Imagine a battery that doesn't just store electricity, but can switch between creating power from fuel and using power to make fuel. To make this work, we need special materials that act like high-performance sponges for ions. Instead of guessing and testing millions of combinations in a lab, this project uses AI to predict the perfect recipe for these materials.
Current fuel cell electrodes rely on critical or toxic materials and the process of finding better alternatives is slow and relies on trial-and-error. This hinders the mass adoption of reversible energy devices for grid stability.
The project is building an AI-enabled discovery pipeline combining multi-scale modelling, spectroscopy, and deep learning to predict high-performance perovskite electrode compositions.
If you are a grid-scale energy storage provider dealing with intermittent wind and solar power — this project developed a method to find materials for reversible devices that store excess energy as chemical fuel. This allows for better integration of renewables into the electrical grid.
If you are a fuel cell manufacturer dealing with toxic or critical raw materials in electrodes — this project developed a knowledge-driven way to discover perovskite oxides that are free or reduced in critical content. This lowers supply chain risk and environmental impact.
If you are a capacitor developer dealing with inefficient carbon-based materials — this project developed a generalized experimental approach to optimize carbon materials for supercapacitors. This improves the energy density of high-power storage devices.
Based on available project data, specific pricing is not provided, but the project aims to reduce costs by eliminating toxic and critical raw materials.
The project is currently developing simplified testing protocols and tools designed to be operable by industrial stakeholders, but it is still in the discovery and modelling phase.
Based on available project data, the project emphasizes open science sharing and harmonised documentation to ensure interoperability and usability.
The project runs from 2023-01-01 to 2026-12-31, suggesting that validated materials and protocols will be available toward the end of 2026.
The project focuses on the 'composition-structure-activity-performance' relation, providing a rational design for electrode materials that can be integrated into SOFC and SOEC devices.
The consortium is heavily research-oriented with 6 research institutes and 3 universities, balanced by 3 industrial partners (all SMEs). With a 25% industry ratio and a budget of over 5 million EUR, the project is designed to bridge the gap between deep theoretical modelling (DFT) and practical industrial application through a 12-partner network across 7 countries.
Contact CNRS (France) regarding the KNOWSKITE-X project coordination.
Contact us to connect with the KNOWSKITE-X consortium for early access to AI-driven material descriptors.
