How ScanWatch uses Artificial Intelligence to Prevent Food Theft at the Self-Checkout Counters

4/12/20232 min read

In supermarkets across Europe, inflation is leading to a surge in food theft, French daily Le Monde reports. Self-checkout machines are becoming a common target for shoplifters. In response, self-checkout makers are turning to Artificial Intelligence to reinforce self-checkout security.

In Eurozone, annual food inflation reached 15% in February, with France at 14.5%, Germany at 22%, and Baltic countries between 25%-30%.

The article interviewed retailers from France, UK, Greece, and Germany, where stores have already put up additional security measures for high-risk merchandise. The article notes that even steaks, meat, and milk products are now regarded as high-risk - with merchandise such as steaks or salmon filets put into plastic security boxes with electronic trackers.

Retailers are often thin-lipped about the issues of shoplifting, with reliable data hard to come by. Yet the French Security Professions Organization noted that food theft has increased by 10% since the beginning of 2023. Police and gendarmerie report a 14.7% uptick in shoplifting incidents. In the UK the figure has increased by 16%. In Spain - by 30.2%.


“Theft issues are becoming the main topic in the retail industry”, said Chairman of the Board at ScanWatch Evaldas Budvilaitis when speaking to Le Monde. 

ScanWatch develops an Artificial Intelligence security platform for self-checkout counters. The platform automatically recognized the items scanned at self-checkout. The product carries a double purpose. Firstly, it simplified the checkout process for the customer. It automatically recognizes bulk unpacked goods - like fruit or vegetables - without the need to select these items from the picklist menu.

“While more than half of customers carry at least one unpacked item in their shopping cart, automatic item recognition increases the speed of checkout by more than 30%”, says Evaldas Budvilaitis.

Secondly, ScanWatch aims to mitigate checkout fraud, when barcodes of scanned products are switched or the wrong item is picked from the picklist, e.g. ‘avocados turning into cucumbers’.

ScanWatch’s Budvilaitis told Le Monde how a small group in Germany exploited self-checkouts to run a ‘cabbage fraud’.

“A retailer I worked with noticed that his cash register statements showed a lot of cabbage being sold. He set up cameras to figure out what was going on. He saw that a handful of fraudsters were buying expensive products, but systematically recording them as cabbage at the self-checkout - a bulky, inexpensive, and heavy item. In this store alone, losses amounted to 1000 Eur per month”.

Such incidents are easily prevented, as the self-checkout registers equipped with ScanWatch, visually identify each scanned item and compare it to the product image. The app also monitors the checkout area and checks whether any products were (intentionally or not) left unscanned in the shopping cart.

“When an issue is detected, a notification on the self-checkout monitor is displayed to the customer. ScanWatch also automatically sends a notification to the store manager and security”, tells Budvilaitis.

Available as a SaaS (Software as a Service) addon, ScanWatch is compatible with self-checkout hardware from all major manufacturers. ScanWatch is already included in the latest self-checkout registers from Partner Tech, a leading POS and self-checkout maker.

Read the full article at: