by Trax Technologies.

In another instance, a recent test at two large retail chains using AI-driven technologies from Shelf Engine and Afresh reduced food waste by an average of 14.8% per store, according to a report from the Pacific Coast Collaborative. The test included more than 1,300 stores and focused on produce, although one of retailers also studied food waste in the deli department.

In addition to cutting food waste, the tests, which sought to optimize inventories to meet projected demand, also yielded improvements in sales and added some labor efficiencies, including reduced ordering time and less time spent managing shrink and restocking.

Applications for AI in the Kitchen

Restaurant operators have increasingly deployed AI solutions to help predict prep levels and optimize inventories.

83 Subs, a Jimmy John’s franchisee operating several restaurants in Kentucky, Ohio and Indiana, deployed a solution from technology provider ClearCOGS to help manage prep levels for the restaurants’ sandwich-making operations. For example, the technology generates prep sheets that help managers determine if they need to bake another cycle of bread as the day winds down. The solution decreased the operator’s bread waste by 53%, according to ClearCOGS.

Veza said that in addition to cutting costs by reducing food waste, AI-based inventory planning can also free up labor that can be better deployed to focus on enhancing the customer experience.

“I think that’s really what AI is doing for these industries, and that’s the value that it can offer their customers when it’s done well,” she said.

Veza also emphasized the role that AI can play in minimizing the impact of food waste on the environment. Not only does it help reduce the amount of food that ends up in landfills, but it also stands to help optimize food production further upstream, so that less resources are used in growing, processing and transporting food that ends up going unused.

AI also holds promise as a tool to add efficiencies throughout the supply chain, including in areas such as delivery route optimization, sourcing and operational efficiency, according to a recent IBM Think article.

Overcoming Challenges of AI in Foodservice

Key to the success of using AI to reduce food waste is that the data used by the AI is clean and accurate. This can sometimes be a challenge when it comes to fresh-product inventory and seasonal items, according to the Pacific Coast Collaborative report.

Another hurdle is the cost of the systems, said Veza, many of which may not be accessible to all operators due to pricing.

“When you see that initial cost, it can be hard to understand what the potential ROI could be,” said Veza.

She suggested that chains pilot AI solutions in a small number of locations first. If they can achieve good results, it can become an easier sell for additional locations, she said.

Increasingly, however, technology companies are incorporating AI into preconfigured, modular solutions, which is reducing upfront costs, she said.

Another challenge: operating procedures at the store level need to be adjusted to make the best use of the data that the AI platform is generating. For example, if the AI suggests prepping a certain amount of something, that forecast needs to be incorporated into the location’s standard operating procedures.

“There has to be a step where the prep cook says, ‘I’m going to check and see what the dashboard says,’” said Veza. “It’s a simple change that just needs repetition, but once it becomes part of the daily practice, then you’re going to truly see the results.”