Why it matters to you
Drowsiness is a major cause of car crashes, but new project funded by the U.S. Army has developed an algorithm for detecting sleepy drivers based on their EEG brain wave readings.
Next to inebriation, speeding, and inattention, people driving while drowsy is a major cause of car crashes. Well, forget about cranking loud music, drinking convenience store coffee, and keeping your windows rolled down, because the United States Army has an alternative solution: reading your brain waves.
In a new research paper, funded by the Army’s Human Research and Engineering Directorate (HRED) unit, computer scientists describe an algorithm for detecting driver drowsiness based on their electroencephalogram (EEG) signal.
“Since EEG signals directly measure the brain activities, theoretically EEG based approaches should be more reliable [than other solutions],” co-author Dr. Dongrui Wu told Digital Trends. “More importantly, it may be possible to predict the driver’s drowsiness from EEG signals before it actually happens, which gives us ample time for interventions.”
More: SmartCap monitors brain waves in machinery workers and truck drivers to keep them alert
As part of the study, 16 different participants had their brain waves monitored while carrying out a simulated drive down a highway, complete with occasional lane changes. The algorithms the team had developed were then used to analyze the EEG results and predict drowsiness.
“The main real-world application is driver drowsiness estimation to prevent accidents,” Wu continued. “In the paper we reported an application scenario of highway driving, but the technique can also be applied to other professionals including pilots, soldiers, and construction vehicle drivers. Another possible application is to monitor in real time the sleep status of a person from EEG, in the hospital or at home. Beyond EEG and brain-computer interfaces, the algorithm proposed in this paper can also be used in many other applications that involve people and hence individual differences, [such as] affective computing, which involves estimating the emotions from a person’s voices or other body signals.”
For now, Wu said the researchers are continuing to improve the algorithm they’ve developed for more accurate detection. They additionally want to experiment with different numbers of, and locations for, EEG sensors to get the most accurate predictions possible. By doing this, they hope to achieve the optimal compromise between performance and portability.
After that we guess it’s just about finding the best way of jolting drivers awake and keeping them that way. Until autonomous cars have been fully perfected, that is.