Edeka Jäger uses the AI-based solution from Diebold Nixdorf for age verification at the self-checkouts in its 24/7 store at Stuttgart Airport. When a customer scans an item that is not allowed to be sold to minors, the technology provider’s software dubbed Vynamic Smart Vision / Age Verification automatically determines the person’s age. In 80 to 90 per cent of cases, staff intervention should no longer be necessary.
Edeka retailer Florian Jäger explained to The Retail Optimiser that automated age verification is not used in his supermarket when in staff-less operation during night hours between 10 pm and 6 am. Light alcoholic drinks, such as beer or wine, are then located in a locked area that is only accessible to people over the age of 18 after scanning their ID card. Edeka Jäger does not sell spirits and tobacco products at night. These products are stored in cabinets that can be closed with roller shutters. The Retail Optimiser reported.
”We have five self-service checkouts in use during the day where our customers can make cashless payments,” explains Florian Jäger. These have integrated cameras installed. If an age check is necessary, it takes a picture of the shopper. Algorithms use artificial intelligence to analyse the customer’s facial features and estimate their age within a few seconds. The photo is then deleted for data protection reasons.
Algorithms determine age with an accuracy of two years
Florian Jäger confirms to The Retail Optimiser that, according to his experience, the system shows that the results are accurate within plus or minus two years: ”However, we have decided to only automatically authorise the purchase if the age determined is five years above the legally required minimum” Retailers can set this safety buffer themselves. Florian Jäger explains that there are no legal concerns regarding the use of the system. ”However, the retailer is ultimately liable,” he adds.
If the system blocks a purchase, it immediately sends a signal to the store employees. At Edeka Jäger, a light on the SCO then flashes. ”Staff at the attended POS systems have a clear view on the self-service checkout area,” explains Florian Jäger. The employees can then react immediately and authorise the transaction using their mobile device. ”We want to disrupt the process as little as possible,” he explains his approach. The branch office also receives a message at the same time.
More AI in the pipeline
Florian Jäger spoke to The Retail Optimiser about further plans. Following the successful deployment at the airport location, he plans to install self-checkouts with age verification in his other supermarkets too. Although the use of the terminals has increased inventory discrepancies, the savings and benefits of the technology are greater than the losses, he explains. Location and layout of the store also play an important role in the risk of theft and fraud.
”I’m also planning to use another AI-based solution to help identify manipulation during self-checkout,” explains Florian Jäger. He is also thinking about using the existing in store camera system to better identify potential theft using artificial intelligence.
