Independent thesis Advanced level (degree of Master (One Year)), 20 credits / 30 HE credits
As more and more people conduct vigilance-based activities at times other than the traditional daytime work hours, the time utilization will continue to escalate in the next century and will further increase the risks of sleepiness-related accidents.
This project, which is commissioned by Scania CV AB, is to nvestigate the potential of a method for sleepiness detection belonging to esium AB. Our objective is to examine whether Scania CV AB should continue with the investigation of the patent method, and in that case, which patent parameters, that indicate sleepiness, should be more closely inquired. The purpose with the method of patent is to discover a sleepy driving behaviour. This method is based on the interaction that appears between the driver and the vehicle. The interaction consists of small spontaneous corrections with the steering wheel that in this report is called micro communication. How well the interaction is functioning can be measured in degree of interaction, which shows how well the driver and the truck interact with each other. The interaction between the driver and the vehicle is in this report looked upon as answers and questions with a certain reaction time, which appears with a certain answered question frequency. The differences in the signal’s amplitudes are measured in variation in amplitudes.
Experiments to collect relevant signals have to be conducted in order to investigate the potential with the method of the patent. It is eligible to collect data from a person falling asleep, which implies experiments conducted in a simulator. The experiments are executed in
a simulator, one test when they are alert and one when they are sleep deprived. Tests are also executed in a Scania truck. The purpose with these experiments is to collect data of the subject’s normal driving pattern in a truck and to investigate if it is possible to obtain
acceptable data in a truck.
The sleepiness experiments have indicated that the micro communication takes place in a frequency range of 0.25 to 6.0 Hz. The variables that have been found to detect sleepiness with high reliability are the reaction time and the degree of interaction presented in spectra.
The validation experiments have shown it is possible to collect exact and accurate data from the lateral acceleration and the steering wheel torque. But, there is more noise in the signals from truck then there is in the signals from the simulator.
This method for sleepiness detection has, according to the authors, a great potential. However, more experiments have to be conducted. The authors suggest further sleepiness experiments only conducted during night time. The subjects are sufficiently alert in the beginning of the test to receive data from normal driving behaviour. Physiological measurement could be interesting to have by the side of the subjective assessments as an additional base for comparison.
Institutionen för konstruktions- och produktionsteknik , 2006. , 135 p.
Sleepiness detection, micro communication, interaction, steering wheel torque, lateral acceleration