The Surveillance Revolution and the Media Revolution

Study the history of IT and you see multiple evolutions and integrations. But perhaps the most powerful and the most fundamental is in progress right now. It is the gradual taming of communications and media. Chaos is abroad. The publishing industry is in metamorphosis with its revenues migrating either to Amazon or to Google. The music industry no longer knows what it is or how to conduct itself. Television will soon be engulfed by the same set of challenges, and in all probability, it will struggle to meet the challenge.

What does that have to do with surveillance?

Although the surveillance industry is huge (estimated to worth worth $150 billion annually, far bigger than Hollywood, for example) and continues to grow, it’s not where I’d go looking to spot important trends. However, it looks like there’s one in progress there. I was briefed by a company called VidSys recently, which is causing some consternation to its competition in the video surveillance industry.

I’ve always regarded the video surveillance industry as a bit of an IT backwater - selling closed systems to non-IT people so they can hire poorly-paid staff to sit in-front of banks of boring screens for hours on end.

I’ve talked to people who’ve done that job. It’s less exciting than reading a chinese telephone directory. Mostly such employees spend their time listening to iPods or surfing the web, rather than watching the bank of screens that flit from one unchanging scene to another - that is, unless they’re in a Las Vegas casino trying to spot grifters. Bored people don’t deliver good security.

What to do?

What VidSys and some other companies in the surveillance market are doing is using rules engines and video algorithms of various kinds to alert security staff to possible incidents and to help handle them. One situation, for example, that VidSys deals with is tracking an intruder. The problem is that you may have tens or even hundreds of cameras located across a large geographical area or several large buildings. Once you’ve detected an intruder, you need to switch from one camera to another to keep an eye on where he is and how he’s moving, and that’s something that the VidSys software can help with. It simply presents all the cameras that could be the next one the intruder appears on and the security man simply indicates which one it is. Normally, the outcome is that the intruder is apprehended with ease.

VidSys is, by no means the only player that can provide functionality of this kind - and anyway, it’s not the point of this article. What is interesting to me about VidSys, aside from such functionality, is that it provides a genuine software platform for managing all the surveillance assets that a company has. The diagram (above) depicts the “platform” and almost explains itself.

From my perspective, it is similar to quite a few software architectures I’ve seen. It has a transport layer (from multiple information sources/devices). It has a translation layer, for passing data from one format to another. It has an intelligence engines layer which allows multiple software components (algorithms) to plug in and, of course, it has a presentation layer.

VidSys will probably do well in the surveillance market with this, because it avoids “rip and replace” and even works with analog video sources. It’s easy to adopt. It may even transform the market, forcing competitors into offering something similar. However, what also interests me more about it, is that it provides a hint to a new market of the future where you process video for reasons other than surveillance - like entertainment or studying crowd behavior or learning how to play golf better or whatever.

A platform like this makes it possible. It’s the shape of things to come.