Materials Discovery

Use AI to accelerate discovery and design of novel materials with optimized properties and sustainability.

Business impact

• Time to market — AI reduces discovery and validation time, speeding product launch timelines
• R&D efficiency — Automated workflows and simulations increase throughput and reduce manual effort
• Innovation rate — Generative AI enables exploration of novel materials beyond traditional methods
• Cost reduction — Lower experimental and computational costs through AI-driven predictions and automation
• Sustainability metrics — Early integration of lifecycle assessment reduces environmental impact of materials

Data requirements

• Experimental synthesis data (Structured) — Used to train models and validate AI-generated material candidates
• Computational simulations (Numeric) — Provide physics-based property predictions for candidate materials
• Scientific literature and patents (Text) — Source of synthesis recipes and material properties for model training
• High-throughput screening results (Numeric) — Feed real-time data into closed-loop discovery systems
• Lifecycle assessment databases (Structured) — Enable sustainability impact predictions integrated into design

AI methods and techniques

• Generative AI — Generate novel material candidates conditioned on target properties
• Predictive AI — Predict material properties and synthesis outcomes from candidate structures
• Agentic AI — Autonomously plan and execute experimental workflows in self-driving labs
• Symbolic AI — Incorporate expert knowledge and rules for retrosynthesis and validation

AI models and model families

DiffSyn, Graph Neural Networks, Large Language Models, Deep Learning Simulators, Physics-informed Neural Networks, Generative Diffusion Models

Industries

Real-world evidence

14 documented case studies on record.

Companies using this: Asahi Kasei Corp, Cusp AI, Dunia Innovations, Lawrence Berkeley National Lab, Massachusetts Institute Technology MIT, Microsoft, Microsoft Research, Murata Manufacturing, North Carolina State University, Orbital Materials, SES S.A., Sepion Technologies, University Liverpool, Zhejiang University.