Nerves for a cyborg
Psychiatrist Nathan Kline and engineer Manfred Klines, who coined the word "cyborg" in 1960, imagined powerful half-humans, half-robots who possess unprecedented strength and "superpowers". Only by becoming a cyborg, they believed, could a person really master outer space.
However, the idea of supplementing the human body with machine components turned out to be in demand not only, and not so much in space. Too many people need this supplement not to become superman, but to regain simple human abilities – for example, the ability to get keys out of their pocket.
Tens of millions of people in the world have experienced amputation of one or more limbs. Until the second half of the 20th century, they would all have to be content with a piece of wood and metal instead of a hand, which could be somehow wielded, but not to do delicate operations. But now progress in the field of robotics allows us to make mechanical hands that are able to work almost the same as real ones – they can move their fingers, grab objects. Bionic prosthetic hands are becoming more and more common. But so far, no technically satisfactory solution has been found to the main problem – connecting a robotic limb to a person.
How to make an electronic-mechanical hand obey commands from human nerves? Engineers and doctors used many different approaches – implanted electrodes, tried to remove electrical signals directly from nerves, from muscles, tried to subordinate prostheses to contractions of the remaining muscles. In the end, sooner or later they managed to achieve success. But in all cases, a months-long process of teaching the prosthesis and the patient the skills of teamwork was required. Even now, when bionic prostheses are no longer made in laboratories, but in workshops, when large companies are engaged in this, mutual "training" remains a complex and time–consuming process - primarily for the patient himself.
The GalvaniBionix team, consisting of MIPT students and postgraduates led by Timur Bergaliev, set itself the task of facilitating the process of mutual adaptation, making it as simple and fast as possible. "Until now, a person has been learning to control a prosthesis, to understand it. We are making a system that understands a person," says Bergaliev.
The device created by his team registers and processes electromyograms, that is, electrical signals arising in muscle cells. With the help of machine learning algorithms, the program learns to correctly recognize certain "muscle commands", adapting to a specific person.
Bergaliev's group has already tested how the prototype of the device works under the control of a person with an amputated limb – he managed to move the cursor around the screen using "muscle signals".
The backbone of the team is four graduates of Phys Tech: Timur Bergaliev, Evgeny Zhvansky, Anatoly Kostin and Nikita Orlov. They had the idea to create an interdisciplinary project at the intersection of biophysics, electronics and data analysis. Full-scale work began in September 2014. The funding came from their own pocket, in addition, funds were received from the student projects fund of the 5top100 program, from the Bortnik Foundation.
Already in April 2015, the project won the Russian stage of the Microsoft ImagineCup-2015 competition, but there is still a selection for the international stage of this competition in Seattle.
The GalvaniBionix team has signed an agreement with the Galatea research and production company, which is a leader in the production of mechanical prostheses in Russia; the project participants are implementing their control system into existing mechanical prostheses. There is communication with foreign colleagues, although there are no specific agreements yet.
"Ultimately, we plan to develop our own bionic prosthesis that adapts to a specific person," says Bergaliev.
However, he and his comrades do not intend to limit themselves only to prosthetics. In their opinion, based on their prototypes, it is possible to create control systems for exoskeletons, robotic systems, for military robotics. The operator will be able to control the manipulator as if it were his own hand – this will shorten the training time and increase the reaction speed.
"In parallel, we are developing a project of wearable electronics for sports and fitness: electromyography methods allow you to work with muscles, analyze and plan training loads," says Bergaliev.
According to him, today there is no objective way to measure the condition of muscles during training. If professional athletes are constantly under the control of experienced coaches, then amateurs do not have such an opportunity, and everything is done "by eye", which often leads to injuries: "Our device will allow you to monitor the load level and warn the athlete if it is close to the limit."
Portal "Eternal youth" http://vechnayamolodost.ru19.05.2015