Walmart’s Cash-and-Carry store operation Sam’s Club has implemented AI receipt verification technology in 120 locations. This represents approximately 20 per cent of the retailer’s stores. The solution, developed in-house by Sam’s Club, uses AI and computer vision recognition to confirm that Scan&Go shoppers have paid for all items in their shopping carts and eliminate queues at the exit area. The retailer plans to roll out the solution to all of its 600 stores by the end of the year.
Previously, employees had to manually check each receipt at a dedicated Scan&Go area at the exit, resulting in long waiting times for customers. Now Sam’s Club’s newly developed technology captures images of carts using a combination of AI and vision recognition to verify payment for all items in a shopper’s basket. The system thus helps to reduce sales losses due to shrinkage and speeds up the self-checkout process. In stores where the technology has been deployed, customers leave the exit area 23 per cent faster than before.
After paying, Scan & Go users at Sam’s Club pass through a high exit gate on their way out of the store. A camera is mounted on the gate above the customer to identify the products in the shopping cart. The technology captures images of the basket and transmits them to an employee’s handheld device. The store associate sees a list of the products the customer has purchased as well as the images of the basket on the device. The technology can flag an item and alert the employee to do a scan check of the basket.
Challenges with full shopping carts
It is questionable whether Walmart’s subsidiary can achieve reliable results using only computer vision recognition. It is unclear how the system performs when multiple products are stacked on top of each other in a very full cart, especially since Sam’s Club does not put a limit on the number of items that can be purchased through Scan&Go. A video from Walmart showing how the technology works depicts shopping carts with only a few items.
Supersmart, a provider of similar verification technology, works with a floor-mounted scale that uses self-learning algorithms to check whether all items have been scanned correctly at the self-checkout. In April of this year, The Retail Optimiser reported on the first Spar supermarket in Israel validating self-scanning with Supersmart.
