Legged robotic systems use ground contact and the reaction forces they provide to achieve agile locomotion. However, uncertainty combined with the discontinuous nature of contact can lead to failures in real-world environments with unexpected height variations, such as rocky hills or curbs. To dynamically travel through extreme terrain, this work introduces the use of proprioception to estimate and respond to unknown hybrid events and altitude changes, and a two-degree-of-freedom tail to improve control independent of contact. Simulation results of unforeseen height differences show that our method can stabilize locomotion at height differences of up to 1.5 times the leg length.

This work was presented at the 2022 ICRA Workshop on Legged Robots.
Extensive summary: http://www.andrew.cmu.edu/user/amj1/papers/Proprioception_and_Tail_Control_ICRA_WS.pdf
Workshop site: https://leggedrobots.org/index.html

The code is implemented as part of the Quad-SDK, open source full-stack quadruped robot controller:
https://github.com/robomechanics/quad-sdk

Please take the opportunity to connect and share this video with your friends and family if you find it helpful.