IIT Delhi develops AI system that can run real laboratory experiments

AILA, the Artificially Intelligent Lab Assistant, autonomously operates complex laboratory instruments, manages experiments, and analyzes results, marking a significant step in AI-driven scientific research and accelerating India’s AI for Science initiative.

Researchers at IIT Delhi, in collaboration with teams from Denmark and Germany, have developed an AI system named AILA that can perform actual laboratory experiments. Unlike conventional AI models such as ChatGPT, which are limited to data analysis or answering queries, AILA can independently control instruments, run experiments, and interpret results. The study, published in Nature Communications, focused on automating atomic force microscopy (AFM), a key technique for studying materials at the microscopic level.

PhD student Indrajeet Mandal, first author of the study, explained, “Tasks that previously took a full day, like adjusting microscope settings, now take just 7–10 minutes with AILA. It significantly accelerates research.” Professors M. Anoop Krishnan and Nitya Nand Goswami highlighted that AILA’s ability to independently manage AFM experiments represents a transformative shift in AI applications for scientific research.

Opportunities and challenges in AI-driven labs

While AILA demonstrates remarkable capabilities, researchers also identified challenges. The AI occasionally deviates from instructions, raising safety concerns. Mandal noted, “It’s like knowing how to drive but navigating real city traffic is different.” The team emphasizes the need for enhanced safety protocols as lab automation expands.

This development aligns with India’s AI for Science initiative and recent funding by the Anushandhan National Research Foundation (ANRF) to promote AI-driven research. Prof. Krishnan believes that systems like AILA can democratize access to advanced research tools, positioning India at the forefront of global AI-powered scientific discovery.