Advanced Intelligence for Materials

Accelerating Discovery with DFT & MLIP

We combine the accuracy of physical theory with the speed of AI in the first stage of materials discovery to revolutionarily expand the exploration space.

• DFT & MLIP: By combining quantum mechanics-based first-principles calculations (DFT) with Machine Learning Interatomic Potentials (MLIP) that accelerate them, we perform large-scale atomic simulations and elucidate complex reaction mechanisms.
• Virtual Screening: Through ultra-fast Virtual Screening that explores vast chemical spaces, we precisely predict promising candidate materials and optimal properties before experimental stages.

Realizing Closed-Loop Discovery Beyond Human Intuition

We automate repetitive and resource-intensive experimental processes and implement a Self-Driving Lab where AI designs the next experiments autonomously.

• Closed-Loop Discovery: We connect the entire research process—from material design, synthesis, analysis, to evaluation—in an automated loop.
• AI-Guided Optimization: AI-based optimization algorithms learn from experimental data in real-time, minimizing the number of experiments needed.
• Beyond Human Intuition: Through data-driven exploration, we discover new chemical compositions and innovative materials that transcend researcher experience and human intuition.

Bridging the Gap via Process Modeling & Multiscale Analysis

We research data-driven scale-up technologies so that basic research achievements in the laboratory can translate to mass production processes in actual industrial settings.

• Multiscale Analysis: We perform Multiscale Analysis spanning from microscopic properties to macroscopic performance, elucidating the correlations between materials and systems.
• Process Modeling: Through Process Modeling and digital twins that learn the relationships between data and process variables, we predict performance in manufacturing processes and reduce trial and error.