In his post, Brooks talks that “too close” to The amount of robotics The humanoid when he fell a few years ago. He did not dare to approach one during a walk since then. Brooks notes that even in advertising videos from humanoid companies, people are never shown close to the movement of humanoid robots, if they are not separated by furniture, and even then robots are dragged only minimally.
This safety problem goes beyond random waterfalls. In order for humanoids to play their promised role in healthcare and factory, they need certification to work in zones divided with people. Current walking mechanisms make such certification almost impossible in accordance with existing security standards in most parts of the world.
Humanoid robot Apollo.
Credit: Google
Brooks predicts that for 15 years there will indeed be many robots called “humanoids”, performing various tasks. But by irony, they will look not like today's two -legged cars. They will have wheels instead of legs, a different number of hands and specialized sensors that do not have similarities with human eyes. Some will have cameras in their hands or look down from their funds. The definition of “humanoids” will change, just as “flying cars” now means electric helicopters, and not aircraft capable of the road, and “cars with independent driving” mean vehicles with remote human monitors, and not really autonomous systems.
Brooks claims that billions that currently invest in the fact that today's tough humanoids are only studying dexterity for vision only disappears. Academic researchers achieve more progress in systems, which include feedback from sensory MIT approach The use of a glove that conveys the sensations between human operators and the hands of the robot. But even these achievements remain far from a comprehensive sensation of touch, which ensures human dexterity.
Today, few people spend their days near humanoid robots, but Brooks's rule is a practical warning about problems, ahead of someone who has built these cars for decades. The gap between advertising videos and a deployed reality remains large, measured not only over the years, but also in fundamental unsolved problems of physics, sounding and safety.
