Purdue Helps Robots Navigate Unstable Ground
Purdue University researchers are stepping in to teach robots how to navigate the swaying deck of a boat or the uncertainty of a moving walkway — enabling their practical use in more real-world situations. Watch this video to see robots find their sea legs.
“These robots are used for inspection, maintenance, disaster response, surveillance, security, and many other applications,” said Yan Gu , associate professor of mechanical engineering. “But there remains a major challenge to overcome for legged locomotion, and that is reliable operation on a moving surface.”
Transcript
00:00:00 In the real world, we can use legged robots for inspection, maintenance, firefighting, disaster response, emergency response and surveillance security. There are many applications of legged robots. We know Boston Dynamics has been constantly releasing very impressive videos of legged robots. Their robots can do a back flip and jump to high surfaces. These robots, they have amazing performance capabilities when addressing stationary surfaces. But once this floor begins to move, we see that their performance can deteriorate really fast. There remains a major challenge for legged locomotion which is to enable reliable operation on a moving surface: for example, ships, aircraft, trains, commuter rails. Our approach is to explicitly consider the movement of the ground and incorporate the movement into the dynamics model of the robot. So the robot has about 30 joints and they have various sensors on it. On top of it,
00:01:03 it has a LIDAR sensor, and in front of the robot there's a depth sensor which allows the robot to see the terrain right in front of it. The treadmill is very capable: it can perform both rocking motion as well as a sway motion. Also has a split belt, so the two belts can be actuated separately, and we can change their speed separately. Also has force plates, which can measure the ground reaction force on the robot, and it's also integrated with the motion capture cameras. We have 20 motion capture cameras that can capture the true movement of the robot while it walks on the moving surface, so that we can use the movement data to analyze how well our algorithm can enable reliable locomotion. Right now, we are just taking baby steps. We have derived new models, new estimation algorithms, and new control approaches that can make a robot walk reliably on a moving surface.