10th June 2025, Battery Show Europe, Stuttgart– London-based AI startup Monolith is doubling down on its mission to reinvent battery development, presenting its ongoing work in developing its next generation of AI technology for battery R&D. At this year’s Battery Show Europe, the company showcased its agentic battery expert – an AI tool being developed to act as a virtual scientific co-pilot for engineers tackling the toughest challenges in battery R&D.
Think of it as ChatGPT with a PhD in physics, electrochemistry, manufacturing, and a stack of peer-reviewed research papers – domain expertise all tuned specifically for battery development.
Monolith CEO and Founder Dr. Richard Ahlfeld took the stage to present the promise of AI agents in the battery field, with Monolith’s R&D companion that will integrate mid-workflow, helping engineers navigate the complexities of physical and chemical reactions, test planning, and design constraints. “We’ve taken testing from years to months with our software – now we’re working to collapse research cycles further,” Dr. Ahlfeld told a packed crowd of OEMs and suppliers.
While still in development, early test results are promising, showing potential to dramatically reduce R&D costs and accelerate time-to-market – two make-or-break metrics in the fiercely competitive battery landscape.
AI Equipped with Powerful Domain Insights
On stage, Dr. Ahlfeld gave a closer look at how an AI Agent can accelerate slurry formulation and electrode manufacturing for an LFP cathode in a lithium-ion battery. It’s a tough challenge, given the complex non-linear material interactions, performance trade-offs such as energy versus power density, and the difficulty of precisely controlling the manufacturing process. Manufacturing parameters like the mass loading per surface area, porosity, and the flattening and compression process known as calendering are all difficult to control precisely, especially in certain combinations.
Designed to complement Monolith’s current Next Test Recommender technology, the AI Agent is being developed to leverage domain expertise and show you how to use advanced modelling most effectively.
Unlike generic large language models, the AI Agent presented is being fine-tuned with industry-specific knowledge. It’s being trained by Monolith’s world-leading battery experts with vetted, high-impact research papers in the fields of material sciences, electrochemistry and physics, as well as expertise from hundreds of projects delivered for world-leading industry giants. The system can also learn and remember constraints from within the business including manufacturing bottlenecks, and material costs – and adapts testing strategies accordingly.
If your optimal formulation found in the lab is within a range that is impossible or uneconomical to mass-produce, the AI Agent can tell you that this is not your optimal formulation after all. Set to help slash research time and costs, this tailored artificial intelligence technology could be a game-changer for engineers who are currently drowning in complex simulations, slow iterations, and expensive physical testing programmes.
