Mimic pairs a robotic hand with an exoskeleton to capture human motion directly
Olivier__OG · x · 2026-07-21
Mimic pairs a robotic hand with a wearable exoskeleton for direct human-to-robot transfer
Mimic is building the M1 robotic hand for manufacturing work, and the key idea is how it learns.
- Instead of relying on video-to-robot transfer, it uses a U1 wearable exoskeleton with the same structure as the M1 hand.
- A human wears the exoskeleton, performs the task, and the system captures real finger motion.
- That motion transfers more directly to the robot hand.
Technical details highlighted in the post:
- Tendon-driven design
- Motors in the forearm, cables running through the fingers
- 15 degrees of freedom and 21 joints
- Force sensing and automatic grip adjustment when an object is about to slip
The post frames this as a practical step toward production-grade physical AI, not just another robotics demo.
More from Embodied
- EPO improves 3D foundation models by aligning edges, poses, and depth — ducha_aiki · 2026-07-21
- Applied Intuition launches Dana, an agentic platform for physical AI apps — a16z · 2026-07-21
- Applied Intuition launches Dana for physical AI, alongside a16z interview on one billion machines — a16z · 2026-07-21
- Hearing Aids Embrace AI: Auracast and OmegaAI Reshape Personal Audio Broadcasting — BrandonSawalich · 2026-07-21
- Applied Intuition says physical AI is moving from concept to product with its new Dana platform — a16z · 2026-07-21
- Argus improves indoor panoramic 3D reconstruction with covisibility and geometry transformers — ducha_aiki · 2026-07-21