Calibration-free non-invasive neuromotor input
Wearable, non-surgical sensing (surface EMG, EEG, optical) that reads intended movement or covert gestures well enough to drive a computer continuously, for any user, with no per-person training.
State of the art
Meta's generic sEMG neuromotor interface achieved 0.88 gesture detections per second and 20.9 words per minute handwriting on uncalibrated new users (Nature 2025); productized as the Ray-Ban Display neural band.
- Reads muscle commands, not cognition; silent or purely mental input remains unreachable
- Non-invasive bandwidth ceiling of roughly tens of bits per second
- Signal drift with electrode placement, sweat, fatigue and body composition
Who’s working on it
- Meta Reality Labs (CTRL-labs)Company
Shipped the sEMG Meta Neural Band with Ray-Ban Display glasses in September 2025, trained on data from nearly 200,000 participants
- Patrick KaifoshResearcherReality Labs at Meta
Co-led the generic neuromotor interface paper with uncalibrated cross-user decoding (Nature 2025)
- Thomas ReardonResearcherReality Labs at Meta
CTRL-labs co-founder; co-led the Nature 2025 sEMG paper
- Mudra Wearable ↗Company
Sells EMG-based neural wristbands for controlling Apple and other devices with hand gestures
- Dario FarinaResearcherImperial College London
High-density surface EMG decomposition into motor-unit spike trains
What to invest in
| Kind | Investment | Scale |
|---|---|---|
| research | Scaling cross-subject decoder training data and adding non-muscular channels (eye, optical brain sensing) to raise bandwidth | $$ $10M – $100M |
Depends on
Required by
No other capability builds on this directly.