As I reported shortly after the attack, video captured at the scene was instrumental in identifying the two brothers who allegedly did it. On April 18, the FBI posted a 30-second video clip and still images taken from video of the suspects. Within hours, one of the men, Tamerlan Tsarnaev, died in a shootout with police, and the other, Dzhokhar Tsarnaev, was captured shortly thereafter.
The FBI has been criticized for failing to share critical intelligence on the threat posed by Tamerlan Tsarnaev in particular, but no one is questioning the effectiveness of the agency's use of video. And since my early report was posted, I've heard from several experts about some of the latest technology developments in video surveillance.
Video surveillance is among the myriad technologies being driven by the consumer market. Amazon, Costco, Home Depot and other retailers sell full-blown systems that combine cameras, networking and monitoring capabilities. Earlier this year, I was checking in at Tampa International Airport for a flight to New York, where it was snowing. The airline check-in agent, who was from New York, pulled out a smartphone to show me, via a live video feed, the snow piling up in his driveway.
Ivideon, a startup in the market, combines conventional video surveillance, from closed-circuit TVs, webcams and IP cameras, with Web capabilities such as online archiving, the ability to plot cameras using Google Maps, and feeding video to websites and even social media.
When thousands of cameras run 24/7 in cities like Boston, New York and Washington, D.C., video surveillance quickly becomes a big data challenge. Analytics and automation technologies are the only answer. The FBI is developing facial recognition capabilities as part of its $1 billion Next Generation Identification program. And IBM sells video correlation and analysis software that provides facial recognition, real-time alerts and situational awareness.
Behavioral Recognition Systems (BSR) is developing software that goes beyond the ability to respond merely to preprogrammed objects by learning about an environment, creating "memories" and providing real-time notifications when the software detects something out of the ordinary. On May 14, the vendor announced that it had completed interoperability testing of its flagship product, AISight (pronounced "eye sight"), with Cisco's Video Surveillance Manager, a sign that such advanced capabilities are moving into the mainstream.
In one example of how this technology will be applied, Portland, Ore.'s public transit system, called TriMet, will use BSR's software to monitor bridges and overpasses. The system will learn the difference between a bus or light rail train and, say, a pickup truck that shouldn't be in the area.