UNN pioneers the implementation of neural robot control based on memristors
Memristor-based technology will enhance the control of robotic complexes using brain signals. The system developed at the UNN Educational Electronics Design Centre will make neural control of robots faster, more accurate, and more energy efficient.
The new R&D product can find applications in robotics and medical prosthetics by improving the interaction of patients with wheelchairs, prosthetic devices, exoskeletons and other devices controlled by brain signals. Memristor-based neural control will also help to modernise UAV operator systemsby making UAV piloting more responsive and reliable.
"The more complex the action, the greater the number of brain signals that must be recorded and processed. Thanks to the memristors' compact size and fast performance, no extra computing power will be needed for this task. Neural signals will be processed on small mobile processors and sent to the control system using wireless communication. The new electronic components will reduce power consumption, thereby reducing the device's weight and dimensions. The memristor-based neural control system will be portable, compact, and energy-efficient,"noted Sergey Shchanikov, a co-developer and senior researcher at the UNN Memristor Nanoelectronics Research Laboratory.
An interdisciplinary team at Lobachevsky University has developed a prototype of such a system, where the operator controls the movements of the robot using motor imagination.
"First, the operator learns to imagine different actions, while the brain signals are recorded and analysed. The recorded brain signals are then sent for analysis to the neural control system, a unit with a memristor chip which can be located both on the operator and on the robot or prosthesis itself,"explained Susanna Gordleeva, co-author of the development, winner of the Russian President's Prize in Science and Innovation for young researchers, professor at the UNN Neurotechnology Department.
Brain signals are captured using a classical EEG helmet and transmitted via Wi-Fi to a card with a memristor chip, where the command is processed and sent to the robot. The operator can adjust his command in the process, for example, by changing the direction of the robot's movement.
"Research on neurocontrol of robotic systems at Lobachevsky University has been underway for over a decade. Our new development marks a new milestone in this field. Memristors have been employed in neurocontrol for the first time in Russia. It is important that all the electronics we use are domestic products and technologies. The memristor chips developed by our team are manufactured in the UNN Laboratory for Memristor Nanoelectronics in collaboration with the Sedakov Research Institute of Measuring Systems," said Alexey Mikhaylov, Head of the Research Laboratory for Memristor Nanoelectronics at UNN.
The project also involves scientists from the Research Laboratory forStochastic Multistable Systems of the UNN Research Centre"Physics of Solid-State Nanostructures", as well as students and postgraduates from the UNN Faculty of Physics.
