The function of controlling robotic structures using the power of thought is known to scientists for years. There are two types of brain-computer interface: risky and expensive implant installation for smooth control or the use of a special "helmet" with reduced accuracy of command input. The team of American researchers presented an improved version of a safe way to control the robot, which does not require a visit to a neurosurgeon.
In an article published in the journal science Robotics, researchers at Carnegie Mellon University describe a combination of sensing and machine learning techniques to create a brain-computer interface for users wearing hats with special EEG sensors. The development team announced the creation of the first brain-controlled robotic arm with smooth movements. Previously, this level of interaction with the prosthesis was theoretically available only for models connected to the intracranial implant.
"Despite the technical challenges of using non-invasive signals, we plan to bring this safe and cost-effective technology to people who can benefit from it," the developers said in a press release.
To test their system, the scientists asked the participants of the experiment to direct the robot manipulator to the cursor moving on the computer screen. The robotic arm showed the ability to continuously track the cursor in real time without sudden movements, which is unusual for a non-invasive control system.
In the future, the new interface will be used not only for medical purposes, but also to create a number of "ubiquitous technologies that help everyone, like smartphones."