Blind humanoid walks, plays soccer and lifts suitcases with joint encoders only — accepted at Humanoids 2026
Jan_R_Peters · x · 2026-10-09
TU Darmstadt and DFKI's paper Blind Dexterity is accepted at IEEE Humanoids 2026 (Dec 6-9, Santa Clara), with code to be released.
Key points:
- On a Unitree G1 humanoid, policies rely solely on joint-encoder proprioception — no cameras, IMU, force/torque, or tactile sensors
- Demonstrated skills: push-resilient bipedal walking without IMU feedback, active one-foot soccer ball trapping, seeking and lifting a suitcase by its handle, and mounting a randomly placed skateboard
- Core insight: joint-encoder readouts evolving under compliant contact act as a whole-body tactile channel — policies actively probe the environment, making object states decodable from short proprioceptive histories, extracted by compact task-specific state estimators trained separately from the policies
- Authors argue joint encoders plus compliant actuation are already a practical substrate for whole-body dexterous manipulation, on which richer sensing can be layered
More from Embodied
- LLMs play Connect 4 via robot arms in MuJoCo; fable 5.1 wins at $17.40 in API calls — eigenron · 2026-10-09
- Prediction: frontier models as real-time robot policies at 10k-100k tok/s within 1-2 years — ATTlKA · 2026-10-09
- Hugging Face launches Robotic Episodes Viewer for 24k+ LeRobot datasets — mishig25 · 2026-10-09
- Delete object info from observations and PPO learns to search anyway — TU Darmstadt on its Humanoids 2026 paper — Jan_R_Peters · 2026-10-09
- CARE certifies VLA inference speedups up to 10.8x with statistical guarantees — UMCP · 2026-10-09
- Corporate robot bets graded: only one binding order with unit counts found — lukas_m_ziegler · 2026-10-09