Researchers develop brainwave-detecting algorithm to prevent drowsy driving

Park Sae-jin Reporter() | Posted : November 27, 2019, 11:10 | Updated : November 27, 2019, 11:10

[Gettyimages Bank]

SEOUL --Researchers at a university hospital have developed a new algorithm that analyzes a driver's brainwaves to predict whether the driver will fall asleep while driving within two seconds and prevent accidents.

Seoul National University Bundang Hospital said in a statement on Wednesday that its team of researchers has adopted machine learning, an automated study method for artificial intelligence (AI), to develop an algorithm that collects and analyzes a driver's brainwaves to decide whether the driver will fall asleep behind the wheel.

The hospital said that researchers have found a change in certain brainwave frequencies occurs when the motor response and eye movement of test subjects were slowed due to drowsiness. The patterns of brainwave frequency change occurrences were then studied by an AI through machine learning to find the exact moment when drivers fall asleep. Researchers said that the detecting process for drowsy driving takes less than two seconds.

"By applying this algorithm to the driver, we can predict momentary drowsiness by collecting very short brainwave signals," professor Yoon Chang-ho was quoted as saying. The professor said that the newly-developed algorithm can be upgraded to be applied to small wireless devices to reduce accidents by drowsy driving.

According to Korea Express Corporation, South Korea's expressway operator, about 60 percent of fatal road accidents in 2018 were caused by drowsy driving. Dozing off behind the wheel is extremely dangerous as it is hard for drivers to recognize if he or she is falling asleep. Also, the majority of accidents caused by drowsy driving take place on expressways where cars travel at high speed.
 
© Aju Business Daily & www.ajunews.com Copyright: All materials on this site may not be reproduced, distributed, transmitted, displayed, published or broadcast without the authorization from the Aju News Corporation.