by turbo

August 11, 2015

Robotics studies are moving along at such a quick pace these days that it is often tough to identify the significant turning points in the development of this modern technology. One significant step forward is a new robot being developed at MIT called Hermes. Like other robots, it’s completely made of inorganic materials and electronics, but its actions as well as its reflexes are human as it is controlled by a human operator. In order to acheive this, the robot feeds back relevant information concerning its equilibrium and spacial positioning in real-time so the human controller has the ability to readjust his/her feet and physical body accordingly. That human-generated data in turn is then relayed in real-time back to the robot, thus keeping it balanced. This technique addresses one of the major issues for humanoid robots ie. keeping upright on both feet is no mean feat. Simply put, the reflexes of the machine to a changing surface or from poke in the rear are essentially human. In addition, the mechanical graspers of Hermes are also special. These three-fingered claws can easily pick up objects, crush tin cans or even pour coffee. Utilizing joysticks designed into the human exoskeleton, the human controller is able to grab items in front of the robot as well as interact with the environment with more precision compared to a machine of this type would normally have. There are plenty of benefits to having a human controller command a robot, as the machine can take advantage of hundreds of years of human brain evolution. In return, the human can take advantage of the robot’s resilience by being able to stay out dangerous situations for example, after a natural catastrophe or travel to another world. To top it all off, the robot also comes with is a 3D headset that makes it possible for the human controller to see whatever in front of Hermes. Whether attempting to maneuver through a set of enclosures or trying to pick things off the floor, it gives the operator a greater feeling of awareness of the robot’s environment.