anymal

Collection of quadrupedal robots configured to work in CHAMP development framework

A simulation environment made in Gazebo to estimate lidar odometry performances under different lidar specifications

A legged locomotion project

Personal Website of Marko Bjelonic

Path planning with local estimations

GPU-parallel RL training workspace for NVIDIA Isaac Lab: ANYmal locomotion + Franka reach with RSL-RL PPO, scripts, and 8GB-VRAM defaults.

Mujoco simulation of ANYmal C with the purpose to validate and evaluate the haptiquad library, capable of estimating ground reaction forces and base rench

Waypoint-based trajectory tracking for quadruped robots using Quadruped-PyMPC

Deep-RL hierarchical controller enabling an ANYmal B quadruped to self-right from any fall and resume walking (TRPO, MuJoCo). MVA Robotics project.

Rosbag interface with the purpose to validate and evaluate the haptiquad library, capable of estimating ground reaction forces

A PyTorch implementation of the TD3 (Twin Delayed Deep Deterministic Policy Gradient) algorithm applied to quadruped robot locomotion control. Trained on two environments of increasing complexity: the MuJoCo Ant (8 degrees of freedom) and ANYbotics' ANYmal (12 degrees of freedom) in the Gym Jiminy physics simulation.

PPO ANYmal C with variable impedance control across 6 terrains in Isaac Sim