London-based AI startup Polaron has secured €6.7 million in fresh funding to tackle one of industry’s hardest problems: understanding how materials behave in the real world — and using that knowledge to design better ones.
The round, led by impact-focused fund Racine2 with participation from Speedinvest, Futurepresent, and several angel investors from the industrial AI ecosystem, will be used to grow Polaron’s engineering team, scale its generative design tools, and meet rising demand from automotive and energy customers.
Founded in 2023 as a spin-out from Imperial College London, Polaron positions itself as the “intelligence layer for materials science.” The company combines generative AI with deep domain expertise to help engineers understand how manufacturing processes shape materials at the microscopic level — and how those structures ultimately determine performance.
“For 150 years, industry has used machines to shape materials,” said CEO and co-founder Isaac Squires. “Now, we are teaching machines to understand them.”
Turning microstructure into intelligence
At the heart of Polaron’s technology is a long-standing scientific principle: the way a material is processed determines its internal structure, and that structure determines how it performs. Properties such as strength, ageing, failure, yield, and lifetime are governed by microscopic features like grains, pores, phases, and defects — features that are visible under the microscope but notoriously difficult to interpret at scale.
While manufacturing itself is highly automated, understanding materials remains slow and manual. Engineers often rely on isolated tools, bespoke scripts, and trial-and-error experimentation to connect processing choices with real-world performance.
Polaron aims to change that by training AI models directly on microscopy images and measured material properties. The result is a platform that can automatically characterise materials, interpret microstructures, and explain why a material behaves the way it does — in minutes rather than weeks.
According to the company, this approach unlocks insights that were previously out of reach, including three-dimensional reconstructions from two-dimensional images and rapid detection of complex microstructural features.
From lab insight to factory floor
Using learned process–structure–performance relationships, Polaron’s system explores vast design spaces to identify optimal material configurations and the manufacturing conditions needed to achieve them. The company says this bridges a critical gap between laboratory research and industrial production, across metals, ceramics, polymers, and composites.
That focus on manufacturability is what attracted investors.
“What impressed us about Polaron is its focus on the point where materials innovation often breaks down: translating scientific insight into manufacturable reality,” said Florian Obst, Principal at Speedinvest’s AI and Infra team. “By grounding AI in real microstructural data and industrial constraints, Polaron is accelerating how advanced materials move from research into production.”
Early impact in electric vehicles
Polaron says its technology is already being used by engineers at global manufacturers, including electric vehicle makers responsible for more than a third of worldwide EV production. In one application, the platform supported the design of new battery electrodes, delivering energy density improvements of over 10%.
With its new funding, Polaron plans to scale these capabilities across more industries and use cases — building what it describes as a foundational intelligence layer for the physical world.
As manufacturers race to develop stronger, lighter, longer-lasting, and more sustainable materials, Polaron is betting that the next leap forward won’t come from more trial and error, but from machines that can finally see — and understand — what materials are made of.
