NIIME: Active Exoskeleton Arm Platform with Neural Network Muscle Control

Main article: Exoskeletons

2023: Exoskeleton development

Employees of the Molecular Electronics Research Institute (NIIME) have developed an exoskeleton designed for use in rehabilitation medicine. This was announced on February 1, 2023 by the head of the Department of Investment and Industrial Policy of the city of Moscow, which is part of the Complex of Economic Policy and Property and Land Relations of the capital, Vladislav Ovchinsky.

NIIME has been a resident of the Special Economic Zone (SEZ) of Moscow since 2015. The company conducts research and development work in the field of micro- and nanoelectronics. As of February 2023, over 600 highly qualified specialists work at NIIME, including three academicians and two corresponding members of the Russian Academy of Sciences, more than 70 doctors and candidates of sciences. This development will be in demand primarily in medicine, for example, in the rehabilitation of patients after a stroke, - said Vladislav Ovchinsky.

The creators of the exoskeleton believe that the main drawback of the models available in the world is their control principle. To make such exoskeletons move, you need to press the remote control button. In industrial exoskeletons, control is often transferred to the operator. Another drawback of most samples is the pre-prescribed trajectory of the exoskeleton, regardless of the user's desires and physiology.

Employees of the resident of the SEZ "Technopolice Moscow" - the Research Institute of Molecular Electronics - have created a platform for an active exoskeleton of the hand with neural network muscle control for solving medical, industrial, rescue and special tasks. It is based on the author's management principle, - said Gennady Degtev, General Director of the Technopolice Moscow SEZ.

The exoskeleton use market is growing annually, demonstrating high potential in industries where enhanced human physical capabilities are required. For example, in the elimination of the consequences of emergency situations, construction or industry.

We propose to non-invasively read electromyographic signals from human muscles and record the actual position of the hand. A trained neural network from this data will predict which trajectory is most likely in a period of 0.2-0.3 milliseconds. Thus, we get an active gain in vital movement, quickly and accurately predicting the trajectory of different movements, taking into account the individual physical characteristics of the user. The author of the idea and the leader of the team is a researcher at the laboratory and a master's student at the Moscow Institute of Physics and Technology German Karnup, "said Oleg Telminov, head of the NIIME target search laboratory for the study of neuromorphic systems.

As of February 2023, the developers are consulting with the rehabilitation centers of Sechenov University, the National Medical and Surgical Center named after N. I. Pirogov, the rehabilitation center "Three Sisters."