Tag: crowdsourcing

London Riots and Video Surveillance, Pt.2

My last post was about the lack of any apparent deterrence of rioting from CCTV. However that’s not to say that video surveillance is proving of no use to the authorities. However the way it is being used says a lot about both the limits of CCTV and the general problem of analysis of video images.

As part of ‘Operation Withern’, the investigation into the rioting, the Metropolitan Police have set up a special section of their website, London Disorder Images, as well as on Flickr, which is essentially crowdsourcing the identification of suspects. Despite being the most well-resourced police force in the UK, the Met lacks the resources, time and expertise to analyse and identify everyone it wishes to identify itself, and with widespread popular anger about the riots, they are banking on opening up the process of surveillance and identification as being more efficient and effective – and they may well be right.

Of course, with the problems of lighting, angle, distances, and image quality, the images vary in identifiability – and bear in mind that the few posted so far are probably amongst the best ones – and no doubt there will be many misidentifications. And, in addition, hundreds of people are already being processed through magistrates courts without much need to video evidence. But it is a tactic we are seeing more and more in many places (e.g. Toronto, following the G20 disturbances).

Problems with Crowdsourcing Surveillance

Bruce Schneier has a nice little piece which is saying similar things to what I’ve been saying over the last couple of years on the subject of ‘crowdsourcing’ or opening closed-circuits of surveillance. He critiques the Internet Eyes scheme and Texas Border Watch and others. This is also the subject of the paper, ‘Opening Surveillance?’ that Aaron Martin of LSE and I presented at the S&S conference in London in April, and which will hopefully be coming out in the journal’s conference special early next year…

US military crowdsourcing communications

Marketing site, Brandchannel, reports on a US Army program, the Joint Tactical Radio System (JTRS or ‘Jitters’), which they say is going to crowdsource video surveillance on the battlefield. Actually, if you watch the embedded video piece from the US Army itself, you’ll see that the program is much more fundamental than this, it is about integrating different radio systems and trying to make the best use of scarce EM bandwidth in order to allow all kinds of more efficient communications – which would of course include video surveillance data or any other kind of data sent over wireless.

However, all is not what Brandchannel thinks. According to Global Security, the JTRS program was already in trouble back in 2005 and rumours of its demise continued to circulate – Wired’s Danger Room reported on this back in 2007. It is still in existence but has been scaled back, the contractors have switched and the costs have risen to more than $1Bn.

The latest bit of boosterism, and claims from the JTRS people that the system will include the ability for troops to access surveillance images from military UAVs and could be in place by 2014, comes therefore in this context, and also in the context of the hacking of US military surveillance drones by insurgents using cheap Russian TV downloading software. One of the really interesting things about this is how the context of military expertise is changing: one of the key justifications for all this is the concept that US troops will already be familiar with handheld devices and streaming video etc. Network-centric warfare turns out to be no different from kids using their iPhones  to watch movies… if, of course, it ever actually works.