Fatigue is one of the primary causes of serious traffic accidents. Safety experts from Mercedes-Benz are working on a system which promptly recognises typical signs of fatigue in many cases and can warn the driver. This new development should be fit for series production in a few years.
FATIGUE AT THE WHEEL – A DEADLY DANGER In 2005, police found that fatigue was the cause of 3034 accidents, 1786 of which were collisions causing injuries. This figure represents around 0.5 percent of all serious accidents. However, experts estimate that unidentified cases may make up 10-20 percent of accidents. The probability of having a fatal accident due to fatigue is more than 2.5 times greater than from all other causes of accidents. According to an investigation by German insurance companies, fatigue causes every fourth fatal motorway accident. Accident researchers in other countries have confirmed these findings.
ACCIDENT AVOIDANCE ORIENTED TOWARDS REAL-LIFE ACCIDENTS While developing driver assistance systems, Mercedes-Benz has always used modern technology to seek ways to improve safety in critical driving situations. The standard systems for passenger cars, the Electronic Stability Program (ESP®) and Brake Assist (BAS), are well-known examples of this. With the fatigue recognition project, Mercedes-Benz continues its commitment to accident avoidance by focusing on real-life applications.
FATIGUE RECOGNITION PROCESS IN TESTS During an eye-blinking observation, an infrared camera directed at the head of the driver permanently registers the eye-blinking frequency, recognises when the eyes close for a period of time during microsleep and then responds with an audible warning. Measurements by an electroencephalogram (EEG) also provide objective indicators of fatigue. Yet another detection method analyses driving dynamic data such as steering and braking behaviour and sets off an alarm if, for example, the driver has not moved the steering wheel for a longer period of time.
COMBINATION OF MEASURED AND EMPIRICAL VALUES Fatigue is very complex and has different individual effects. That is why many criteria, such as individual driving style, drive length, time of day and current traffic situation are utilised to recognise fatigue. Using a continual comparison of such data with stored empirical values, an individual driver profile is created with whose help signs for the start of fatigue are recognised, based on a probability calculation. It is only when the sliding transition from awake to tired, meaning when a high ability to concentrate becomes a significant attention deficit, can be unequivocally recognised that the driver can be warned in time.
