NVIDIA GEAR Lab Releases ENPIRE for Autonomous Robot Experimentation
NVIDIA's GEAR Lab, in collaboration with CMU and UC Berkeley, has unveiled ENPIRE (Agentic Robot Policy Self-Improvement in the Real World). This system allows AI coding agents to autonomously design, run, and iterate real-world robot experiments without human intervention. ENPIRE achieved a 99% success rate on high-precision tasks like pin insertion and GPU installation. The project also discovered a 'physical scaling law,' showing that deploying more parallel robot setups significantly accelerates research progress, and agents could share successful strategies.
ENPIRE represents a breakthrough in robotics research automation, drastically reducing the manual labor involved in running robot experiments and accelerating the development of robust robotic policies through autonomous learning and parallel experimentation. The open-source nature can foster widespread adoption.
Learn one new AI thing every day.
Daily Deck sends you seven plain-English cards like this every morning. Free.
Start free