Burden of Proof Falling on Video Analytic ManufacturersBy John Honovich, Published on Oct 16, 2009
A public rejection of video analytics by an ADT systems designer is a good example of the decline in sentiment towards video analytics over the last few years. It's a clear example of the challenge that faces video analytic manufacturers.
In a discussion post on LinkedIn [link no longer available], the designer asked for a recommendation on:
"A system comprising 750+ IP cameras of which 10% will be conducting video analytics. This needs to include detection of the following - slips, trips and falls, loitering, target tracking, running, left objects, abandoned vehicles, crowd management/leevel of congestion and theft detection.
After a series of manufacturers pitching their products and a few industry people questioning the maturity of video analytics, the ADT designer decides against analytics, responding:
"Following a meeting the general consensus is that just concentrating on getting a fully operational system with 750 cameras viewing at 25fps is challenge enough, especially along with the other security disciplines that require integration. Video analytics to that level are are a nicety that can come later. I've spoken to a lot of very experienced people and it's not there yet. No doubt it will be one day."
The response is reasoned, measured and common even among larger systems.
Burden of Proof
While most video analytic manufacturers have been quiet over the last year, the video analytics market is now more in need of public proof and validation than ever. BRS Labs, the most vocal of the video analytic manufacturers, is at least controversial and may be detracting from the overall cause.
A number of manufacturers are tired of criticisms by this site and others. However, they offer little public or independent proof of their success and practical value. In the face of increasing disappointment and without new evidence, it is hard to build confidence in these technologies.