Berkeley's TANGO Enables Whole-Body Humanoid Navigation

UC Berkeley's TANGO, presented at CoRL 2026, is a whole-body VLA model that predicts 29-DoF joint actions directly from egocentric RGB input for humanoid navigation. Trained purely in simulation, it achieves zero-shot transfer to real robots in cluttered 3D environments.

2026-09-24 ~ 2026-09-24 · 2 related posts

• TANGO: Sim-Only Trained Whole-Body VLA Gives Humanoids Zero-Shot Navigation on Unitree G1 — chris_j_paxton · 2026-09-24
• TANGO: Whole-Body VLA Maps RGB to 29-DoF Actions for Humanoid Navigation, Trained Fully in Simulation — berkeley_ai · 2026-09-24