Robot Dog Does Parkour, Climbs Rubble

A robot dog created by ETH Zurich researchers has learned parkour, climbing over obstacles and navigating difficult terrain in a new video. 

The team said they used machine learning to teach the quadruped robot, dubbed ANYmal, new navigation and mobility capabilities. 

ANYmal used a trial-and-error method of learning, adapting its response to obstacles depending on past experience. The robot uses its camera, as well as an in-built artificial neural network, to identify the terrain and select an appropriate response. 

“Before the project started, several of my research colleagues thought that legged robots had already reached the limits of their development potential,” said study co-author Nikita Rudin. “But I had a different opinion. In fact, I was sure that a lot more could be done with the mechanics of legged robots.”

In tests, ANYmal showed proficiency in negotiating obstacles using dynamic maneuvers such as jumping and hopping, as well as navigating rubble and other tricky terrain found on building sites or in disaster areas.

For navigating the latter terrain type, the researchers used machine learning to teach ANYmal how to identify gaps and recesses in piles of rubble, as well as how to identify the correct maneuvers to respond to the situation. 

Related:Robotic Inspection Tech to Automate Offshore Sites

“As a result,” the team said, “the quadrupedal robot is now better at gaining a sure footing on slippery surfaces or unstable boulders,

“ANYmal is soon also to be deployed on building sites or anywhere that is too dangerous for people – for instance, to inspect a collapsed house in a disaster area.”