Kroger Steps Up AI Deployment to Combat Checkout Losses

Editor’s Note: Billions of dollars are lost by stores each year due to items not being properly scanned at checkout. Just as Badger autonomous robots rely on our eagle-eye computer vision and AI to improve the accuracy of inventory shelf scans, digitally savvy grocers like Kroger are working with the same technologies to further streamline and improve the checkout process.


  • Kroger has signed a deal with Irish artificial intelligence firm Everseen to use the company’s technology to help stop items from passing through self-checkout stations in the grocer’s stores without being scanned, according to an emailed press release.
  • Everseen’s system uses computer vision cameras to detect if a shopper fails to scan an item and discreetly alerts a store employee, who can intervene before the customer leaves without paying for the merchandise.
  • ​Kroger began installing the Everseen equipment in stores in March and plans to deploy it at as many as 2,500 supermarkets as part of an effort to combat shrink.


Kroger’s arrangement with Everseen is part of a broad effort by store operators to stem the billions of dollars in losses they incur when shoppers and store employees neglect to scan items at check-out counters, either intentionally or by mistake.

Retail shrink is rising at a fast clip, with losses attributable to shoplifting, theft by store workers and organized retail crime amounting to $61.7 billion last year — a more than 20% increase from $50.6 billion in 2018, according to figures released in July by the National Retail Federation. Put another way, shrink ate up 1.62% of retail sales in 2019 compared with 1.38% in 2018, the organization reported.

In an illustration of the challenges they face in reducing those figures while also demonstrating their tech savvy to consumers, retailers are hoping that tools such as artificial intelligence and high-resolution cameras will help them reduce theft as well as automate the checkout process. 

Companies like Standard Cognition, Amazon, Grabango and Mashgin have come up with ways to let customers entirely bypass checkout counters with technology that is able to keep a closer watch than humans can on merchandise. For example, the cameras Grabango uses to track what people remove from shelves can see the same level of detail as the human eye, enabling the system to identify products even if they look highly similar to one another, according to the company’s chief business officer, Andrew Radlow.

Retailers like Sam's Club and New York's Westside Market have integrated computer-vision systems with their scan-and-go programs to recognize products faster, while Amazon Fresh's Dash Cart uses similar technology to log each item as shoppers place it in the cart.

A video on Everseen’s website notes that while the eye is remarkably good at discerning color and other details, it is able to focus on only one image at a time — a key disadvantage that the company says its technology is designed to overcome.

Methods to spot retail theft have grown in importance for grocers as the industry looks for ways to speed up the checkout process. Stores have used self-checkout stations for years as a way to reduce friction, and Walmart recently eliminated traditional checkout aisles in one of its stores to test the reaction from consumers. The coronavirus pandemic has also put a spotlight on self-checkout systems because using them can help customers avoid contact with store employees.


This article was written by Sam Silverstein from Grocery Dive and was legally licensed through the Industry Dive publisher network. Please direct all licensing questions to