Robot Dog can learn to get up after a fall

There are movies where we see comic scenes of people pushing robots unable to get up after a fall. This seemed like a good way to get away from our robot masters, at least until now, where researchers have actually taught a robot dog how to get up after a fall.

In the study published in the journal Science Robotics (via Engadget), the researchers formed a network of neurons via a computer simulation. They then practiced it on a robot dog by repeatedly throwing it literally to test their workout, and it worked, as the robot dog finally learned to get up.

According to the researchers, these movements were taught during the simulation, which means that the ability to get up depends entirely on the way the robot has learned, not on specific maneuvers previously programmed or integrated into the robot. The use of simulation is explained by the fact that it is a means of training faster and cheaper.

It also makes it possible to better adapt to various situations that will be useful if never deployed in the real world, where the robots could get up independently of the field or at least in theory.

Deposit Robots. Learn more about AI (artificial intelligence) and science.