The ‘All Seeing Eye’!

In my article in Issue 4 2007 of Intelligent Transport, I concentrated on technological developments in CCTV, and the relationship to transport systems. In this article, I want to ‘zoom out’ and review the wider picture, considering the role of CCTV in society, at the same time as focussing on the detailed implications for transport systems.

CCTV is ubiquitous, but surprisingly it still retains fantastic support by the vast majority of the public, despite all the talk by some organisations and high-profile individuals that Britain maybe sleepwalking into a surveillance society.

In contrast, it has been generally seen as a ‘caring father’ rather than ‘Big Brother’. To my knowledge, there has only been one finding of the European Court of Human Rights relating to CCTV, since the Human Rights Act was introduced, and whilst the court found in favour of the complainant, the incident dealt with a specific event rather than any of the major principles.

Equally, totally fallacious but widely quoted ‘statistics’ such as the 4.5 million cameras in the UK and being watched on over 30 systems each day, have failed to dent the public enthusiasm for CCTV, and still the majority of new systems or extensions to systems installed, are being initiated by requests from the public, rather than being foisted upon them.

An interesting analogy must be the introduction of ‘speed cameras’ which, when first introduced, did receive some public support as a traffic safety measure, but when the public got the impression it was more about ‘revenue generation’, they were renamed on warning signs as ‘ safety cameras’ in an attempt to retain public support. Since then, the criteria has been changed by the Government to the degree that a large percentage of the cameras were considered to no longer comply with the requirements (i.e. they had little safety value) and have been removed. This is the last thing we would want in respect of Public Space Surveillance.

And thus it is with CCTV. Public Space CCTV, and those systems on ‘Planes and Boats and Trains’ or any transport system, will only retain public support whilst they are seen as enhancing ‘community safety’ and the use is proportional to the risks and the benefit achieved. It is simply not a miraculous panacea for all problems.

At this point, I think it is important to differentiate between the two types of CCTV systems: those with a pro-actively monitored control room such as are frequently found at transport hubs and terminals, using multi-functional cameras capable of the ‘Pan, Tilt and Zoom’ (PTZ) functions; and those systems frequently found in buses, trains, ships, taxis and in fact all means of transport, using fixed view cameras (or a PTZ facility on automatic tours of preset positions) largely not pro-actively monitored (although the driver or guard may well have a small monitor so he can respond to a disturbance). Images are recorded but reviewed only if an incident is later reported.

In my view, the greatest risk to public support lies in the pro-actively monitored situations, and largely results from ‘over-enthusiastic’ human decisions rather than the technicalities of the system. Some years ago I visited a control room monitoring bus lanes. The operator had control of many cameras focused on the bus lanes and whenever another vehicle encroached, the number and image was recorded and an enforcement notice generated to collect the fine. It didn’t seem to matter to the operator what caused the driver to encroach (for example a car blocking the other lane or dropping off a passenger), it was like ‘shooting fish in a barrel’, with the operator counting £80, £160, £240 etc as the number of instances increased, his only interest was income generation.

Use of cameras for traffic enforcement seems to be following the same route as ‘speed cameras’ and risks loosing public support. Only recently the Local Government Association warned Councils not to use CCTV to catch people dropping small amounts of litter in the town, street parking enforcement or people riding bikes on the pavement. That is not why such ‘crime’ cameras were installed, Community Safety is one aspect, and as we found with ‘speed cameras’, income generation is entirely separate. The first has the support of the public and the second is considered totally ‘over the top’.

An alternative approach to these scenarios is available with what has been dubbed ‘Talking CCTV’. It is not a new idea but recently has become increasingly popular, and is very good at deterring minor incidents without the ‘Big Brother’ implications. No, the camera doesn’t talk, but a loudspeaker connected to the control room is located adjacent to the camera (in some cases utilising the same transmission path) enabling an operator who sees a minor infringement to provide a reminder that for example littering is an offence. It is also useful in more serious offences to warn those involved in an assault that the Police are on their way and all events have been recorded. Equally, it does reinforce the essential element to enhance that ‘feel safe factor’ related to support for CCTV, that someone is actually watching the cameras. Could this have value in the transport situation where the driver might see something but be physically unable to respond immediately? It is however generally accepted that it is totally inappropriate to use the technology in the opposite sense, of recording conversations by members of the public, although the often found ‘Help Points’ where a traveller can contact the control room are a massive benefit in retaining that ‘feel safe factor’ and bringing incidents to the awareness of the operators.

In contrast, the fixed, largely un-monitored transport CCTV systems are less likely to infringe privacy issues and risk diminishing public support, as they are only viewed post event, and then to identify what occurred and who was responsible.

This does however raise an issue which many people have expressed to me; the public just perceive CCTV as one overall technology and give little thought as to exactly how it is being operated. They see signs saying ‘CCTV is in Operation’ and might even see and recognise a camera. Their immediate reaction is that they are being ‘protected’. In the case above, a woman was seriously assaulted in line of sight of a camera. She felt certain that the police would be rushing to her aid any minute. They didn’t, it was a ‘dummy’ camera. The ‘camera’ owners lost the case and I understand paid many tens of thousands in damages, but this hardly compensated the victim for loosing her faith in the ability of CCTV to protect her.

