IIT Delhi Unveils AI Agent That Mimics Human Scientific Research

• IIT Delhi's AILA AI agent runs lab experiments autonomously
• Operates Atomic Force Microscope; cuts tasks to 7-10 mins
• Paradigm shift: AI now conducts real science independently

Researchers at the Indian Institute of Technology (IIT) Delhi, in collaboration with scientists from Denmark and Germany, have developed  AILA (Artificially Intelligent Lab Assistant) , an advanced AI agent capable of autonomously conducting real-world scientific experiments, marking a paradigm shift from AI as a digital tool to a physical laboratory performer.

This major innovation, which is the basis for an automatic operation of an ultrahigh, resolution imaging device by AILA, has been described in the journal Nature Communications in an article entitled "Evaluating large language model agents for automation of atomic force microscopy." AILA can now command an AFM (Atomic Force Microscope), which is a very sensitive nanoscale imaging tool, completely on its own.

The AI agent manages the intricacies of the situation: changing parameters on the fly from the feedback it gets, performing experiments, gathering data, analyzing results, and deciding what to do next without any kind of supervision, phenomena that researchers have been known to take years to acquire."

Also Read:  IIT Delhi Launches 6-Month Online Executive Brand Program

The study's first author PhD scholar Indrajeet Mandal highlighted that the work of setting up a microscope to take high, resolution pictures, which used to take a whole day, is now done in just 7, 10 minutes. "AILA is really a great help for me in my daily lab work and it makes my research progress at a much faster rate, " he said.

The supervisors Prof. N.M. Anoop Krishnan (Department of Civil Engineering and Yardi School of AI) and Prof. Nitya Nand Gosvami (Department of Materials Science and Engineering) pointed out the jump: "Before, AI was only a tool that could help you write about science. Now it can actually do sciencefiguring out experiments, performing them on a real instrument, getting data, and understanding the results."

This improvement is in line with the India's AI for Science initiative which is meant to speed up discovery in the areas of energy storage, materials for environment, and advanced manufacturing, etc. AILA, by doing away with the repetitive but very precise parts of lab work, is the harbinger of faster innovation cycles, less human error, and more people getting access to advanced research instruments which in turn is making Indian scientists global leaders in the field of autonomous experimental science.