Ever felt sleepy at work? For most of us, it’s just a sign that we need a break.
But for people who drive or operate heavy machinery, drowsiness can be dangerous—and even deadly.
Drowsy driving alone causes hundreds of fatal car accidents each year in the U.S., and the National Safety Council warns that fatigue is a major risk in industries like construction and mining.
To help prevent these kinds of accidents, engineers at the University of California, Berkeley, have developed a new type of earbuds that can detect when someone is getting sleepy.
These special earbuds work by monitoring brain activity, just like an electroencephalogram (EEG), which is a test doctors use to measure electrical signals in the brain.
Traditionally, an EEG requires several electrodes to be attached to the scalp, but these new earbuds do the job with built-in electrodes that make contact with the ear canal.
Although the signals they pick up are weaker than those from a standard EEG, the researchers have shown that their device, called the Ear EEG, is still sensitive enough to detect alpha waves.
These brain waves typically appear when you close your eyes or start to fall asleep.
Rikky Muller, the senior author of the study and an associate professor at UC Berkeley, got the idea after buying her first pair of Apple AirPods in 2017. She thought they would make a great platform for recording brain activity.
“We believe this technology has many potential uses,” she says. “Detecting drowsiness is just one example—it could also help with sleep monitoring or diagnosing sleep disorders.”
Designing these earbuds wasn’t easy. To get accurate readings, the electrodes need to touch the skin well. In traditional EEGs, this is done with flat metal electrodes stuck to the scalp.
But it’s much harder to make earbuds that fit comfortably in a variety of ear shapes and sizes while still providing a good seal.
Earlier designs from other research teams used gels to improve contact or made custom earpieces for each person. But Muller’s team wanted to create something that anyone could use without special preparation.
They worked with other researchers at UC Berkeley to design a dry, reusable earbud in three sizes—small, medium, and large. The final product includes multiple electrodes that gently press against the ear canal and uses flexible materials to ensure a comfortable fit.
In their latest study, the researchers improved the design and added machine learning to analyze the data.
They tested the earbuds on nine volunteers who wore them while performing boring tasks in a dark room. The earbuds accurately detected when the volunteers were getting sleepy, even when the signal quality wasn’t perfect.
Muller and her team are now exploring other uses for the Ear EEG, which can also pick up signals like heartbeats, eye movements, and jaw clenches. “Wireless earbuds are something we already wear all the time,” Muller points out. “That’s what makes this technology so appealing—it doesn’t require anything extra.”
The study was published in Nature Communications and was supported by several research grants.
