This robot dog just taught itself to walk

The team’s algorithm, called Dreamer, uses past experiences to build models of the world around them. Dreamer also allows the robot to perform trial-and-error calculations in a computer program as opposed to the real world, by predict potential future outcomes its potential actions. This allows it to learn faster than it can do. Once the robot has learned to walk, it will continue to learn to adapt to unexpected situations, such as resisting being toppled by a stick.

Lerrel Pinto, an associate professor of computer science at New York University who specializes in robotics and machine learning, said: “Teaching robots through trial and error is a difficult problem, made more difficult by the long training time. that such teaching requires. Dreamer shows that deep reinforcement learning and world models can teach robots new skills in a really short amount of time, he said.

Jonathan Hurst, a professor of robotics at Oregon State University, said the findings, which have not been peer-reviewed, make it clear that “reinforcement learning will be a foundational tool in the future of robotic control. “.

Remove the simulator from training robots with many privileges. The algorithm could be useful for teaching robots how to learn real-world skills and adapt to situations like hardware failure, Hafner says — a robot could learn to walk with a disabled motor, for example. Malfunction in one leg.

Stefano Albrecht, an assistant professor of artificial intelligence at the University of Edinburgh, says this approach could also have huge potential for more complex things like autonomous driving, which require sets of tools. complex and expensive simulation. A new generation of reinforcement learning algorithms, Albrecht says, can “quickly catch up with how environments work in the real world.

Source link


Kig News: Update the world's latest breaking news online of the day, breaking news, politics, society today, international mainstream news .Updated news 24/7: Entertainment, the World everyday world. Hot news, images, video clips that are updated quickly and reliably

Related Articles

Back to top button