Report: MBTA Deploying "Most Expensive Bus Surveillance System Ever"
Despite a bus fleet that is overdue for replacement, the MBTA is installing what it says is the most extensive surveillance system on a major transit system in the country. As reported in today's Globe, a total of 225 buses will be equipped with three cameras each, capable of providing live, high definition video feeds to the MBTA's control center as well as to MBTA police cruisers. While many buses already have surveillance cameras focused on fare boxes, the MBTA's Senior Director of Security and Emergency Management, Randy Clarke, complains to the Globe that, because the video is stored on the bus, investigators must go to the crime scene, the bus, to retrieve the video. Now, he says, "we can see it instantly." Not satisfied with live, instant high definition video, Clarke also told the Globe he seeks to recruit the entire mass transit public, saying, "What we’re trying to do is make everyone on the bus work as an extra set of eyes to help report suspicious behavior or criminal activity." But as Transit Police Superintendent Joseph O'Connor notes, "crime is relatively low on buses."
According to the ACLU of Massachusetts, responding to a proposed Cambridge surveillance camera policy:
In fact, most studies show that cameras have no meaningful impact on criminal activity whatsoever. Other studies show that cameras have a displacement effect: they simply drive crime into areas where there are no cameras. And at least one study, from Chicago , showed that crime actually increased when surveillance cameras were installed on public transit platforms.
The surveillance system, costing almost $7,000,000, is funded by a grant from the Department of Homeland Security (DHS). According to IPVM, an organization that provides market research for surveillance camera systems, the cost of $21,000 per bus makes this the "most expensive bus surveillance ever", Bus surveillance systems usually cost a few thousand dollars per bus, IPVM says. DHS is expected to provide .8% of the MBTA's capital funding for the next five years amounting to $51,000,000 according to the MBTA's draft capital spending plan (pdf). DHS funding has been so extensive that the MBTA hired a budget analyst dedicated only to defining security priorities. That funding, estimated to be $150,000,000 since 2004, has funded projects that include transitioning from 700 analog cameras to were to a 6,000-7,000 camera digital system. Since it is impossible for humans to monitor so many cameras round-the-clock, the MBTA has started exploring pilot projects with BRS Labs, a "behavioral analytics" company. BRS Labs "AISight" is said by the company to be:
advanced, intelligent software that uses Behavioral Recognition™ technology to learn—on its own—about the environment and objects it observes in each camera’s field of view.
According Mass Transit Magazine, many local transit authorities are adopting AISight, among them the The San Francisco Municipal Transportation Agency (SFMTA). According to BRS Labs' Vice President David Gerulski "AISight's behavioral recognition technology gives SFMTA the ability to monitor dramatically larger numbers of cameras in real time." AISight is also used to monitor the World Trade Center, ground-zero for the 9/11 attacks. According to the company, gross sales doubled from 2012 to 2013 and it expects even greater growth in 2014. BRS Labs also says that it has a commitment to privacy:
BRS Labs is concerned about the privacy rights of individuals everywhere. AISight is implemented as a software-enabled enhancement to video surveillance system hardware. It is the responsibility of the buyer to assure that the placement of all cameras and the content of all video streams comply with applicable local ordinances, state laws and federal regulations. The artificial intelligence inside AISight makes no attempt to identify specific persons or objects. The system does not recognize individual faces or read vehicle license plates. Instead, it teaches itself patterns of behavior by generalizing from what it observes in the video stream. The creators of AISight consciously and deliberately designed the system to protect and respect individual privacy.
The artificial intelligence techniques used by AISight are one of a series of technologies that will, according to Visual Computation researchers at Queen Mary University of London,( pdf) transform video surveillance systems. Intent profiling using microexpressions and other behavioral cues is increasingly possible. Crowded scene analysis, including the ability to track the movements of individuals, is also advancing. Made possible by increasing quality of cameras and the machine learning and data processing techniques of "big data", this work is being fueled by the need to create products that process the data being created by government security mandates.
While the MBTA is increasing their own visibility into what transpires on buses, it's not helping one find a bus in busy, complex bus stops like Central Square. Their multi-million dollar, multi-year "wayfinding" project considers those outside the scope of their work.
Disclaimer: The author is a member of Cambridge's Public Transportation Advisory Committee. The views expressed are his own.
Editor's Note: This article is an Open Media Boston guest editorial.