AI for Transdisciplinary Science (AI4TranS)

Scientific discovery increasingly depends on our ability to integrate vast amounts of data, knowledge, and expertise across disciplines. The AI for Transdisciplinary Science (AI4TranS) pillar develops advanced AI methods that accelerate scientific discovery by bridging disciplinary boundaries. We create generalisable, structure-aware, and compute-efficient AI models that integrate heterogeneous scientific data, domain knowledge, and physical principles across multiple modalities, fidelities, and scales.

By bringing together fundamental AI research and diverse scientific expertise, we develop transferable methodologies, strengthen cross-disciplinary research capabilities, and address strategically important scientific challenges aligned with Singapore's research and innovation priorities.

Advancing scientific discovery requires AI methods that can learn from diverse data, incorporate scientific knowledge, and reason across complex systems. Our research develops the foundational AI capabilities needed to accelerate discovery, improve scientific understanding, and enable breakthroughs across disciplines through the following areas:

Research Focus

Foundation Models

Developing foundation models that capture scientific knowledge and support reasoning, prediction, and discovery across diverse scientific domains.

Graph and Geometric Learning

Advancing AI methods that learn from structured data and complex relationships, enabling the modelling of molecules, materials, biological systems, and other scientific phenomena.

Multimodal and Multi-fidelity Learning

Integrating heterogeneous data sources, modalities, and levels of fidelity to generate richer scientific understanding and improve predictive performance.

Generative AI and Simulation

Creating generative models and AI-driven simulation methods that accelerate hypothesis generation, design exploration, and scientific innovation.

Spatiotemporal Modelling

Developing AI techniques that uncover patterns, interactions, and dynamics across space and time in complex scientific systems.

Uncertainty Quantification and Robust AI

Building reliable and trustworthy AI systems that quantify uncertainty, incorporate scientific constraints, and support robust decision-making in real-world scientific applications.

AI for Scientific Workflows

Developing intelligent systems that assist researchers throughout the scientific lifecycle, including knowledge extraction, experiment planning, automated analysis, simulation, and discovery workflows, helping to accelerate and scale scientific research.

Application

We apply our research to accelerate discovery across a broad range of scientific domains, including:

Materials Science, Surfaces and Catalysis

Medtech

Developing AI systems that support clinical decision-making, longitudinal human-AI interaction, and safety-critical recommendations, enabling more effective and trustworthy healthcare delivery.

Hub of the Future

Creating intelligent systems that operate within complex physical-social environments, supporting mobility, infrastructure, public services, and adaptive urban operations through anticipation, planning, and coordination.

Manufacturing

Enabling autonomous planning, process optimisation, and intelligent operations across manufacturing environments, improving efficiency, resilience, and adaptability while maintaining safety and reliability.

Collaborate With Us

Addressing today's scientific challenges requires close collaboration between AI researchers and domain experts. The AI4TranS pillar serves as a methodological and theoretical bridge across research groups, departments, and organisations, enabling partnerships on frontier scientific problems that require advances in both AI and science.

We welcome collaborations through joint research projects, industry partnerships, postdoctoral fellowships, and graduate research programmes. Together, we aim to develop the next generation of AI-driven scientific discovery.