ЮФУ разрабатывает биоморфного робота с мемристивными технологиями

The robot's limbs will be able to feel and transmit sensations to the operator

In Taganrog, at the campus of the Southern Federal University, a robotic complex (RTC) has been developed that mimics the shapes and functions of a human hand. The key difference of the system is the memristive sensor modules implanted in the limbs, which reproduce tactile and nociceptive memory at the hardware level. Along with this project, a new educational program has been launched to train electronic engineers and microelectronics specialists to work with neuromorphic devices.

The Laboratory of Neuroelectronics and Memristive Nanomaterials (Neuromena) at SFedU has been developing in this field for over a decade. Memristors — elements whose resistance depends on the previous history of currents and voltages — are used here as a "hardware" neural network. Thanks to these devices, the robot "feels" the stiffness, temperature, and consistency of the surface, just like a human does thanks to neurotransmitters in the skin.

Our scientific group has been working on the development of memristive nanomaterials and their use in creating a wide range of promising electronic devices for over 10 years: from designing a modern computer platform and the architecture of the future, to the neuromorphic structures themselves for various applications.
Vladimir Smirnov, Head of the Neuromena Laboratory, Ph.D. in Engineering, Head of the Department of Radio Engineering Electronics and Nanoelectronics at SFedU
Robotic arm and glove for control
Robotic arm and glove for control

The first sample was a "glove — robotic arm" tandem in "shadow" mode: the operator's finger movements are accurately copied by the model. The system combines sensors and a memristive processor, minimized in size and energy consumption.

Our task is to miniaturize devices, a balanced and lightweight system that consumes very little energy and retains human-likeness in terms of shape.
Maxim Kartel, fourth-year student

The mega-grant project brought together over 30 participants: graduate students, candidates, and doctors of science. Among the supervisors are Professor Park Bae Ho from South Korea and an expert in nonlinear dynamics, Professor Viktor Kazantsev from UNN named after Lobachevsky.

The immediate plans include expanding the robotic complex to a full "anatomical" set with sensory limbs, as well as applying the developments in rescue operations and in the manufacture of bioprostheses.

Read more materials on the topic:

Every third Russian dreams of a robot assistant with AI

Highly accurate and easy to control: a prototype of an industrial robot consuming 30% less energy has been developed

A unique reconnaissance robot with an unusual principle of movement has been developed in Russia

Now on home