Mind control for exoskeletons (STEAM Register)
A brain-computer control interface for a lower limb exoskeleton has been developed by scientists working at Korea University and TU Berlin.
The interface works by decoding specific signals from within the user’s brain via an electroencephalogram (EEG) cap. This allows users to move forwards, turn left and right, sit and stand simply by staring at one of five flickering light emitting diodes (LEDs).
Each light flickers at a different frequency. When the user focuses on a specific LED, it is reflected within the EEG readout. This signal is identified and used to control the exoskeleton.
A key problem has been separating these precise brain signals from those associated with other brain activity, and the highly artificial signals generated by the exoskeleton, reports Steve Pritchard for The Institute of Physics.
“Exoskeletons create lots of electrical ‘noise’,” explains Klaus Muller. “The EEG signal gets buried under all this noise – but our system is able to separate not only the EEG signal, but the frequency of the flickering LED within this signal.”
Although the test were performed on healthy individuals, the researchers say the system has the potential to aid sick or disabled people.
“People with amyotrophic lateral sclerosis (ALS) [motor neuron disease], or high spinal cord injuries face difficulties communicating or using their limbs,” continues Muller. “Decoding what they intend from their brain signals could offer means to communicate and walk again.”
Training on the system takes only a few minutes and the flickering lights were carefully screened for epilepsy on participants prior to them taking part in the research. The researchers say are now working to reduce the ‘visual fatigue’ associated with longer-term users of such systems.
“We were driven to assist disabled people, and our study shows that this brain control interface can easily and intuitively control an exoskeleton system – despite the highly challenging artefacts from the exoskeleton itself” concludes Muller.
The results are published today (Tuesday Aug. 18) in the Journal of Neural Engineering.
Next Article: Hitchhiking robot attacked by vandals in Philly