Counter-Terror – Fact and Fiction

Duncan Graham-Rowe

July 2005

The continuing bomb threats in London have left most of us in no doubt that we are now living in the shadow of terrorism. By insisting that terrorists will not be allowed to affect our way of life and by using expressions like 'war on terror', politicians imply that we can somehow do something about this threat. But what exactly? TV series like 24 and Spooks may offer a glimpse into the inner workings of the counter-terrorism techniques used by the intelligence community, but how realistic are they? Can technology really help protect us?

Smart CCTV

Technology certainly has the potential to help protect us. Even before the collapse of the Twin Towers it had been widely recognised that certain public places, such as subway systems and airports, were prime targets for terrorist attacks and posed a particular challenge for protection teams. Consequently, technologies aimed at shielding such vulnerable hotspots from a range of threats have been under development for some years.

Much progress has been made in developing smart CCTV analysis software that can monitor camera feeds for 'unusual behaviour' that is statistically out of the ordinary. It can pick up anything from someone preparing to throw themselves in front of a train, to a fight breaking out, to someone leaving an item of luggage unattended at an airport.

'Smart CCTV analysis software can monitor camera feeds for unusual behaviour.'

One example of this technology is the Video Motion Anomaly Detection (VMAD) system developed by Roke Manor Research in Hampshire. Another system has been co-developed at Kingston University with technical partner Ipsotek and with research funded by the European Commission.

This system has in fact already been tested at Liverpool Street and Mile End Underground stations in London and is currently being trialed in Rome's Termini Underground station. But its emphasis is on monitoring crowd flow, and although it is able to spot intrusions into prohibited areas, unusual lingering and abandoned objects, it has a limited abilitiy to spot suicide bombers.

With 1400 CCTV cameras on the London Underground network, this smart CCTV analysis software addresses the problem of having many cameras and not enough pairs of eyes to study the footage. But despite its potential for preventing suicides, fights and other petty crimes, it has very limited value in preventing a terrorist bomb attack of the sort we have recently witnessed. In this case, it is likely to prove useful only after the event. Since suicide bombers attack by blending in with the crowd, intelligent CCTV that spots unusual behaviour is unlikely to aid prevention.

Scanners

Another approach is to create a cordon to try and detect explosives as they are brought in to stations and airports. Ultra-sensitive chemical sniffing sensors, for example, can identify molecules based on how fast they move in an applied electric field – a technique developed by Bonner Denton, a chemist at the University of Arizona.

Denton is currently adapting the technique to suit a handheld scanner. Although this might be useful for spot checks, it would entail monitoring every commuter at rush hour, bringing transport systems to a standstill. The result would be bottlenecks that in themselves are ideal targets for terrorists.

Far better would be to somehow screen travellers automatically as they enter a building or station. Within days of the London bombings, reports emerged that some stations were set to have a new type of device installed which would do precisely this, detecting anyone carrying weapons or explosives.

'Terahertz scanners can detect plastics and other non-metallic materials.'

These somewhat appropriately named terahertz scanners are so-called because of the range of frequencies they are designed to detect. Terahertz radiation sits between microwaves and infrared on the electromagnetic spectrum. It is invaluable as a scanning tool because it can pass through clothing but not through skin – ideal if you are trying to catch people with concealed weapons or explosives.

There has been much debate about the use of these scanners because of sensitivities over their ability to reveal people as naked, leaving the technology open to abuse. But now the merits of the scanners appear to outweigh their drawbacks, making them very attractive.

Unlike the metal detectors normally found at airports, terahertz scanners (sometimes called millimetre wave scanners) can also detect plastics and other non-metallic materials. Terhertz waves are far more useful in detecting a range of different materials than metal detectors and it is possible to discern the actual chemical signatures of concealed objects, including explosives. Though this can be done with X-rays, terahertz scanners use non-ionising radiation that doesn't carry health risks, making it far more suitable for mass screening of commuters.

'The only way to tackle this kind of threat is with good old-fashioned police work.'

However, despite these developments, it is worth remembering that none of these technologies are infallible. A study published just three days before the London bombings revealed that even if the technology was affordable and reliable enough to deploy across the board, there will always be some suicide bombers that manage to slip through the net.

