Fatigue at the wheel: Mercedes-Benz developing warning system for motorists

According to official statistics in Germany, around 0.5 per cent of all serious traffic accidents are caused by "over-fatigue". Last year, the police put a total of 3034 accidents down to fatigue, including 1786 collisions resulting in personal injury. This equates to some 0.5 per cent of all serious accidents. Experts believe that the true figure is much higher, however, as it is often impossible to detect and verify signs of fatigue by the time accidents are reconstructed. Various scientific studies estimate that between 10 and 20 per cent of serious traffic accidents can be attributed to drowsiness. According to an investigation carried out by insurance companies in Germany, fatigue is responsible for one in four fatal motorway accidents. The probability of being killed in an accident resulting from fatigue is 2.5 times higher than for all other causes of accidents.

These findings are corroborated by accident researchers outside Germany. The American National Highway Traffic Safety Administration (NHTSA) estimates that every year over 100,000 accidents are caused by driver fatigue in the US, in which 1500 people are killed and a further 71,000 injured. The NHTSA claims that the risk of an accident or a near-accident occurring increases four to six-fold if driving when tired. Meanwhile, a study into fatigue at the wheel conducted by scientists in Canada established that one in five drivers has at some time nodded off whilst driving.

Two thirds of fatigue-induced accidents occur at night-time, with one in two taking place in light traffic. The majority of incidents of this type happen in the early hours of the morning between 2 and 6 a.m., but they can also occur in the middle of the afternoon.

Mercedes assistance systems honed for real-life accident scenarios

When developing driver assistance systems, Mercedes-Benz works at all times on the basis of what happens in real-life accidents and deploys the very latest technology in its ongoing search for new ways of improving driving safety in critical situations. In the past, the Stuttgart automotive brand has focused on particularly serious accidents resulting from driver errors, rear-end collisions, accidents at night-time and accidents involving pedestrians. This has enabled Mercedes-Benz engineers to obtain remarkable results, as confirmed by the accident statistics:

- Fitting the Electronic Stability Program (ESP®) as standard has brought about a reduction of over 42 per cent in the proportion of serious accidents resulting from driver errors in the accident breakdown for Mercedes passenger cars.

- Brake Assist (BAS), that was likewise invented by Mercedes-Benz and has featured as standard in its passenger cars since 1997, helped to cut the rate of rear-end collisions (per 10,000 new vehicle registrations) by eight per cent for Mercedes passenger cars. At the same time, BAS has reduced the proportion of serious pedestrian accidents by 13 per cent.

The latest project for detecting fatigue at the wheel has yet again seen the experts at Mercedes apply their committed, practically minded approach to avoiding acci-dents. Development of this brand new system got underway several years ago with a series of tests in the driving simulator in Berlin. This was followed by a programme of motorway driving at night-time which 150 test persons have participated in to date, covering over 250,000 kilometres in the process.

Different fatigue detection techniques undergoing trials

Over the course of the practical investigations, the expert team of engineers, cyberneticists, mathematicians, computer scientists and psychologists is testing out a variety of methods for detecting driver fatigue as soon as it sets in. One of these techniques is the eye-blink observation method: an infrared camera directed at the driver's head permanently monitors the eye-blink frequency, enabling microsleep to be detected the instant the eyes stay closed for a certain period of time. A warning signal sounds in the car's cockpit in response.

In order to obtain objective indicators of fatigue, other physiological readings such as the electroencephalogram (EEG) are used. Yet another method is based on the analysis of dynamic driving data, such as steering or braking characteristics. One of the systems triggers an alarm if the driver does not move the steering wheel for a prolonged period of time.

Detecting fatigue by combining measured and empirical values

However, initial results obtained by the engineers in Sindelfingen indicate that the observation of individual criteria alone does not allow reliable detection of tiredness. Fatigue is a highly complex phenomenon that can manifest itself in all manner of ways. In view of this, Mercedes-Benz will make use of a whole host of factors for fatigue detection, including the individual driving style, the duration of the journey, the time of day and the current traffic situation. By continually comparing this data with stored empirical values, the system will be able to compile an individual driver profile and use probability calculus to determine whether the driver is exhibiting the first signs of fatigue.

The principal goal of the development project is to detect the gradual transition from the state of being fully awake to being tired, in other words from high powers of concentration to a clear attention lapse, and warn the driver before he or she becomes over-fatigued and thereby jeopardises safety.

Source: DaimlerChrysler