Camsafe AI-powered solar security cameras
Last year, an old friend of Silverink's had a problem ...
This friend runs a successful London-based security firm, with hundreds of guards stationed across the UK and a fast-growing army of security cameras. But the problem was the cameras: there was only one company on the far side of the world that could deliver exactly what our friend needed.
These cameras did a lot of things right. They were solar-powered and used mobile networks, so they could operate without power in remote locations. But the hardware was unreliable, the software poor and the service patchy. Oh, and they cost a bomb.
Naturally we said - "we could do better!"
So began a lengthy process specifying, locating and testing many, many cameras. Mostly from the Far East. We eventually found a Chinese supplier willing to partner with us on the development of security features we considered essential - and ultimately supplied a superior product for 1/3 the price.
Alongside this hardware R&D we built a fast, scalable and user-friendly cloud-based service that allows staff to monitor, process, classify and securely distribute these sensitive images. In record time.
Eighteen months later and Camsafe has 180 cameras across the UK and over 2 million successfully processed images. Our cameras operate in remote locations without power supply and will run for months without maintenance. If they detect movement, they send images to our platform within seconds for human/machine analysis and potential escalation.
This success has brought its own issues, with the sheer volume of images becoming a challenge.
So we added Artificial Intelligence to analyse and classify images as they come in.
This doesn't remove humans from the process altogether. But if a machine can reliably filter 95% of images that do not contain either humans or vehicles, that means us humans can better focus on the images that do require detailed analysis. And that is a spectacular efficiency gain.
Upshot: This has been a pretty intense project. Think twice before you say you can do something better. But most important of all, our friend no longer has a camera problem.