Beyond Teleop: The Key to Figure 03's Ladder Climbing Lies in Autonomous Exploration
PTrubey · x · 2026-08-02
Following a demo of the Figure 03 robot climbing a ladder fully autonomously, industry experts discussed the underlying training methodologies.
Commenters pointed out that the current demo likely relies on copying teleoperated movements—for instance, the bot moving its head to see where its hands are placed, mirroring the human operator. In contrast, Optimus has a very wide field of view, making such head movements unnecessary. The commenter argues that the ideal training approach should move beyond simply copying teleop data, and instead allow bots to experiment autonomously in simulation and real life (using a safety harness) to figure out solutions on their own.
Related event: Figure F.03 demo highlights autonomous climbing(6 posts)→
More from Embodied
- iPhone Duo preorders sell out in 3 minutes around NYC, deliveries slip to late November — BenBajarin · 2026-10-16
- UMD's PAIR framework bridges perception and action in VLA models via bridge tokens — Kaixi Feng · 2026-10-16
- CI-MSE offline metric hits -0.87 Spearman correlation with real robot task success vs weak raw MSE — chris_j_paxton · 2026-10-16
- List of 15 best robotics researchers on X, from Jim Fan to Chelsea Finn and Fei-Fei Li — FinanceYF5 · 2026-10-16
- Fake Apple listing for dual-screen 'iPhone Duo' jokes it's the best phone for watching your AI agents — AIFlow_ML · 2026-10-16
- Inside Auki's Hong Kong office: a fake grocery store powering robotics data collection — LoganGrasby · 2026-10-16