How Dunia Innovations Is Industrialising AI‑Driven Materials Discovery

The company describes this approach as a materials acceleration platform, where AI models learn directly from real laboratory results rather than relying solely on simulation or historical datasets.

Why Physics‑Informed AI Matters

Many AI approaches to chemistry rely purely on statistical correlations. Dunia instead combines machine learning with physical and chemical constraints.

This physics‑informed approach helps models generate candidates that are not only theoretically promising but also experimentally feasible, reducing wasted laboratory effort.

According to climate‑tech accelerator Third Derivative, combining physics‑informed AI with robotic experimentation could enable the discovery of electroactive materials up to 10× faster and at roughly one‑third the cost of conventional R&D processes.

Focus on Energy and Catalysis

Dunia’s early research targets materials critical to the energy transition.

Examples include:

• Electrocatalysts for green hydrogen or ammonia production
• Materials for CO₂ conversion into chemicals or fuels
• Advanced catalysts used in industrial decarbonization

The company’s work specifically emphasizes electroactive materials used in power‑to‑X processes—technologies that convert renewable electricity into fuels or chemical feedstocks.

These areas are particularly well suited to automated experimentation because catalyst discovery often requires testing large numbers of material combinations.

Collaboration Through the ASCEND Consortium

One confirmed example of Dunia’s industrial collaboration is the ASCEND catalyst‑discovery consortium in Germany.

The project brings together:

• Dunia Innovations
• BASF
• Siemens Energy
• Helmholtz‑Zentrum Berlin
• The Fritz Haber Institute of the Max Planck Society
• Technical University Berlin

Germany’s federal government has committed €30 million to the initiative to accelerate AI‑driven catalyst development for energy‑intensive industries.

The goal is to combine computational modeling, automated experimentation, and industrial expertise to create more efficient catalysts for the energy transition.

Funding and Deep‑Tech Backers

Dunia was founded in 2022 and has raised roughly €10.6 million (about $11.5 million) to expand its autonomous laboratory platform.

Investors include:

• Elaia
• redalpine
• the European Innovation Council (EIC)
• Pace Ventures
• Kindred Capital
• Deep Science Ventures
• Anglo American

These backers reflect a growing European focus on AI‑driven scientific discovery and climate technologies.

Industrial AI Ecosystem Around the Platform

Although no public documentation confirms a specific partner stack for a large Dunia facility, several major industrial collaborations around AI, robotics, and digital manufacturing illustrate the broader ecosystem in which such labs could operate.

For example:

• Siemens and NVIDIA are collaborating on industrial AI and digital‑twin technologies that simulate complex industrial systems.
• ABB Robotics is integrating NVIDIA Omniverse simulation into RobotStudio to train and deploy physical AI for industrial robotics.

These technologies—digital twins, AI simulation environments, and autonomous robotics—are the same foundational building blocks needed to scale automated laboratories and AI‑driven experimentation.

What Autonomous Labs Could Mean for European Deep Tech

If systems like Dunia’s scale successfully, they could reshape the economics of scientific discovery.

Autonomous labs promise several advantages:

• Faster discovery cycles for new materials
• Larger experimental datasets for AI training
• Reduced R&D costs
• Continuous experimentation rather than batch research projects

That matters because advanced materials underpin major industries including energy, chemicals, electronics, and manufacturing.

Europe has historically been strong in materials science but weaker in large‑scale AI infrastructure. Platforms that integrate AI software with automated physical experimentation could help bridge that gap and keep more deep‑tech innovation within the region.

The Bottom Line

Dunia Innovations represents a new wave of AI‑native scientific infrastructure—laboratories where algorithms, robots, and real‑world experiments operate as a single system.

The company’s core technology—closed‑loop, AI‑driven materials discovery—is well documented and already being applied to catalysts and energy materials. Claims about a future large‑scale Berlin “GigaLab,” however, remain unconfirmed in publicly available sources.

What is clear is the broader shift: the future of materials discovery may not be a traditional lab bench, but an autonomous research factory where AI designs experiments and robots run them around the clock.