The study, carried out by Edward Kaplan at Yale University in Connecticut, was a mathematical analysis of how effective networks of terahertz scanners and explosives sensors would be at detecting pedestrian bombers in time to stop them. He found that even with widespread deployment, this technology would not be able to reduce the number of casualties in any meaningful way.

In any case, in a fight against suicide bombers, any technology that is deployed to detect bombs on site is flawed from the start, because the bomber is prepared to die with his victims. According to Kaplan the only way to tackle this kind of threat is with good old-fashioned police work, in the form of intelligence gathering by humans. With a little intelligence, the fight against terror turns into an exercise in tracking terrorist suspects – this too is an area of work that has its own high-tech aids.

Tracking suspects

ID cards and biometrics passports are both due to be introduced in the coming years, at least partly in the name of counter-terrorism. But it is still far from clear how these systems will help in the battle. At the moment, it looks unlikely that it will be mandatory to carry an ID card and the police will not be allowed to stop people and demand them anyway. Similarly, biometrics may well turn out to be very good at ensuring the person bearing a passport is in fact the owner, but that isn't much help if a terrorist obtained the passport using false documents.

However, there may be other ways to identify people that don't involve forcing the rest of society to relinquish their civil rights and privacy unnecessarily. How? In a word – satellites.

According to movies like Enemy of the State and TV series like 24, spy satellites are now powerful and accurate enough to detect the difference between an unarmed, innocent passer-by and a sniper. Some conspiracy theorists would even have you believe that the lenses in these satellites are powerful enough to read the paper over your shoulder from more than 100 kilometres away in space. Whilst both of these scenarios are unlikely, the truth of the matter is that we are not privy to what intelligence services can and can't see. Satellite tracking is a useful tool, certainly in the military and presumably in the world of counter-terrorism too. Although it may or may not resolve quite the level of detail seen in 24, the capabilities of satellite detection systems are certainly improving all the time.

'We are not privy to what intelligence services can and can't see, but satellite tracking is a useful tool.'

But soon, spy satellites won't be necessary to find people. Later this year, the first of 30 satellites will be launched as part of the European equivalent to the Pentagon's global positioning system (GPS). This constellation of satellites, called Galileo, is designed specifically for non-military uses. It will greatly improve civilian satellite positioning devices, with the ability to pinpoint a precise location as accurately as the Pentagon's GPS system and with a reliability that potentially outstrips it.

Then there are mobile phone location-based services (LBS). This is positioning technology that works out roughly where you are by detecting your nearest mobile phone base station, and in some cases how far you are from it. When combined, Galileo and LBS will make tracking people simpler, quicker and more precise than ever before.

But despite the paranoia people may have about 'being watched', by itself there is nothing about satellite positioning technology that allows for people to be located or tracked. The satellites themselves merely beam down time signals which the receivers passively monitor. The clever bit lies in the receiver's brain which calculates a position by comparing discrepancies in signals from different satellites. Receivers do not transmit the location back to the satellites or anywhere else for that matter. All they do at the moment is to calculate a position. But where this technology becomes most powerful, and perhaps where the paranoia might be warranted, is when the receiver has the ability to transmit the position it has calculated. By giving a receiver the ability to transmit using the mobile phone networks, the device can not only broadcast your position but also, unlike normal GPS receivers, it can work indoors – this is the kind of tracking device that the 24 team plant on their suspects.

'Later this year, 30 satellites will be launched as part of the European equivalent to the Pentagon's GPS – it is called Gallileo.'

The convergence between GPS and mobile technologies is already being used in trials in the UK to track ex-offenders like paedophiles and violent recidivists. But with the current GPS system there are shadows and blackspots where high-rise buildings block satellite signals, making it possible to slip off the radar. And there are too few GPS satellites to get an accurate fix anyway. Galileo and improved LBS will ensure that there are fewer places to hide.

So the technology works and is likely to get better and more widespread, especially as mobiles start to have built-in GPS receivers. So the question is, why on earth would anyone want to carry such a device? The answer to this lies in the fact that Galileo is being launched purely as a commercial venture, to make money. The idea is that service providers will offer mobile phone users bundles that include attractive location-based services tailored to their needs. The benefits of these, we are told, are likely to be so high that people will be willing to sign up.

It is possible to tell an awful lot about someone from their use of the Internet, by combing which websites they have visited and how long they have spent at each page. In the same way it will be possible to glean information about a person from where they go in the physical world.

