TEAM PROJECT
Push-robust walking on the Booster K1
A humanoid walking policy trained to recover from pushes, benchmarked in sim and on the real robot.
With Riddhi Chauhan and Hridik Hingorani
In progress
A humanoid walking policy trained in mjlab with PPO to recover from pushes. The benchmark is the largest push the robot can recover from in each direction, measured in simulation and on the real robot.