That was a dummy camera, but my point is; could a similar situation arise on a bus or a train, where notices say CCTV is in Operation, but in fact no one might be watching the images ‘in real time’? Certainly it would have the saving grace, compared to a dummy camera, that it would hopefully record the exact details of the event, with good enough image quality to provide the investigators with ‘identification quality’ images. But I do say that more in hope than in anticipation!

So, what in fact is CCTV? Ask the public and they will say ‘the cameras on the street’, or ‘those poor quality, grainy images we always seem to see on programmes like Crimewatch’. And, in the main they are right, not principally the camera itself, but the end result – the value of the images captured, and in many ways the camera itself is just a small, but an important part of a very complex series of relationships between the many facets involved. No camera has ever jumped off its mounting to arrest a villain or provide a record of what transpired without the other human and technical interfaces. This is as true for all transport systems as much as any CCTV system.

The key to any effective CCTV system is to understand fully all these relationships from the outset, to have total clarity in exactly what you are trying to achieve and how all these facets might interact to best achieve your objectives. Usually, this means obtaining evidential, identification quality images of everyone present (equivalent to head and shoulders shots), and full coverage of any incident wherever it occurs – on a bus, in a train, or indeed any transport location.

Few, if any, single un-controlled cameras are capable of achieving this, as a camera targeted on the entrances will need to be focused in to get head and shoulders, whilst even if in a train carriage or bus it will need a wider angle view to see the whole scene, and if looking longitudinally along a railway carriage will not have the depth of vision to obtain good quality images from one end to the other, particularly if the carriage is fairly full and you might be interested in the actions of a persistent pick-pocket. The systems now operating on a large proportion of the London bus fleet recognise this and have up to 16 cameras on each bus, eight with internal views and eight looking outward to the street scene. This latter idea recognises the fact that driving around London every day they are bound to record evidence of some activity in the street scene of interest to the Police, and so the systems were designed from the outset to play a larger role in community safety, than solely the safety of passengers. Eight internal cameras might at first seem excessive but just recall the quality of images of those involved in the bus bombing and the subsequent failed attempt.

But in this context, things might change with 360 degree vision, mega-pixel cameras with exceptionally high resolution, capable of considerable enlargement, and the birth of ‘intelligent moving cameras’ where an array of fixed view cameras can direct a PTZ to zoom in on any subject or action which falls within the criteria programmed into the software. In my view, both systems are still in their infancy and comparatively expense, but may develop in the future.

So camera types, and camera locations have been selected to cover the area and provide good quality evidential images. Practically, digital recording is the only option for moving transport solutions, but there are so many choices between the systems available and a deep knowledge of compression systems, image resolution, frame rate and the storage media is essential to make the most appropriate choice amongst them. For a non-technical person like me, I want to see the end result, rather than understand all the technical jargon. Get a recording from the equipment of a frame with a wide angle view containing people or a car number plate at a reasonable distance and try to enlarge and enhance both it, and the following frame. Can you identify a person or read the number plate or does it just pixelate, and how much movement is there between frames? If it were a fight, a knife attack, or a gun incident, could the critical part of the incident be lost between two frames?

A critical issue is how long you retain the images? On buses or trains, it would be impractical to keep days or weeks of images on the actual transport vehicle and no one else would be able to access them when needed, so it is far more practical to download to a main system when back at the depot. But how long should they be kept there? In the past, I might have said probably the transport company would be aware relatively soon after an incident occurred as someone would have complained to them so keeping the images any length of time would probably not be necessary. But terrorism and other major incidents have brought new dimensions to this, particularly in relation to transport hubs and modes of transport, where an event may have occurred but the investigators might want to see any ‘practice runs’ or previous ‘reconnaissance’ in the weeks before the event to identify others who might have been involved. This is an issue I can’t advise and suggest you discuss it with the Transport Police.

Now you have masses of data on your storage medium, from hundreds or thousands of cameras, with probably at least 30 days storage. The next question is how you deal with downloads? A minor, localised incident should cause you no problems, downloading to disk or some other media, but in the event of a major incident affecting a large city, the Police will probably want every image you have. How easily can they download them without upsetting all your other daily routines? That is, you will still want to be recording new images whilst the older ones are being downloaded. Can your system manage that within an acceptable period? Check with your supplier and the Police that both can handle that.

In the moving transport environment, CCTV technology and particularly the transmission systems between camera and recording media face problems not found in static installations, and Dave Gorshkov who chairs the technical working group of the American Public Transport Association recently told the delegates at our last conference in April, how his organisation has been working to improve the way that CCTV is used across US public transport networks. The basis was when the Department of Homeland Security asked APTA to look at various aspects of CCTV with a view to looking at the video analytics that needs to go into some of their installations. But he commented that ‘in order for video analytics to be effective, we need to ensure we have an appropriate minimum baseline for our CCTV systems. So the first task we undertook was to develop a technical standard that our members could use and move forward with, not simply a recommendation that we need cameras’.

His working group produced a detailed 55 page report entitled ‘Technical Recommended Practice for the selection of cameras, digital recording systems, digital high speed train-lines and networks for use in transit related CCTV systems.’ It is an exceptional document and more details can be obtained by visiting www.cctvimage.com May 2008 issue, Conference presentations – page 34 www.apta.com and www.aptastanards.com.

It is, in fact, essential reading for all those involved with CCTV in transport, and whilst many of the technical issues are far above my ability, this and other documents relating to CCTV in transport provide useful advice.