In the future, one of the areas of application of robotics will be the use of robots in scenarios to eliminate the consequences of natural disasters, man-made disasters and other places where the presence of people will be associated with risks to their health and life. But before robots can operate in such difficult conditions, they will have to learn to cope with great physical strain, acquire the necessary dexterity and coordination of movement at a level surpassing that of humans.
In recent years, significant achievements in the above area have mostly concerned four-legged robots, while the development of bipedal humanoid robots is moving much slower for well-known reasons. And so far, such robots are not able to cope with lifting loads, with applying force to something without falling down at the same time. But researchers at the Massachusetts Institute of Technology and the University of Illinois have managed to develop a new way to control balance for remote-controlled bipedal robots, which partially solves the problems mentioned above.
The introduction of new algorithms and the use of new sensors and other technologies is now the only way for bipedal robots to become capable of working in challenging environments. The robot HERMES (Highly Efficient Robotic Mechanisms and Electromechanical Systems) used by the researchers has a fairly simple design, it consists of the torso, standing on two lower limbs. The robot is controlled by a special vest worn by the operator. This vest digitizes information about the movements of a person and transmits this information to the robot, which repeats the movements with maximum accuracy.
In addition, the operator’s vest provides feedback, allowing a person to feel the loads experienced by the robot’s structure when performing any action. This and some other functions of the vest allow a person to fully control the robot, keeping his and the robot’s equilibrium. As tests have shown, this feedback works great in all cases, while walking, running, jumping and doing something.
The system that provides feedback and control to the HERMES robot is capable of self-learning and improving itself using data on human movements, after a while the robot becomes capable of performing movements it could not or did not know how to do before.
And the video below is a confirmation of everything that is said in the previous paragraph.