AI-based battery cell production comes to EAS’s LFP line via KiBa-Pro

KiBa-Pro splits AI-based battery cell production into five work packages, with correlation analysis feeding production data and cell modelling back into each other in a continuous loop

(Image courtesy of EAS Batteries GmbH)

EAS Batteries has started KiBa-Pro, a three-year consortium project applying AI-supported modelling and digitalisation to lithium iron phosphate cell production at its Nordhausen line. Backed by €2.4 million in German federal funding, the project will first digitalise the production line, then use sensor data and a virtual cell model intended to predict service life and performance before physical trials begin. For procurement and quality teams tracking Europe’s LFP capacity build-out, it’s an early signal of how manufacturers plan to compress development cycles while getting ahead of the EU battery passport requirement.

Building the data foundation for AI battery cell production

KiBa-Pro runs from July 2026 to June 2029 under EAS Batteries’ coordination. The first step covers comprehensive digitalisation of EAS’s existing production line, adding sensor technology and modern interfaces so that production, material, and quality data flow into a central infrastructure covering the entire manufacturing process.

Most digitalisation efforts in battery cell chemistry manufacturing stop at dashboards and monitoring. KiBa-Pro instead treats the production line itself as a data source for a downstream AI model, a distinction that will determine whether the project delivers genuine predictive capability or simply better visibility into existing processes.

Modelling service life before cells leave the line

The consortium will build a virtual LFP cell from the collected data. AI methods will analyse correlations between production parameters, including coating thickness, porosity, material throughput, and temperature profiles, and the resulting cell characteristics. The project aims to predict effects on service life, capacity, and performance before physical production trials run.

That sequencing reverses the usual order of process development, where physical trials generate the data that later informs simulation. EAS and its partners are betting that a sufficiently detailed virtual model can catch process errors early, cutting scrap rates and shortening the loop between a production change and its measurable effect on cell quality.

The research consortium powering AI battery cell production

RWTH Aachen’s Center for Ageing, Reliability and Lifetime Prediction of Electrochemical and Power Electronic Systems handles cell characterisation and AI model development, led by professors Dirk Uwe Sauer and Weihan Li. CARL brings an established track record in battery ageing and lifetime prediction research, giving the project a scientifically grounded route from raw production data to usable degradation models.

KIT and Batalyse manage research data management, automated evaluation of measurement data, and digital product passport preparation at cell level. KIT’s Institute for Nanotechnology contributes its Kadi4Mat research data infrastructure, already deployed across multiple materials science and battery projects. Batalyse supplies commercial software for standardising and correlating electrochemical test data at scale, and Omron joins as an associated partner contributing automation and data acquisition expertise.

Regulatory timing behind the EU battery passport

KiBa-Pro’s data infrastructure work lines up closely with the EU battery passport requirement. Under Regulation (EU) 2023/1542, EV and industrial batteries above 2 kWh placed on the EU market need a battery passport from February 2027, and that passport depends on manufacturers holding structured, traceable production and material data. A production line that already collects this data through sensors and central infrastructure starts several steps ahead of that compliance deadline rather than treating it as a separate reporting exercise.

The project also sits within the German Federal Ministry for Research, Technology and Space’s Battery Research Framework Concept, specifically the “Scaling Research and Digitalisation” action field, which funds production processes suitable for series manufacturing. Projektträger Jülich manages the funding on the ministry’s behalf.

What success looks like for European LFP manufacturing

EAS and its partners have set explicit long-term targets of at least 10 percent higher cell service life and at least 15 percent higher production yield. Those figures give the project a measurable bar rather than an open-ended digitalisation promise, one a specialist audience can hold the consortium to as results emerge from 2027 onward.

The timeline calls for reference cell batches and established data platforms during the first year, with the virtual cell and predictive AI models following in later phases before industrial validation closes the project in 2029. Whether the 10 and 15 percent targets hold under production conditions will say as much about the transferability of AI-based process modelling to other cell chemistries as it will about EAS’s own manufacturing gains.

Stay ahead in the electrification revolution. Explore more breakthroughs from leading innovators in electric commercial transport – visit our news page.