In the United States, an estimated 38 percent of food—around 89 million tons—goes unsold or uneaten, and a majority goes to waste. That’s almost 145 billion meals’ worth of food, equivalent to $473 billion or about 1.8 percent of U.S. gross domestic product. The production, transport, storage, preparation, and ultimate disposal of all this food produces the same amount of greenhouse gas emissions as nearly 100 coal-fired power plants, making it a significant contributor to climate change.

Fortunately, there is an abundance of solutions to this global problem. And a handful of these solutions are incorporating artificial intelligence (AI) to analyze data—such as inventory, sales, or even images of discarded food—and generate valuable insights to drive greater reductions in waste. Machine learning, a subset of AI, has the capacity to process significantly more data than a human and can, for example, to make predictions like how many apples would sell on a rainy day in October in Boston based on past trends. Solutions incorporating AI include:

Enhanced Demand Planning: This solution is used primarily in the retail sector with the potential to be applied to the foodservice sector. It utilizes past data to analyze ordering trends against purchasing behavior, giving retailers the ability to order more effectively and efficiently. In 2022, the Pacific Coast Food Waste Commitment conducted a pilot study with two large retailers that used AI solutions Shelf Engine and Afresh to improve order accuracy, which led to a 14.8 percent average reduction in food waste per store. If the entire grocery sector were to implement these solutions, an estimated 907,372 tons of food waste could be prevented, representing 13.3 million metric tons of avoided CO2e emissions and more than $2 billion in financial benefits for the sector.

Shelf-Life Extension: This solution is used primarily in the manufacturing and retail sectors, and employs sensors and other technology to collect data about when food spoils. Once the data is analyzed, users can make a more informed decision about when to move food from ripening rooms to transportation vehicles or store floors. For example, Strella offers a solution to support banana ripening through the use of proprietary sensors and machine learning to approximate the stage of ripening; then it issues recommendations that help ripeners improve consistency of color and maintain banana shelf life.

Waste Tracking and Analytics: This solution is used primarily in the manufacturing and foodservice sectors. Cameras capture and track waste, identifying hotspots that users can focus on to implement waste reduction efforts. Winnow offers an AI-enabled waste tracking solution that recognizes food being thrown away without the need to change operational procedures. Their solution has helped companies like Ikea and Guckenheimer cut food waste in half. Leanpath, another waste tracking and analytics provider, recently integrated generative AI into its platform to analyze a kitchen’s food waste data and recommend specific actions to prevent that waste from reoccurring. Its new AI-enabled features include the use of smart, probing questions to engage culinary team members about their ideas on how to reduce wasted foods.

These are just a few examples of how AI is being integrated into solutions, and we’re just beginning to understand the full potential of AI as a tool to drive reductions in food loss and waste. It will be exciting to see how this industry develops in the near future.

To learn more about solutions to food waste, including those that incorporate AI, visit the ReFED Solution Database and sign up for the ReFED newsletter to get updates delivered to your inbox.