Google DeepMind Unveils Whole‑Body Robot Control Model to Advance Autonomous Machines

DeepMind announced a new artificial intelligence model that can command the full range of movements of a physical robot, extending control beyond isolated limbs to coordinated whole‑body actions, and positioning the system as a central focus of its latest robotics research release.

The model is documented on DeepMind’s official research website, where it appears among recent publications detailing advances in embodied AI, robot manipulation, and whole‑body control, and is also listed in the DeepMind models catalog for public reference.

Researchers say the system learns to generate joint‑level commands that synchronize arms, legs, and torso, enabling smoother and more adaptable robot behavior in complex tasks, and the approach leverages large‑scale simulation data to train the policy.

Industry observers note this development marks a step toward more versatile autonomous machines, potentially influencing future robotics applications in manufacturing, logistics, and service environments, and they anticipate broader adoption as the technology matures.