In a gymnasium in Minnesota, a small quadrotor drone flies effortlessly through hoops. But look for its pilot, and it's not obvious who it is, because it's being flown by someone's brain.
Scientists at the University of Minnesota have designed and tested the world's first noninvasive, brain-controlled helicopter in a breakthrough that could one day allow people who are paralyzed to operate wheelchairs, robotic arms or electronics using only their minds. The study is described in the Journal of Neural Engineering.
The project was carried out by Bin He, whose lab has been working on brain-computer interfaces for more than 10 years. To fly the drone, a commercially-available Parrot AR.Drone 2.0, the pilot simply needs to "imagine" making a fist. A cap with 64 electrode sensors picks up the pilot's brainwaves and directs them through a computer, which wirelessly transmits them to the drone. To fly left, the pilot imagines making a fist with his left hand – to fly right, he imagines making a fist with his right hand. Imagining making a fist with both hands flies the helicopter up.
"We've demonstrated for the first time that this kind of thing can be done nonintrusively," says He. Previous brain-computer interfaces have had to connect a chip to a subject's brain in order to accurately control a device using brain waves.
Before flying, the average pilot needs to train for about 10 hours – a computer picks up the electric signals sent by the brain so that it can interpret them later, when the pilot is actually flying the drone. He says that merely thinking about making a fist causes the brain to react in a similar way to when a person actually makes a fist, making the technology potentially useful for people who are paralyzed.
"The brainwaves contain signatures for each person, but there are some common features from person to person," he says. "There's a high degree of similarity between imagining or thinking of movement as compared with actually moving that part of the body."
He said he wanted to use a helicopter because it is inherently more difficult to control than an object on a flat surface. In order for the technology to be useful for people in the future, they have to be able to effortlessly perform complex movements.
"It's not that complicated to move a car on a surface compared with the three dimensions of a helicopter in the sky," he says.
After the team's successful test, they are in the early stages of designing a test for people who have been paralyzed.
"This has to be a step-by-step process, but we'd like to move to disabled patients soon," he says. "We want to allow them to control a wheelchair or switch the channel on a television using just their thoughts."