Menlo Research open-sources Asimov 1 humanoid's locomotion policy and Isaac Lab training code
freelerobot · x · 2026-10-07
Menlo Research has open-sourced the locomotion policy checkpoint for its Asimov 1 humanoid robot along with the full training framework (repo: cyclotron, BSD-3-Clause, 118 stars).
- Built on Isaac Lab, the code ships reward definitions, hardware-tuned actuator configurations, and domain randomization settings
- Lets developers train and evaluate locomotion policies in simulation before deploying to real hardware — "you can start training a humanoid before you own one"
- The foundation is meant to be adapted to hardware changes and new gait experiments; the robot itself is also available for order
More from Embodied
- Seeed engineer to give ROSCon UK 2026 deep dive on ROS2Control interface for STS3215 servos — kamathsblog · 2026-10-07
- Self-Compensating VLA Boosts Robot Arm Task Success by Over 30 Points — POSTECH · 2026-10-07
- Agility's Digit 5 launches at $150K bill of materials as $300M order unlocks skill by skill — xmercury_one · 2026-10-07
- Humanoid robot shrugs off kicks and shoves in live ROBOPARTY balance demo — CurieuxExplorer · 2026-10-07
- Asimov 1: an open-source, repairable humanoid robot with 5kg-payload arms for $499 deposit — freelerobot · 2026-10-07
- Indian builders make offline AI hardware that listens and remembers without apps — alysha_lobo · 2026-10-07