By Katherine Devine, Director of Business Case Development, WWF
ChatGPT has been all over the news for its ability to create well-written concepts with minimal prompting, leading many to herald a new era of artificial intelligence. But it’s not the only game in town in terms of innovative AI. WWF, in collaboration with the Pacific Coast Food Waste Commitment (PCFWC), Afresh, and Shelf Engine, conducted pilots using AI purchasing systems in two different grocery retail chains to reduce food waste and improve profits. The results were impressive: food waste was reduced by 14.8% per store on average.
The AI technology was able to reduce waste by increasing the accuracy of ordering and inventory requirements, leading to more precise ordering and stocking, and higher sales. Where traditionally these tasks have been left to human perception and calculation, AI churns through enormous amounts of historical and real-time data to better align ordering with demand. This also increases the freshness of the produce on the shelves, a great benefit for consumers. While the biggest waste reductions may be seen in early implementation, the algorithms only get better at predicting customer behavior over time, which can then help to reduce waste of seasonal items (think: pumpkins in the fall or holiday baking items) that may only be in demand once a year.
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.
When you consider that grocery retail is responsible for more than 10 percent of the kind of surplus food that leads to waste in the US (and that almost 35 percent of all food is wasted), the implications of using AI to further reduce grocery retail waste are considerable. Not to mention that it’s a relatively easy to implement solution once the AI providers map retailers’ data. It also more than covers its costs once it’s in place. And, in addition to increasing profit and reducing waste, the AI solutions also increased labor efficiency by up to 20 percent per store through automating inventory tracking and reducing the amount of time spent on ordering and restocking shelves. With time saved from these AI solutions, grocery retail can focus on better serving consumers, and educating about how to reduce waste in the home, where 37 percent of all food waste occurs in the US.
The pilot success has led one of the solutions, Afresh, to move beyond fresh produce and expand to meat, seafood, and other categories in grocery retail, enabling it to better serve its customers and create greater impact at scale. Shelf Engine is already working in categories beyond produce as well. WWF estimates that if the results from the pilot were replicated in these other departments across the retail industry, an estimated 1.1 million tons of food waste (equivalent to 2.8 million tons of CO2E emissions) could be avoided.
AI won’t solve all of our problems and, as critics of ChatGPT point out, it may be creating some of its own (1). But there are clear cases where it can have a considerable positive impact as seen in the PCFWC pilots, where retailers, their employees, and their customers are all seeing the benefits. It’s time for the industry to fully take advantage of these easy-to-implement AI solutions that are big wins for their pocketbooks and the planet.
To read the full case study, Using Artificial Intelligence to Reduce Food Waste in Grocery Retail, click here.
Recent Comments