Israeli provider of scanless store technologies Trigo is launching a solution that promises to detect possible cases of shoplifting using computer vision and AI. It uses CCTV cameras to anonymously track customers’ movements and identify items they pick up from the shelves. The software then compares these with the goods scanned and generates an alarm in the event of discrepancies.
The solution is designed to work with a wide variety of checkout options, including traditional manned checkouts, self-checkouts, self-scanning and self-scanning with mobile devices. It provides security personnel with real-time alerts while the suspects are still on site. Customer privacy would always be protected, and biometric data would not be used or stored.
The solution can be deployed on existing CCTV infrastructure. Therefore, no significant investment in new hardware should be necessary. The provider states that only a connection to an existing network video recorder (NVR) needs to be established and integration with the POS system is required.
Scanless stores are still in trial stage
Trigo software enables operation of scanless stores and can also be used for inventory monitoring and shelf layout control. Both Rewe Group and Tesco have invested in the provider’s technology. Auchan and Edeka are also testing the solution. Aldi Nord has since abandoned its pilot project in Utrecht, the Netherlands.
While retail companies are now rolling out an increasing number of unmanned mini supermarkets using different solutions, attempts to operate larger store formats without scanning have not yet progressed beyond pilot stage. Apparently, AI and computer vision alone are currently not capable of clearly identifying every single item. In the spring of last year, media reports stated that Amazon Go had over a thousand people in India working at monitors to help in cases of doubt. The Retail Optimiser reported.
Theft in retail is on the rise
Trigo clearly expects to successfully market its software as a problem solver facing the growing threat to retail margins posed by inventory losses. In fact, retail theft has increased significantly in recent years. For 2024, EHI Retail Institute calculated a 23 percent increase in reported shoplifting, which corresponds to 15 percent in terms of value. In the United States, shoplifting figures rose by 93 percent within four years, as industry data and FBI statistics show.
Many retail companies will welcome a solution that automatically analyses CCTV camera images without tying up staff and reliably flags attempted thefts. However, doubts remain as to the extent to which the software can better existing solutions at reliably detecting all or almost all cases. Even a comprehensive camera system will not be able to accurately capture each customer movement in all situations. If the Trigo software can clearly recognise and identify picked up items, the question arises if scanning was still necessary at all.
