The first prototypes of brain-computer interfaces that don’t require invasive brain implants have already been developed by researchers. Within several years, brainwave-controlled robotics are likely to have become relatively normal. So think a team at Carnegie Mellon University in Pennsylvania who have been demonstrating a robotic arm controlled through brainwaves picked up by headgear as light as a shower or swimming cap.
The cap contains electrodes able to pick up brain waves and AI which filters out the ‘noise’, so only relevant signals are acted upon. Until now, the only way to pick up brainwaves clearly enough to interpret commands was through implants placed beside the brain that were both expensive and high risk.
Electroencephalogram caps are covered in 57 electrodes that picked up the electrical activity coming from the brains of volunteers participating in the R&D project. EEG caps, as they are called, are not a new invention and have been tested before. However, it was difficult for machines to pick the ‘right’ information out of all of the electrical impulses crackling out of brains. A combination of improved sensor technology and machine learning developments are now moving close to meaning EEG caps will be enough for users to communicate reliably precise commands to robotics.
Described in the Science Robotics journal, the Carnegie Melon team have shown how their robotic arm solution allowed for smooth movements rather than the kind of jerky, disconnected movements achieved until now with a comparable headset. The hope is that within a decade or so, new prosthetics will be almost as functional as biological limbs with delicate movements able to be controlled be the wearer subconsciously in much the same way as an actual arm, leg, hand or foot. The new technology could also be of huge help to stroke victims, actively training their brains to move affected limbs, initially supported by robotic devices.
As well as in medicine, there is also a lot of investment and effort going into the development of brain-robot interfaces in the field of industrial robotics. Professor Bin He, professor of biomedical engineering at Carnegie Mellon believes that brain-computer interfaces may one day, quite possible within 2-3 decades, become as pervasive as smartphones are today.