SE3 Labs Releases First Open YAM Real-Robot Teleop Dataset: 500 Bimanual Episodes
paigeinsf · x · 2026-11-19
SE3 Labs released real-robodojo-teleop-500 on Hugging Face, the first open RoboDojo teleoperation dataset recorded on a real YAM station (CC-BY-4.0, LeRobot format), so any lab with a YAM setup can train and test on it:
- Scale: 500 successful teleop episodes (6.3h total) across 10 real bimanual manipulation tasks, 50 each
- Hardware: two 6-DoF YAM arms with parallel grippers, three Intel RealSense D405 stereo cameras (top-down, both wrists), 30 fps, six 640×360 video streams (color + infrared per camera)
- Contents: 14-dim state/action vectors (joints + grippers), velocities and torques, per-view camera intrinsics/distortion, natural-language task instructions, and initial object layouts
- Evidence of value: the same data took Dream-1 from 8.5 to 22 on their real-world benchmark
Citing a blog critical of RoboDojo sim (built on NVIDIA Isaac Sim) — "the simulation is not even wrong" — the author argues LLM-powered agents plus minimal real-world calibration is the right path for embodied intelligence: real data is slow and expensive, sim scales, so compute can solve the robot data problem.
More from Embodied
- 1.5 years under review: researchers urge releasing AI papers immediately after submission — carlosdponx · 2026-11-19
- Purple metallic robot dog spotted walking indoors: 'Greetings from the future' — technollama · 2026-11-19
- Sound-based robot destroys tumors without cuts: histotripsy treats 2,000+ patients — TinfoilTricorn · 2026-11-19
- Behind the scenes of a robot dog field expedition at the Matterhorn — ki_ki_ki1 · 2026-11-19
- Reddit user's AI device Dot proactively followed up on his ear condition next morning — HannsGruber · 2026-11-19
- Self-trained pixel RL agent beats Elden Ring boss Margit in simulated gym — Sl33py_4est · 2026-11-19