Chennai, Aug 24: Indian Institute of Technology Madras researchers have developed an Artificial Intelligence platform that can speed up the discovery of sustainable, high-performance materials for industries ranging from EVs, aerospace, renewable energy to marine infrastructure.
The researchers have created one of the world’s largest publicly available databases of advanced metallic alloys by using Large Language Models to automatically extract and organise decades of scientific knowledge scattered across more than 10,000 research papers. The AI-driven framework enables scientists and manufacturers to rapidly identify materials that deliver both superior performance and improved environmental sustainability.
The databases and supporting software have been made freely available through the Alloy Tattvasar platform and GitHub, enabling researchers, startups and industries worldwide to accelerate materials innovation.
The research is expected to significantly reduce the time and cost required to discover next-generation engineering materials. It will help researchers, startups and manufacturing companies avoid repeating expensive experiments while supporting India’s efforts toward greener manufacturing, reduced dependence on critical raw materials and accelerated clean-energy technologies.
Traditional materials discovery is often slowed because valuable experimental data is buried across thousands of journal articles in text, tables and figures, making it difficult for researchers to systematically compare materials or reuse existing knowledge. Manual extraction of such information is labor-intensive, time-consuming and prone to errors.
IIT Madras researchers tackled this challenge by developing an automated AI pipeline that can read scientific literature and extract detailed information on alloy compositions, manufacturing processes, testing conditions. The platform could also review 350+ material properties without needing significant human intervention.
The research was carried out at IIT Madras by Mr. Aravindan Kamatchi Sundaram, Dual Degree student, Mr. Mohit Chakraborty and Sai Mani Kumar Devathi, BS (Data Science) students and research interns, Mr. B. Pabitramohan Prusty, doctoral scholar, under the guidance of Dr. Rohit Batra, Faculty at IIT Madras.
The work received funding from the Anusandhan National Research Foundation, the Defence Research and Development Organisation’s Directorate of Industry and Academia (DRDO-DIA), and Wadhwani School of Data Science and AI at IIT Madras. Computational resources were provided by the Robert Bosch Centre for Data Science and AI (RBCDSAI) at IIT Madras.
The findings were published in Advanced Science (https://doi.org/10.1002/advs.75916), a leading peer-reviewed open-access journal published by Wiley that publishes high-impact research across materials science, engineering, life sciences, medicine and physics.
Elaborating on this research, Dr. Rohit Batra, Assistant Professor, Department of Metallurgical and Materials Engineering, IIT Madras, said,
“Artificial intelligence is transforming how we discover new materials. Instead of spending years manually collecting data from thousands of publications, our framework automatically builds structured databases that can be used to identify sustainable materials much faster. By combining materials performance with environmental and socio-economic indicators, we enable researchers to design alloys that are not only technically superior but also better suited for a sustainable future.”
Dr. Rohit Batra, also the Samar and Jayanti Paul Early Career Chair Professor, added,
“The platform generated two comprehensive databases containing more than 185,000 structured records, making them the largest publicly available multicomponent alloy databases in the world. We also integrated sustainability indicators—including environmental, economic and social parameters—to identify advanced alloys that combine high performance with lower environmental impact.”
Unlike earlier AI-based materials databases that typically extracted fewer than 25 properties, the IIT Madras framework captures over 350 different material properties while simultaneously recording the exact processing and testing conditions under which each measurement was obtained. This enables more accurate scientific comparisons and significantly improves the quality of AI-driven materials design.
The platform also employs Retrieval-Augmented Generation (RAG), allowing the AI system to dynamically retrieve the most relevant examples while extracting information, resulting in state-of-the-art extraction accuracy from both text and tables.
The researchers demonstrated the practical value of the database by identifying promising high-entropy alloy compositions across three major industrial sectors:
ØLightweight structural materials that can reduce fuel consumption and carbon emissions for automotive and aerospace applications.
ØSoft magnetic materials for electric motors, transformers and next-generation electric vehicles.
ØCorrosion-resistant alloys for marine infrastructure, offshore engineering, chemical processing and energy systems.
In each application area, the AI-assisted analysis identified candidate alloys capable of matching or exceeding the performance of conventional materials while offering improved sustainability characteristics.
Going forward, the research team plans to expand the AI system to extract information from figures and microstructural images, incorporate life-cycle assessment methodologies for deeper sustainability analysis, and extend the framework to polymers, ceramics and composite materials, further accelerating the development of environmentally sustainable engineering materials.