'Galileo and location-based services will make tracking people simpler, quicker and more precise than ever before.'

Even if a subscriber is on an anonymous pay-as-you-go contract, it may still be possible to identify and so track them. For example, foot soldiers with a clean record who have been recruited by a terrorist organisation may reveal their affiliations simply by where they go.

The big 'but' in all this is that it assumes intelligence about who to track in the first place. Furthermore, as society becomes increasingly hooked on these kinds of communication technologies for recreational and commercial ends, it may in fact become even easier for people to disappear simply by opting out of the technologies altogether.

So the question arises, where does the intelligence on who to track come from?

Data dredging

If there's one thing that distinguishes Jack Bauer and his team at the Counter-Terrorism Unit in 24 from other intelligence shows, it's the copious amounts of keyboard-tapping that goes on. If the programme is anything close to reality, the secret to catching terrorists lies in the ability to cross-reference multiple databases to build profiles. As dull, and perhaps disappointing, as this might sound, this kind of data-mining is an extremely powerful tool and probably likely to become the most effective early warning tool for detecting terrorist activity – besides actual infiltration of a terrorist network, of course.

But what Bauer's colleagues appear to do so efficiently by hand is now becoming possible automatically – the ability to create a profile of someone by sifting through enormous amounts of data. Just by cross-referencing telephone records, bank and credit card details, utility bills, job records and much more it is possible to create a detailed profile of a complete stranger. In fact it is even possible to find people this way – with or without mobile phones and satellites.

'The secret to catching terrorists lies in the ability to cross-reference multiple databases to build profiles.'

Similarly, attention can be alerted to a particular individual just by ploughing through other, even less obvious, information. For example, by monitoring the purchasing of certain chemicals it may have been possible to prevent the Oklahoma City bomber, Timothy McVeigh, from killing 168 people in 1995. The same could also be said for the World Trade Center attacks, where intelligence officers failed to react to warnings about the attack because it was subsumed in so much other information.

Such lessons demonstrate that the information sought is often out there, it is just a question of finding it. The trouble is it is a job for which human beings are not particularly well equipped. But for computers, on the other hand, it is a very different story.

One system, called Knowledge Aided Retrieval in Activity Context (KARNAC), is being developed to specifically profile different types of terrorist attacks and to determine possible future terrorist incidents. Developed by Applied Systems Intelligence in Roswell, Georgia, the system works by scanning through information derived from a range of structured and unstructured databases, such as gun registrations, driving licences, residential and criminal records, as well as the Internet, newspapers and county records.

'Systems do exist that may help to detect terrorists even before they have started planning their attack.'

A similar approach is being applied for detecting infectious diseases and potential bioterrorist attacks long before doctors are aware there is a problem. A doctor examining a patient may not immediately assume that a patient presenting with a rash and a fever has smallpox. But the new system, called Lightweight Epidemiology Advanced Detection and Emergency Response System (LEADERS), could by observing similar symptoms in neighbouring hospitals. In such circumstances LEADERS would tell the doctors to isolate the patients and inform the relevant government agencies, such as the US Center for Disease Control.

Of course, in the context of terrorism this sort of action would save lives but would come at the tail end of an attack. Other systems, however, do exist that may help to detect terrorists even before they have started planning their attack.

'It seems that preventing more murders means better intelligence.'

Terrorist attacks are usually funded by illicit money. Find the money and you can find the terrorists. But banks are not really equipped to spot this kind of money laundering activity. Even automated systems that use rules to spot abnormal behaviours can quickly find their rules outdated. It is a proverbial needle in a haystack. But one UK company, called SearchSpace, has developed a data-mining tool for banks that uses artificial intelligence techniques similar to the CCTV system, to spot unusual activity within bank accounts. The system is now being tested by a large number of banks and has already led to money laundering investigations in a number of different countries.

Ultimately there is no one solution to the problem of terror and as often as not there are flaws in the technology. The media is currently obsessed with the potential of systems that can detect bombs in vulnerable hotspots, but these fall far short of prevention: their main use is in the tragic aftermath. It seems that preventing more murders means better intelligence. And despite the various pros and cons, one thing seems clear: the further upstream of the process the technology is, the more effective and powerful it appears to be, and the more likely counter-terrorism is to succeed in its ultimate aim of prevention.

Find out more >>

top ^