Intelligent Video Stirs the Pot
Here is an email thread I had with “D”, a SecurityDreamer subscriber.
D > Cernium’s strengths/weaknesses are exactly opposite of what you wrote
As a stand alone system, it is very easy to deploy as it does not require integration with a platform. Perceptrak does its own alerting, display, recording, and playback as a self contained system.
ConSpin: Because it DOESN’T integrate with anyone, customization would be very cumbersome. And with the inventor/founder gone, it doesn’t look like Cernium is much committed to improving the product. The trend is toward embedded analytics and Cernium has no such capability. It is strictly server based. Cernium’s niche should be 16-48 camera applications that don’t have and don’t want to invest in a platform such as Endura or Nextiva, and only need basic DVR functions in addition to a pretty good video analytic. Cernium VP of Sales David Keene’s departure might bring about change for that organization but they have lost out on a lot of deals lately.
Me > Thanks. It’s your CON spin on Cernium that is exactly the point. It takes a patient and sometimes creative integrator to apply Cernium in most situations. That’s why I think that sites which would require loads of customization anyway would not mind the extra work for Cernium. Still disagree?
D > I guess I don’t disagree but this does ignore the unique ability of Cernium to parachute in, install a complete turnkey system, then leave in very short order. These "pots and pans" installations could be done in high volume. This is in contrast to the sexy Airports and Seaports that everyone else is after. Contrast with, say, Agent VI. When they sell a system, now they have to go out and drag in a platform to run with, then engage all three parties for integration. Verint, which is married to OV, wasn’t open to that. They liked the analytic, but liked the Nextiva platform better.
I think you are dead on regarding Object Video. They tout 1000 installations (!) and if I hear one more time about how many PhDs they have on staff, I think I will choke. My brother-in-law has a PhD and he washes dishes. Maybe if OV didn’t have so many eggheads in Reston they could spend the money on developing an affordable analytic that actually works. The old guards such as OV and Cernium should have sold last year before this most recent flood of analytics companies have come on the scene. One of the upstarts will seem like a bargain when bought for the same multiple as the big guys. I think OV has run out of letters in the alphabet with funding rounds.
Exotic analytic functions like fallen person and "salience filtering" make a difference in the presale boardroom, but 85% of end user applications need three things: differentiating cars vs. people, tripwire, object left behind. That’s all. Do that cheaply, reliably, with an easy interface to the major platforms — you can make a lot of money. Then sell before you burn too much venture capital.
Who do I like? Agent VI (formerly Aspectus VI) has an embedded analytic that works and they’ve only spent $3 Million so far. If they can land a big platform or offer their analytic FREE to all platforms — now that could be dangerous.
Me > I like Agent VI, too. Looked at them closely on a recent trip to Israel. But don’t like their distribution model.
D > What distribution model? I wasn’t aware they had one!
Me> I mean the strategy of distributing through partnerships with camera manufacturers.
D > With the exception of Cernium, every analytic I know of needs a platform to integrate with. If I were a platform company, I would whore myself out to every analytic out there. The platform is the lamp; the analytic is the light bulb. Sell the lamp and then go buy whatever light bulb the customer wants after the fact.
Thanks for giving me a chance to vent!