AI-driven prediction and inverse design
Start with the performance you need. AI helps identify and design promising materials, narrowing the search.
Unlock the potential of materials to address real-world challenges.
Accelerating materials innovation through research and AI.
Our Core Capabilities
From prediction to validation, we bring science and technology together to accelerate materials discovery and development.
Advanced characterisation
Know how your materials perform. Advanced characterisation validates their properties and helps you identify candidates worth taking further.
Closed-loop autonomous experimentation
Set your research goals. Self-driving labs handle iterative experiments, using each result to guide the next while you focus on what matters.
Cross-domain expertise
Bring your toughest materials challenges to the table. Access a network of specialists in materials science, AI, robotics and engineering across four A*STAR institutes and research partners nationwide.
Our Partners
A global ecosystem of research and industry partners, powered by AI and automation to advance materials innovation.
Four Material Classes, Infinite Possibilities
From lightweight alloys to advanced polymers, we explore four major material classes to discover what's possible, prove what works and move promising materials closer to real-world use.
-
Polymers and organics
Specialty polymers and organic materials designed for new electronics, sustainable packaging and coatings.
-
Metal Alloys
Strong, lightweight and more resilient metals engineered for demanding industrial applications.
-
Ceramics
High-performance materials for extreme conditions, from semiconductor components to water treatment.
-
Composites
Lightweight, high-strength materials for maritime, offshore and high-temperature systems.
←
→
Our Proof
A faster, smarter route to higher-performing materials
UNIVAC Precision Engineering, a Singapore manufacturer of injection-moulded parts for medtech and consumer technology, wanted higher-performance materials to improve productivity and enable complex, multi-material components.
MARVEL-AI used AI to screen candidate alloy compositions against UNIVAC's performance and printability requirements, allowing only the most promising candidates to be 3D printed, tested and characterised.
This reduced the time, manpower and cost of physical experimentation.
The result: the research cycle fell from around six months to four weeks, while a key functional property improved by approximately 200%.
Designing lighter, stronger composites with AI
An advanced textiles R&D organisation wanted lightweight carbon-fibre composites combining strength and thermal performance for next-generation aircraft and UAV applications.
MARVEL-AI combined experimental data, physics-based simulation and generative AI to identify promising material designs before physical testing, reducing the need for lengthy trial-and-error.
The approach cut a typical four-to-six-month fabrication and characterisation cycle to two-to-three weeks, lowering experimental effort and development costs. Physical testing demonstrated the material's performance, giving the partner stronger evidence to decide which designs warrant further investment and scale-up.
Get In Touch
Have a material challenge or technology you would like to validate? Connect with MARVEL-AI at IMRE_industry@a-star.edu.sg to explore Materials-as-a-Service, pilot projects, R&D collaborations or startup engagement opportunities.