Manchester Prize winner uses AI to improve battery materials

THE FIRST Manchester Prize has been awarded to Polaron, a startup which uses generative AI to “speed-up” the development of designs for renewable materials, including for electric vehicles and batteries.

Polaron, a spinout from Imperial College London, has been given £1m (US$1.3m) from the UK Department of Science, Innovation and Technology to expand its technology, which uses microstructural data to improve a material’s performance.

The Manchester Prize was launched in 2023 and aims to award funding to companies developing “artificial intelligence for public good”. The first challenge was focused on AI innovations in energy, environment, and infrastructure.

Isaac Squires, CEO of Polaron, says: “We are thrilled to have won the first ever Manchester Prize – it has been an extraordinary team effort. In the last year, we have turned the research we pursued at Imperial College London into a commercial product, using our AI to reduce years of materials development into a matter of days.”

Microstructural data

Polaron uses machine learning algorithms that can process image data of materials to understand and change their features and behaviours. These images are microscale, meaning the algorithms look at features which can only be seen through a microscope.

The technology can identify particle morphologies and potential particle-cracking anomalies in a material, both in a two-dimensional and three-dimensional state. Once identified, the algorithms can generate new microstructures for those materials – changing the surface areas and volume fractions – and predict how they will behave when in use.

Squires says: “We are now working with our first customers in the battery manufacturing sector to apply Polaron to improve the performance of EVs by extending range and reducing charge times.”

Polaron published a paper last year which demonstrated it could predict the optimal manufacturing parameters of lithium-ion batteries, increasing the energy density of a battery cell by 10%. This improvement is equivalent to adding 20 miles of range to a typical electric vehicle.

Squires added: “While this has been our core market to date, Polaron is material agnostic, and we are already bringing our rapid design capabilities to industrial manufacturing more widely, including alloys, composites, and catalysts.”

The next prize

The second round of the Manchester Prize launched in November last year and is focused on the use of AI in developing clean energy systems.

Ten finalists will be announced in Q2 of this year, and like the first iteration, will receive £100,000 in funding. In 2026, one of these finalists will be selected as the ultimate winner and receive the £1m prize.