Advanced Computing in the Age of AI | Tuesday, April 16, 2024

Taking Aim at Retail’s $40B Annual Non-Scanning Losses 

By hook, crook, negligence or equipment malfunction, $40 billion worth of goods pass through checkout lanes at retail stores around the world each year without being scanned. Be it the fault of consumers (using self-checkout machines) or employees using POS systems, non-scanning cuts into retail’s thin margins, particularly those in the grocery sector.

Everseen, of Cork, Ireland, is emerging from stealth mode after seven years of development with a cloud-based system combining AI, deep learning and analysis of video from in-store cameras that the company says notifies retail store staff of non-scanned items within three seconds. In pilot projects at five of the world’s 10 largest retail companies in the world – all of them grocery chains, four of them in Europe and one of them in the U.S. – Everseen claims a 90 percent success rate detecting items that have not been scanned.

“Retailers are losing $40 billion a year through shrinkage at the point of sale,” said Professor Joshua Bamfield of the Centre for Retail Research. “Products non-scanned are not only impacting profits but also creating supply chain issues leading to out-of-stock products. At the moment, retailers are almost certainly haemorrhaging revenue through their POS and are probably completely unaware of how it is happening, how they prevent it and the scale of the problem.”

The problem is most acute at self-scan checkout lanes, with up to 10 times more non-scans compared to checkouts staffed by cashiers, according to the company. Everseen said it identifies non-scans by using its video analysis and artificial intelligence system that visually recognizes non-scans using existing CCTV feeds and linking them to checkout transaction data.

Alan O’Herlihy, founder and CEO of Everseen told EnterpriseTech the company’s core innovation was combining POS and video data, and then training the system to detect non-scan outcomes.

Everseen's Alan O'Herlihy

Everseen's Alan O'Herlihy

“We take the video stream from the camera and we put it into our cloud-based server,” O’Herlihy said, “then we take the POS information from checkout that the retailer streams to us. We have every line item time-stamped when that product was scanned, so every time someone scans a product it goes into a log file and we receive that data. It’s like a heartbeat, down to the millisecond. We compare it with analysis of the video stream data. You have the POS stream and the video stream, you have a heartbeat of both, and you throw out the differences.”

The system identifies when and where the non-scan took place, and a video of the incident is then sent in near real-time to the mobile devices of store staff, who can discuss the problem with the shopper.

He said most retailers simply don’t know why losses are occurring, so “they’re acting blind – trying to prevent losses but without any real direction.”

Everseen also announced it has secured investment funding from MARCOL Capital Europe, London, along with other investors.

“Our due diligence has shown that this solution has the potential to completely transform the market,” said Pii Ketvel, CEO of MARCOL. “Everseen allows retailers to move more quickly to self-service check-outs by making the user experience better, whilst helping the retailers to reduce their losses.”