14 DAY TRIAL // A new project that combines brain control and artificial intelligence could open new horizons for paralyzed people.
This robotic wheelchair can be maneuvered through thoughts, thanks to a innovative approach called “shared control”, that turns brain signals into computer commands.
José de R. Millán’s laboratory in the Neuroprosthetic Center at the Federal Institute of Technology in Lausanne has developed a wheelchair that features a software that can “translate” a simple command like “go left” into a way of doing so without hitting anything.
The software also understands when the driver wants to go towards a certain object, like a table or a chair.
Patients can control computers, prosthetics and other devices using several technologies, that capture signals from the nerves, muscles or from the brain, and in this case, scientists used electroencephalography, or EEG.
The user has to wear a skullcap and be patient for a few hours a day for about five days, the time to get accustomed to the machine.
The chair is being controlled by the mere thoughts of the user, who imagines that the wheelchair is only a moving part of the body.
Michele Tavella, an assistant in Millán’s laboratory at EPFL in Switzerland, tested and demonstrated the chair's performances: he managed to move around a room, avoiding obstacles along the way, by simply concentrating.
The idea is that if the user thinks of moving the left hand, a specific brain pattern will be recorded by the electroencephalography and interpreted by a computer that will tell the chair to turn left.
“When I want to turn left, I imagine moving my left hand,” Tavella says, “and this is very natural and very quick; I can send a command in about a second.”
Now, even if this is an amazing process, the EEG has a rather limited accuracy, and only a few commands are detectable.
Also, the entire control process demands a lot of concentration and in a cluttered environment it can become very tiring for the user.
José del Millán, director of noninvasive brain-machine interfaces at the Federal Institute of Technology, and leader of the project said that “people cannot sustain that level of mental control for long periods of time” and the more you concentrate, the noisier the signals become and the more difficult is for the computer to interpret them.
On the other hand, the system compensates by its “shared control”: the patient only needs to think the command once – move forward for example, and the two small cameras on each side of the chair, along with the image-processing software will avoid the obstacles along the way.
This prototype uses 16 electrodes that monitor the user's brain activity and so far it has not been tested on any paralyzed patients.
EEG signals can be slow and tricky to work with, says Damien Coyle, a researcher in the Brain-Computer Interfacing and Assistive Technology group at the University of Ulster.
This is why the shared control is a very good solution and Millán's project is one way of putting it into practice.
“The more shared control you have, the better the brain-computer interface, and the faster the person can get from one place to another,” he said.
Millán's team is already working on new ways of making the chair smarter, according to Technology Review.
Watch the video of the wheelchair controlled by thought alone: