Difficult Terrain Now Presents No Challenge to ATLAS Humanoid Robot

Tuesday, December 6, 2016

Difficult Terrain Now Presents No Challenge to ATLAS Humanoid Robot

Humanoid Robots

Four-legged Boston Dynamics robots are well known for their ability to maintain balance in uneven terrain. Now, the same feat has been achieved by researchers using the humanoid robot, ATLAS. Check out the astonishing video!

The Institute for Human and Machine Cognition (IHMC) Robotics Lab in Pensacola, Florida has developed a system by which a Boston Dynamics ATLAS robot can navigate across a series of different surfaces of different sizes and geometries.

As the video below shows, even masonry blocks that have been turned on their sides and placed at various angles in a course present little challenge to ATLAS.

"It vastly extends the set of environments a robot can traverse."
"After a step is taken, the robot explores the new contact surface by attempting to shift the center of pressure around the foot," the researchers stated in a paper that details the walking technique. "The available foothold is inferred by the way in which the foot rotates about contact edges and/or by the achieved center of pressure locations on the foot during exploration."

ATLAS walking on difficult terrain

Related articles
The robot has has no prior knowledge of the terrain laid out in front of it. ATLAS tests the surface one foot at a time, shifting its weight around in the process and finally applying it to the supported foothold. The process is supported by angular momentum provided by the robot’s upper body as the humanoid robot executes some of the more difficult steps.

The project is one of a series of bipedal walking projects being undertaken by the institute.

"The ability to walk on unexpected partial footholds greatly increases the robustness of a robot when employed in cluttered environments," state the researchers. "In addition it vastly extends the set of environments a robot can traverse...In the future, we hope to improve the balancing capabilities of our robot by improving our control algorithms and state estimation and by applying the algorithms to robots with higher joint velocity limits."

SOURCE  DRCihmcRobotics

By  33rd SquareEmbed


Post a Comment