Taking on the drugged and drunk drivers
03 December 2005
IN A government lab in a bleak, misty field in the middle of the English countryside, I am playing with a gadget that looks not unlike a PlayStation Portable. I have to press one button when an arrow on the screen fleetingly points to the right, and another when it points to the left. How quickly I respond is used as a measure of my sobriety.
Forget breathalysers and walking the line: tomorrow's drunk or drugged drivers could be rumbled by police officers wielding just such a gadget. And if the software says someone's psychomotor skills are as impaired as their driving suggests, a sliver of silicon will then be produced to indicate from a saliva sample just what they have been imbibing, injecting, smoking or snorting.
At least, that's the hope of researchers at the UK government's Home Office Scientific Development Branch (HOSDB), who last week opened their rural lab to the media for the first time. They are developing the technology for a fresh campaign against drugged and drunk drivers.
It's not hard to see why. The Royal Society for the Prevention of Accidents says that of the 3221 people who died in road crashes in the UK last year, 590 were killed by the actions of a drink driver. And of 42,636 deaths on the roads in the US, 16,694 people were killed by drunks, according to the National Highway Traffic Safety Administration.
While no one knows for sure how many of the remaining road deaths were caused by drugged drivers, the Home Office believes that the problem is rife. An analysis by the UK's Transport Research Laboratory in 1999 found that the number of people driving while under the influence of cannabis, cocaine, heroin or amphetamines had increased sixfold since 1990. One reason figures on drug-related accidents are hard to compile is the lack of on-scene drug detection technology, says Home Office spokesperson Fiona Midgely. So the HOSDB lab at Sandridge, Hertfordshire, hopes to give the police gadgets they can use to screen errant drivers at the roadside, to help them decide if they need to be taken to a police station for blood tests.
Today's sobriety tests fall well short on two counts. First, assessing how well suspect drivers perform in standard "field impairment tests" (FITs) - making them walk a line, stand still on each leg for half a minute and touch their nose with their fingertip - are highly subjective. Where one police officer may judge that the driver has failed the test, a second may not, meaning some intoxicated drivers may be allowed to drive away. Second, anyone failing a FIT may still escape a trip to the station because breathalysers cannot detect drugs.
The machine I saw is the first of a new generation of intoxication testing devices being developed by HOSDB to improve roadside assessments. Based on a miniature Sony Vaio tablet PC, the Roadside Impairment Test Device (RITD) will present a suspect driver with tests designed to assess his or her manual dexterity and state of alertness. "Unlike the FIT test, this should give the police an objective measure of someone's impairment with a simple pass or fail notification," says the scientist leading the effort. The Home Office asked New Scientist not to identify the individual involved because much of the work at HOSDB is classified.
The RITD prototype now runs six tests, but after trials on 120 drink and drug-taking volunteers early next year, these will be whittled down to three, which should take a driver no more than 10 minutes to complete. One of them will measure a suspect's ability to sustain attention to a task and respond to a stimulus, such as pressing a button according to what type of road sign appears. At the moment there is also a tracking test to check people's ability to follow a moving object on screen. In one reaction-time test, the user has to move a stylus to one of six buttons when a red dot appears near it. Another is the right and left arrow recognition test I tried, though any of these tests could be dropped.
But some specialists question whether the researchers will be able to find tests that are useful across the whole population. Would a 30-year-old cannabis-smoking gamer beat the system, for example, while a sober 50-year-old who has never played computer games in his life seem impaired? "Our ability to use our fine motor control on a computer varies tremendously from person to person," warns Alan Dix, a specialist on human-computer interaction at the British Computer Society. "In particular, certain people vary a great deal in their ability to react to tasks involving left-hand and right-hand discrimination."
Dix suggests that HOSDB needs to find tasks that drill down to basic neural responses for which the practised reactions of an experienced computer gamer would be irrelevant. "If you bang a book on a table and make someone jump, that is not a practised reaction, for instance," he says. "I'm sure this is doable. They just have to find tests where practised reactions are not a major influence, but where intoxicants affect the reaction."
Writing in the journal Perceptual and Motor Skills (vol 101, p 383), Julia Boyle and her colleagues at the University of Surrey, who worked alongside the HOSDB to develop an early version of the RITD, reports that "the device has only been shown to be sensitive to the impairing effects of alcohol in trained subjects". They conclude that much more work is needed to make it a robust "one-off" tester for untrained subjects.
If and when the RITD goes into service, drivers it shows to be "impaired" will then undergo a roadside swab test that will give an instant read-out on the drugs they have taken. At its heart is a laser-based chemical ID technique called surface enhanced Raman spectroscopy (SERS), which was pioneered by Max Fleischmann, of cold fusion fame.
In standard Raman spectroscopy, incoming photons collide with atoms in a material, and the scattered photons emerge at different wavelengths depending on the atom or molecular group that the photon strikes, revealing which chemicals are present. But only one photon in 10 million is scattered this way. SERS boosts this number significantly and so makes tiny amounts of materials, like traces of a drug in saliva, detectable. HOSDB's system, developed with the University of Strathclyde in Glasgow, UK, performs SERS using the surface electronic effects of a very thin layer of silver deposited on a disposable 2-centimetre diameter silicon wafer.
To use the test in the field, a police officer takes a saliva sample by placing a cotton bud under the driver's tongue and then wipes the saliva onto the surface of the silicon wafer. "The machine will give a red light if drugs that affect driving are present," says the scientist leading the project.
However, neither the RITD nor the SERS device is meant to provide evidence that can be used in court. The idea is merely to help police officers at the roadside decide whether someone needs to be arrested and taken for a blood test at a police station. The Home Office hopes trials of both devices will be complete within two years and that they can then be issued to the police.
reproduced by kind permission of
Return to Latest News Page