Watch Argo in action on YouTube!
Argo is an AI home robot designed to bring autonomous navigation, conversational AI, and entertainment to everyday environments. Positioned between today’s static assistants and tomorrow’s humanoid robots, Argo paves the way for smarter home robotics!
As Argo progresses, the need for advanced robotics technologies like manipulation or locomotion becomes critical. To encourage collaboration and innovation, we’ve decided to make key components of our research open-source.
Explore our repositories and deep-dive tutorials:
- Footsteps Planning: Algorithms for planning footstep trajectories for a biped given a global path.
- Locomotion Policy for Quadrupeds: Reinforcement learning-based policy for quadrupedal robots, enabling adaptive and smooth motion in Genesis simulator using PPO.
- EKF State Estimation for Humanoids: EKF model to estimate humanoid state (pose + velocities) based on floating base dynamics.
- Foundation Models for Manipulation: Overview about state-of-art imitation learning techniques for robotic manipulation, enabling generalization across diverse tasks and environments.
- Kinematics & Dynamics: Library to compute kinematics, inverse kinematics and dynamics for N-dof robotics systems based on Denavit-Hartenberg or URDF convention.
- wake-word detection: wake-word detection module for conversational AI.