ML-101 Ergonomic Leader Arm

Open-source leader arm designed for teleoperation and VLA (Vision-Language-Action) data collection.

Key Features

• Improved ergonomics: 50mm joint offset (J2, J3, J4) for better operator clearance
• Coaxial trigger alignment: Natural grip, reduced wrist strain
• SO-101 compatible: Same kinematics, works with existing electronics
• Open Hardware: MIT License, OSHWA certified

Specifications

• DOF: 5+1 (same as SO-101)
• Actuators: Feetech ST3215 series (SO-101 compatible)
• Materials: 3D printed components

Compatibility

• Kinematics: SO-101 compatible
• Servos: Dynamixel MX-28, MX-64, XM430, XM540
• Electronics: Standard SO-101 controller boards
• Software: Compatible with SO-101 control stacks

Improvements over SO-101

• 50mm joint offset for operator clearance
• Coaxial trigger-finger alignment
• Optimized cable routing
• [Add other specific improvements]

Use Cases

• VLA model data collection
• Robot learning from demonstration
• Teleoperation research
• Imitation learning datasets

Community

• Issues: Report bugs or request features via GitHub Issues
• Discussions: Share your builds in GitHub Discussions
• Discord: Discord Server

Citation

If you use ML-101 in your research, please cite:

@misc{ml101,
  author = {Chatzikonstantinou, Ioannis},
  title = {ML-101: Ergonomic Leader Arm for VLA Research},
  year = {2026},
  publisher = {GitHub},
  url = {https://github.com/motionlayer/ML-101}
}

License

This project is licensed under the MIT License - see LICENSE file.

Open Source Hardware Certification: OSHWA UID [PENDING/XXXXXX]

About MotionLayer

ML-101 is developed by MotionLayer, building motion infrastructure for embodied AI.

Other projects:

• Tinymovr motor controllers
• ROS 2 integration

---

Questions? Open an issue or email: info@motionlayer.com