Retail technology company Diebold Nixdorf has started rolling out an artificial intelligence (AI) system to spot shoppers deliberately or accidentally not scanning items at self-checkout.

Designed to prevent the most common sources of loss at self-service and traditional PoS checkouts, the smart vision technology will complement Diebold Nixdorf’s already-live AI-based products, which reduce friction during fresh produce scanning and age verification for restricted sales.

Bringing these technologies together on one platform should result in one of the most holistic anti-shrink offerings on the market with ability to scale. Diebold Nixdorf can deploy its AI-powered options for retailers without disruption through its existing in-store integrations.

Retailers have long recognised the potential that self-service technology offers in terms of customer convenience, optimised floor space and employees’ ability to engage shoppers. However, managing and mitigating the associated risk of theft at the self-service checkout has long been a problem. Poor shopper and employee experiences are also factors that retailers must consider when the self-checkout process is disrupted by the need for human intervention, stock inaccuracies and other points of friction. All this can detract from the convenience of using self-checkouts.

The AI-powered Vynamic smart-vision shrink reduction system (www.dieboldnixdorf.com/en-us/retail/portfolio/software/self-service/vynamic-self-service) is powered by the SeeChange (seechange.com) AI and machine learning cloud platform, and is aimed at satisfying retailers’ need to implement and deploy technologies across multiple store locations and geographies.

“Retailers are committed to innovating, but they want to avoid duplication of efforts and costly implementations,” said Matt Redwood, vice president at Diebold Nixdorf. “Any new innovation deployed inside stores not only needs to be seamless but must also enhance the customer experience. We designed our new AI-powered offerings based on insights we’ve gained deploying millions of PoS and thousands of self-checkouts across the major retailers we work with. We’ve left no stone unturned when it comes to enhancing customer and employee experiences, as well as retailers’ need to scale easily and improve efficiency.”

The AI-based offering can combat common causes of loss, including theft where shoppers deliberately do not scan an item, mistakes where an item is unintentionally not scanned, and situations where shoppers scan only one item of several, use false barcodes or switch them, and pay for some items but leave with more.

The need to verify the age of a shopper for certain types of purchase is a legal requirement in all countries and can account for up to 22% of interventions by a shop employee. By using a consumer facing camera and requiring the shopper’s consent to use the application, an algorithm estimates the consumer’s age in real time, without storing any consumer information in the system. This enables consumers to prove their age privately in less than ten seconds, compared with an industry average of three minutes when employee verification is required.

The AI-equipped self-checkouts additionally speed up the process of purchasing produce that is priced by weight or quantity, such as fruit, vegetables or other loose non-barcoded items. Through a combination of camera-enabled computer vision and trained algorithms, the system identifies the produce and quantity so shoppers no longer have to, and retailers’ margins don’t suffer due to misidentified items.

Ohio-based Diebold Nixdorf (www.dieboldnixdorf.com) has a presence in more than 100 countries with approximately 21,000 employees worldwide.