A new partnership between a Copenhagen-based deep-tech company and Singapore's National University of Singapore aims to create a shared laboratory where artificial intelligence and atomic-scale manufacturing work together to discover and test new materials faster than traditional methods allow. The collaboration, announced in April 2026, represents a significant shift in how researchers approach materials science by automating the entire workflow from digital design to physical synthesis and testing .

What Is This New Materials Discovery Foundry?

ATLANT 3D and the Institute for Functional Intelligent Materials (I-FIM) at the National University of Singapore have signed a Memorandum of Understanding to establish a shared, AI-driven materials discovery foundry within NUS's robotic laboratory at CREATE, which stands for Campus for Research Excellence and Technological Enterprise . The facility will integrate ATLANT 3D's Direct Atomic Layer Processing (DALP) technology and NANOFABRICATOR platform as core synthesis tools within NUS I-FIM's robotic materials hub.

The foundry is designed to support Singapore's AI for Science programme under the National Research Foundation, creating what both organizations envision as a reference model for self-driving laboratories that connect atomic-scale manufacturing with AI-enabled materials discovery . Rather than having researchers manually design, synthesize, and test materials one at a time, the automated system will handle these steps in coordinated workflows, generating data that feeds back into AI models to suggest new material combinations worth exploring.

Which Research Areas Will Benefit Most?

The partnership targets several high-impact application areas where faster material discovery could accelerate innovation. These focus areas include 2D materials and nanoelectronics, advanced semiconductor packaging, quantum materials, catalytic materials discovery, and photonics . Each of these fields faces similar challenges: researchers need to test thousands of potential material combinations to find ones with the right properties, and traditional trial-and-error approaches consume months or years of laboratory time.

The ability to fabricate and test novel material combinations with atomic precision while producing device-relevant structures could fundamentally change how researchers approach these challenges. Rather than designing materials in theory and hoping they work in practice, scientists will be able to rapidly iterate through possibilities, learning from each experimental result to inform the next round of synthesis and testing.

How Will the AI-Driven Workflow Operate?

• Automated Synthesis: ATLANT 3D's NANOFABRICATOR platform provides programmable, atom-by-atom control of matter, allowing the system to synthesize thin-film materials and devices at the atomic scale without manual intervention between design and fabrication.
• Continuous Experimentation: The robotic laboratory performs automated testing and characterization of synthesized materials, generating experimental data that would normally require weeks of manual work by human researchers.
• AI-Enabled Feedback Loops: Machine learning algorithms analyze experimental results and suggest new material combinations to test, creating a closed loop where each experiment informs the next design iteration and accelerates discovery.
• Ecosystem Access: The foundry will be accessible to researchers across participating programs in academia, industry, and government, creating a shared resource that amplifies the impact of the discovery infrastructure.

This integration of atomic-scale manufacturing with AI represents a departure from how materials science has traditionally operated. Rather than treating synthesis and discovery as separate steps, the foundry treats them as parts of a unified, automated workflow where data flows continuously between design, fabrication, and testing .

"An AI-driven materials discovery foundry in Singapore represents our vision for what becomes possible when atomic-scale manufacturing and AI are deeply integrated. Singapore is one of the world's most forward-thinking environments for deep-tech and AI-for-science materials discovery, and we are excited to pursue this with NUS," said an ATLANT 3D representative.

ATLANT 3D

Professor Sir Kostya S. Novoselov, who leads the Institute for Functional Intelligent Materials at NUS, emphasized the practical benefits of combining these capabilities. "We are excited to partner with ATLANT 3D in harnessing new capabilities in atomic-scale fabrication. The ability to fabricate and test novel material combinations with atomic precision, while producing device-relevant structures, would accelerate experimental studies and open new lines of inquiry. I look forward to integrating advanced fabrication platforms into our AI-driven workflows at NUS I-FIM," he stated .

"The ability to fabricate and test novel material combinations with atomic precision, while producing device-relevant structures, would accelerate experimental studies and open new lines of inquiry," noted Professor Sir Kostya S. Novoselov.

Professor Sir Kostya S. Novoselov, Institute for Functional Intelligent Materials, NUS

Why Singapore and Why Now?

Singapore's positioning as a hub for deep-tech research and AI-for-science initiatives made it an attractive location for this collaboration. The city-state has invested heavily in research infrastructure and has positioned itself as a leader in advanced manufacturing and artificial intelligence applications to scientific discovery. By establishing this foundry in Singapore, both ATLANT 3D and NUS I-FIM are betting that the combination of world-class research talent, government support through the National Research Foundation, and access to cutting-edge fabrication technology will create a model that other institutions can replicate .

The timing also reflects broader momentum in the field. Materials science has become a bottleneck in many technology sectors, from semiconductors to battery development to quantum computing. Traditional discovery methods cannot keep pace with the computational power now available to simulate and predict material properties. By automating the experimental validation step and closing the loop with AI, this foundry addresses a real constraint in the innovation pipeline.

What Makes This Different From Other AI Materials Research?

While AI has been applied to materials science for several years, most applications focus on the computational side: using machine learning to predict which materials might have desired properties. This foundry takes a different approach by integrating AI prediction with automated physical synthesis and testing. The system doesn't just suggest that a material might work; it actually makes the material, tests it, and learns from the results in a continuous cycle .

This closed-loop approach is more powerful than prediction alone because it grounds AI models in experimental reality. Materials that look promising in simulations sometimes behave differently in practice due to factors that are difficult to model computationally. By continuously validating AI predictions against real experimental data, the system becomes smarter and more reliable over time.

The foundry also emphasizes scalability and accessibility. Rather than being a proprietary research tool available only to ATLANT 3D or NUS, it is designed as a shared infrastructure that researchers from academia, industry, and government can access. This ecosystem approach could accelerate materials discovery across multiple sectors simultaneously, from semiconductors to catalysts to photonic devices.

The partnership between ATLANT 3D and NUS I-FIM represents a significant step toward making AI-driven materials discovery a practical reality rather than a theoretical possibility. If successful, the Singapore foundry could serve as a blueprint for similar facilities worldwide, fundamentally changing how scientists approach the challenge of discovering new materials with the properties needed for next-generation technologies.