Tech

Sony AI's Ace Robot Outpaces Pro Athletes in Real-World Tasks via Reinforcement Learning

April 30, 2026 · From the archive, written in the Professor's voice before our October 2026 redesign from ai.sony

What happened

Sony AI unveiled Ace, an autonomous robotic system that surpasses professional athletes in dynamic athletic benchmarks. Published in Nature, it leverages advanced force-torque sensors and model-based reinforcement learning for superior agility. Ace excels in tasks like ball catching and obstacle navigation in unpredictable environments.

Why it matters

This highlights integration of multimodal sensors with RL for real-world robotics, changing your approach to embodied AI. Move beyond simulated training to hybrid real-sim loops for robust policies. It redefines workflows, emphasizing sensor fusion for human-level physical intelligence.

Who's doing it

Sony AI's robotics division deployed Ace prototypes that beat human pros by 20% in speed-accuracy metrics on custom athletic tracks. Their Nature publication validates zero-shot generalization to novel environments.

Try it

1. Download Stable Baselines3 via pip install stable-baselines3 in a Python setup with Gymnasium.
2. Set up a robotic sim environment like Isaac Gym, adding force-torque sensor noise via custom observations.
3. Train with PPO algorithm for 1e6 steps; expect policies outperforming baselines by 15-30% in agility tasks. Tutorial at https://stable-baselines3.readthedocs.io/.

Read the original at ai.sony