Times Square bombing — where were the cameras?

Although it’s early still, this bombing doesn’t look especially well-organized, and I expect that we’ll catch the perp soon, given the massive amount of information to be gleaned from inside the car.  That said, I can’t help wondering why New York City, the best-funded and most terror-focused police department in the country, doesn’t already have video of the would-be terrorist getting out of the car.

New York has been spending hundreds of millions of dollars on street cameras.  The lower Manhattan camera project was expected to cost $90 million and to  network 3,000 cameras.  That’s $30,000 per camera.  The project is being expanded to midtown at a similar cost.

Despite all this funding, though, we don’t have pictures of the wannabe bomber.  That’s probably because NYC hasn’t finished installing the midtown system.

But there’s another problem; 3,000 cameras aren’t really enough to take pictures all across lower Manhattan, if what you want is a record of everything that happens for later investigation.  Inevitably, there are lots of blind spots in the system, at the same time that it costs an arm and a leg for each camera, which creates an incentive not to put cameras in low-risk areas.

I think New York has made a mistake in taking this tack.  It is trying to use cameras for real-time detection of terrorists.  That massively raises the cost.  The cameras have to be networked back to a central office, they probably have to have real-time pan and zoom capability.   And even then, it’s hard to see how cameras would help prevent attacks in a scenario like Times Square.  The SUV wouldn’t look any more suspicious to police watching on video than to observers on the scene.

So, in the real world, the best use of cameras is likely to be after the fact — to identify the perp once the crime has been committed or attempted.  If so, we don’t need to waste time on networks or pan and zoom capability.

All we need is a bunch of cheap standalone cameras that provide blanket coverage and keep taking pictures, overwriting their memory every week or two.  That would cost maybe $200 per camera.  For the cost of a single pan and zoom camera, you could put a couple of dozen standalone cameras on every block.  And privacy would be better protected because retrieving the video would be expensive and overt, making it unlikely that the police would do so unless a crime had been committed.

One of my projects at DHS was to move this idea forward, so we challenged industry to design a cheap (under $200), standalone camera for use on mass transit vehicles that could withstand even a massive terrorist blast without losing data.

After I left, DHS actually tested several cheap, explosion-resistant cameras for use on mass transit.  And since no VC blog post is complete until something explodes, here’s a video of the test.  Most of the cameras and data did indeed survive even this ball-bearing-enhanced explosion.

I leave it to the libertarian commenters whether they prefer a big, expensive government camera system (on the grounds that it will protect privacy by failing a lot) over a cheap system that only has value after an attack (but is more likely to provide blanket coverage).