Duke's open-source 31-DoF humanoid boosts visible-reachable workspace from 38% to 97% with actuated eyes
LexiLove · x · 2026-09-25
Duke unveiled Duke Humanoid V2, a fully open-source 31-DoF humanoid with two RGB-D cameras that can move and look independently, arguing robots shouldn't blindly copy human anatomy.
The team introduced the Visible-Reachable Workspace metric, measuring where a robot can both reach and perceive while manipulating.
Key results:
- Actuating the cameras raised visible-reachable workspace from 38% to 97%
- A second independently controlled camera lifted concurrent two-region coverage from 0.45 to 0.95; a third only pushed it to 0.97
- In two-target manipulation tests, the dual-camera setup finished tasks 17% faster while using 19% less mechanical energy
More from Embodied
- Google dev imagines orchestrating AI agents via smart glasses: tap their shoulder, watch their screen — jason_mayes · 2026-09-25
- Climbing analysis in 3D with iPhone LiDAR: open-source SAM 3.1 + ViTPose demos — dosco · 2026-09-25
- Menlo open-sources Asimov 1 humanoid locomotion policy and Isaac Lab training code — freelerobot · 2026-09-25
- Human-Robot Dialogue workshop at IROS 2026: NVIDIA, MIT, Georgia Tech speakers lined up — dhadfieldmenell · 2026-09-25
- YuXiang attends NSF FRR robotics meeting, presenting navigation work at IROS workshops — YuXiang_IRVL · 2026-09-25
- OmniEcho: Spatial Audio Benchmark and Model for Embodied Agents — PKU-VaLuE-Lab · 2026-09-25