"This product may be harmful to your health" could just as well be printed on every car, as it is on packs of cigarettes. That's the opinion of Professor Kare Rumar of the Swedish Road and Transport Research Institute in Linköping. "Being part of road traffic is still the most dangerous everyday activity in most countriesabout 40 times more dangerous than working in industry," says Rumar. Statistics corroborate this grim picture. About 42,000 persons currently die in traffic accidents every year in Europe. Only cancer is associated with a higher risk of fatality. Yet, the number of road accident victims has dropped by almost half since 78,000 were killed in 1970, the worst year ever.
The automotive industry itself is taking part of the credit for this improvement, mainly because cars have become much safer in the last 30 years. Systems like the airbag (introduced on the market in 1981), intelligent wheel brakes such as ABS (1978), ESP (Electronic Stability Program, 1995) and BAS (Brake Assistant, 1996) have reduced the seriousness of accidents or even prevented many. But the automakers don't intend to rest on their laurels. The DaimlerChrysler AG board member responsible for research, Klaus-Dieter Vöhringer, describes the goal as follows: "In the future, it will be possible to avoid half of all accidentsprovided the vehicles involved are equipped with driver assistance systems." Electronic systems are supposed to recognize critical situations early enough to inform or warn the driver, or automatically intervene in an emergency. Vöhringer even believes that they will put the vision of "accident-free driving" within reach in the coming decades: "We have intentionally framed our vision in the most ambitious way in order to exploit all the technological solutions," he says.
The research results give cause for optimism. Today, for example, some 38 % of fatal traffic accidents occur because the driver was distracted or falling asleep; 46 % are due to miscalculations (see chart).
It is still unclear, however, whether these virtual co-drivers operate in a foolproof manner in everyday use. Don't they lull the driver into a false sense of security? Isn't additional information more apt to distract than safeguard? To answer questions like these, the engineers are getting more and more psychologists, anthropologists, ergonomists and neurophysiologists involved: "Drivers don't always react in the same way, or even rationally," says Dr. Werner Reichelt, head of the Man-Machine Interaction unit at DaimlerChrysler. "We want to identify these differences and control their complex interactions."
One example is the flight reflex: If the driver has to avoid an obstacle that suddenly appears, he or she is inclined to engage in exaggerated steering maneuvers. Another example is selective perception: Drivers devote their attention only to what they consider important at the moment, such as the reading of street names or the whining child in the back seat. They might then fail to notice a red light, a pedestrian or the brake lights of the car in front. Researchers still know too little about what goes on inside the head of a driver during emergencies such as these. But they need these insights in order to build useful assistance systems.
With the help of neurophysiology, they now want to find out how the brain works during driving. To accomplish this, the researchers are making use of the fact that parts of the brain involved in thought consume a great deal of oxygen. This can be made visible by techniques such as functional magnetic resonance tomography (fMRT). Here, the test subject lies in a tomograph. Through stereoscopic glasses with miniature display screens, they see a maze, for example, through which they navigate using a joystick. The increased oxygen consumption shows which cells are active during this process.
One initial finding was a marked difference between the sexes. Most men orient themselves in the mazeor on the streets of a cityby drawing a map inside their heads. Women more often feel their way along using landmarks. These insights could improve the navigation systems, says Stefan Hahn, head of Image Recognition at DaimlerChrysler Research. He suggests that, in the future, the displays should provide both "street maps for the map interpreters as well as photos for the landmark interpreters."
The machines of an fMRT examination are too heavy and bulky for mobile operation in a car, however. It is better to use a mobile electroencephalograph (EEG), which measures brain waves via electrodes placed on the head and derives an activity pattern from them. Before year-end, the first series of tests along these lines are expected to be under way in the driving simulator. However, the procedure furnishes cruder 3D images than the fMRT technique. This could be remedied using a method introduced not long ago by Finnish neurobiologistsnear-infrared imaging (NIR). NIR makes use of the fact that active brain cells reflect a laser beam in the near-infrared range differently than inactive cells do. The laser light penetrates the brain to a depth of about six centimeters. Although the convenient-sized NIR devices provide images at least as good as those of fMRT, they have not yet been perfected. Whether the approaches of neurophysiology will ultimately deliver what researchers hope for is something "one can only speculate about," says Hahn. "But it would be a mistake not to sound out their prospects thoroughly," he adds.
One thing is certain already, however: No matter how drivers behave in extreme situations, electronic assistance systems must not incapacitate them, say the German automotive industry and the European automobile manufacturers association ACEA. The full responsibility for driving behavior continues to lie with the driver, they note. Hence, a certain risk will continue to be present in the future as well.
Rolf Sterbak