ANYmal Robot Hiking: A Quadrupedal Adventure

Recent Advances in Legged Robotics

The progress achieved by robotics researchers in the field of legged robots has been remarkably rapid in recent years. A notable example is Oregon State University’s successful completion of a 5K run with their robot, Cassie, just last July.

This week, a team at ETH Zurich revealed their own advancements, demonstrating the extended physical capabilities of their ANYmal robot through rigorous mountain climbing training.

Mount Etzel Ascent

The researchers trained the quadrupedal robot – which shares design similarities with the robots developed by Boston Dynamics – to ascend Mount Etzel. This peak reaches a height of 1,098 meters (3,600 feet) above sea level.

The robot completed the 120-meter vertical climb in just 31 minutes. This is four minutes quicker than the average time recorded for human hikers undertaking the same route.

Impressively, the ANYmal robot navigated the terrain without any stumbles or missteps, a feat not always matched by human climbers.

Control Scheme and Technology

This accomplishment was enabled by a novel control scheme, recently detailed in a scientific publication. It integrates imaging technology with tactile feedback systems.

This combination enhances the robot’s ability to operate effectively on challenging, uneven surfaces and in conditions of limited visibility – common obstacles encountered during hiking.

The robot utilizes this feedback to assess the appropriate level of caution required for each step it takes.

Initial testing of the technology was conducted within a virtual simulation before the robot’s ascent of Mount Etzel.

Researcher Insights

“The robot has learned to integrate visual perception of its surroundings with proprioception – its sense of touch – derived from direct leg contact,” explains lead researcher Marco Hutter.

“This integration allows it to navigate rough terrain with greater speed, efficiency, and, crucially, robustness,” Hutter added.

Future applications for ANYmal include deployment in environments deemed too hazardous for human exploration or inaccessible to conventional robotic systems.

Implications for Robotics

Maintaining stable footing on irregular terrain is a critical challenge that roboticists are actively addressing in the development of quadrupedal robots.

The implementation of real-time processing capabilities like those demonstrated by ANYmal holds significant potential for robots deployed in dangerous scenarios, minimizing risks to human personnel.