A year ago, Veesion and its team won a prize at the 22nd i-Lab innovation competition: a competition that aims to detect innovative technology start-up projects and support the best of them with financial aid and appropriate coaching. In the meantime, they have expanded internationally and even want to develop a new technology for detecting shopping carts of convenience, i.e. a shopping cart in which not all the items have been counted at the checkout.
A real problem to address
The project was born out of a simple observation: each year, shoplifting represents 5.8 billions euros, even though companies invest 7.4 billions euros in their security. Despite this, less than 5% of thefts are detected.
Benoît Koenig, Damien Menigaux and Thibault David, all three alumni or students of the Ecole polytechnique at the time, then had the idea to create an artificial intelligence algorithm that would detect suspicious behaviour. In April 2018, they launched Veesion: a technology for recognising gestures in video content to detect, automatically and in real time, acts of theft in a point of sale.
Their solution adapts to the equipment already present in the point of sale and connects to the cameras.The gesture recognition algorithm developed by Veesion continuously analyses the video stream and is able to detect in real time gestures associated with theft. A notification is then sent to staff, who can take action to apprehend the suspect. The technology can also be used to analyse theft histories and provide personalised recommendations, such as risk areas and time slots, or security breaches.
A proven track record
The next steps? An even greater international development; the Australian and English markets having already been conquered recently. They also want to develop a new system for detecting shopping carts of convenience. Veesion is also expanding its target market, as the technology is now present in supermarkets but also in pharmacies.