Singapore Unveils AI Programme to Accelerate Material Discovery

A new materials laboratory, established through a collaboration between the National University of Singapore and the University of Toronto, is now operational. This facility aims to expedite the creation of recipes for next-generation semiconductor chips and affordable clean hydrogen, moving these technologies closer to real-world application. The laboratory intends to bridge the gap between academic findings and industrial manufacture.

Credit: UNSPLASH

The facility, named the Materials Data Foundry, represents one of eight research projects forming part of Singapore's national AI-for-Science (AI4S) programme. It has secured SGD 10 million in funding. The lab will employ artificial intelligence to conduct thousands of rapid experiments, generating the precise formulations necessary for the mass production of essential future technologies.

These projects mark the first to be publicly introduced under the AI4S programme since its initial announcement. The National Research Foundation supports the overall AI4S initiative with SGD 120 million. Its purpose is to foster the creation of artificial intelligence methods and tools designed to accelerate scientific discoveries across various fields.

Permanent Secretary for National Research and Development, Tan Chorh Chuan, commented on the unveiling of these projects. He stated that the objective is to operate more swiftly and with greater innovation in the domain of discovery. But developing new materials involves complexities beyond theoretical prediction.

The other seven projects within the AI4S portfolio extend into areas such as advanced manufacturing and materials, biomedical and health sciences, as well as aviation and maritime technologies. Mr. Tan also holds positions as chair of the local research agency A*STAR and the Ministry of Health's Office for Healthcare Transformation. He delivered his remarks at the annual scientific meeting, the AI4X-Accelerate Conference.

Brandon Sutherland, the University of Toronto's Director of Research Operations for the Acceleration Consortium, highlighted a challenge within material science. While AI systems are increasingly adept at identifying promising new materials, there is an insufficient amount of data concerning how best to produce these materials industrially. So, the divergence between materials AI can predict and those actually manufacturable continues to expand.

This collaborative effort directly involves the National University of Singapore's Institute for Functional Intelligent Materials and the University of Toronto's Acceleration Consortium. The intention behind the Materials Data Foundry is to address the data deficit by generating empirical, real-world experimental data on materials and their production methods.

Additional organisations participating in the broader AI4S projects include Nanyang Technological University, A*STAR, Imperial College London's first overseas research hub Imperial Global Singapore, and the University of Illinois-affiliated Urbana-Champaign's Illinois Advanced Research Center in Singapore. Furthermore, the University of Cambridge's Cambridge Centre for Advanced Research and Education in Singapore is also involved.

One specific project involves researchers from A*STAR and Imperial Global Singapore. They are developing an AI system to accelerate the search for improved catalysts. Catalysts are substances that hasten chemical reactions and are crucial for the manufacture of cleaner fuels and industrial chemicals.

Discovering new catalysts is a difficult task due to the vast number of potential candidates requiring evaluation. This project aims to reduce the time and expense associated with catalyst discovery by almost five times, achieved through the use of AI to forecast the most suitable options. And another project, involving NUS and Imperial Global Singapore, seeks to ensure the safety and dependability of the increasing volume of AI-generated code.

Researchers on this latter project will build AI tools capable of automatically detecting software bugs, verifying that software operates as intended, and assisting developers and security professionals in auditing critical systems. Other projects include the development of an AI agent to automate the design process for mRNA vaccines, digital twins designed to better forecast farmland conditions in South-east Asia, and an AI system intended to improve disease prediction from full blood-count tests.

Mr. Tan also noted that the AI4S initiative will help cultivate a new cohort of scientists. These individuals will possess proficiency in both scientific disciplines and artificial intelligence. He expressed a desire to position Singapore as a location where AI forms a fundamental component of research endeavours.

• A new laboratory, the Materials Data Foundry, has been established by the National University of Singapore and the University of Toronto.

• This facility will use artificial intelligence to generate production recipes for next-generation semiconductors and clean hydrogen.

• The laboratory, costing SGD 10 million, is one of eight initial projects under Singapore's AI-for-Science programme.

• The AI4S programme, backed by SGD 120 million, aims to accelerate scientific discovery through AI methods and tools.

• Projects also include developing AI for catalyst discovery, ensuring AI-generated code safety, designing mRNA vaccines, and predicting disease from blood tests.

Source: The Straits Times