https://j.people.com.cn/n3/2026/1020/c95952-20510224.html

The ‘Robot Scientist 2.2’ system, developed at the University of Science and Technology of China, serves as a next-generation infrastructure for AI-driven materials science. It integrates ‘bold hypothesis generation’ via large language models, ‘rigorous verification’ through smaller language models, and precise experimentation conducted by robotic systems. This enables the autonomous creation of new materials and the discovery of new knowledge, providing a powerful new tool for scientific research.

The goal of this system is to distill the long-accumulated experience of human scientists into actionable skills, allowing more research and development teams to build upon existing scientific expertise. The “Robot Scientist 2.2” system currently integrates 110 advanced research robots, 196 AI-driven chemistry workstations, 508 command sets for research robots, 3,200 chemical experiment templates, and 310 scientific AI models. The system is already being applied in fields such as biomedicine, green agriculture, new energy, and space exploration.