Just over five years ago Paul opened his third convenience store in Buckingham Park. However, unlike the two other stores owned by Paul, the level of shoplifting was much higher than he was accustomed to. What was interesting was the type of products that were missing, generally higher value with the top shoplifting choices being steak and wine. With these products being targeted it also meant that the managers of the store were better able to monitor the slippage as high value items were much easier to track as they are accounted for individually. Paul started to investigate and gained a good working relationship with the local police and it became apparent that this was not impulse theft, it seemed to be more organised.
“We were puzzled by the extent of the shoplifting. It was very targeted theft of high value items and it seemed that perhaps it was organised by someone stealing to order. Just by luck, my wife was talking to a friend she had met, who turned out to be the sales director of Facewatch, and in a conversation she mentioned the problem. Within a few weeks we decided to become the first trial site for the new facial recognition system.”
In mid-2018 the Facewatch team installed a new facial recognition camera inside the main entrance to the Jubilee Square store. The system automatically scans the faces of customers entering the store against a list of known people who had been previously been caught on the existing CCTV cameras shoplifting or abusing the staff, however, this was just the start.
“It became clear very quickly that the power of the system could only be used if the data (the watch list) in the system was good,” said Paul. “We found that our existing CCTV camera were pretty good at capturing shoplifters but the quality of some of the facial images meant that when they were transferred to the Facewatch system it was more difficult to get an immediate match. We discussed this challenge with the Facewatch team and a more powerful solution was found. The newly installed facial recognition camera was also used to capture CCTV footage of incoming customers and as it was being used in a dual mode for CCTV and facial recognition, we were lawfully allowed to look at these images for 30 days before destroying them. We now have a system that rarely gives us a false match as every image of a thief is almost a perfect headshot.”
Everyday use of the system is now a seamless part of the store manager’s job, the Facewatch smartphone is carried by the manager and when an alert is heard the person is confirmed on the screen and then visually identified. If they are seen stealing or behaving badly, they would be watched by staff and if found to be hiding stolen goods the Police are called.
Since installing Facewatch the shop has seen a reduction in losses of over 25%.
The hardware to run Facewatch includes a standard HD camera and Intel NUC, a mini-PC that is 4×4 inches in size which plays and records video at 4K Ultra HD clarity.