BMW and University of Zagreb Use AI to Improve Battery Cell Production

BMW and the University of Zagreb develop AI models to optimize battery cell production, reduce testing and costs. Explore how this impacts EV manufacturing.

Reducing material use and production time by more than 50% is the target BMW is pursuing with a new artificial intelligence project focused on battery cell manufacturing. This is not a theoretical concept: the models are already working with real production and testing data, directly influencing how battery cells are developed and evaluated.

The initiative, called Insight, is a joint effort between BMW Group and the University of Zagreb. It covers the entire battery cell value chain, from electrode production to final testing and recycling. At the center of the project is the Battery Cell Competence Centre (BCCC) in Munich, where BMW develops next-generation high-voltage batteries. Here, AI systems analyze historical test results alongside real-time production data to predict performance and optimize process parameters.

This approach directly addresses one of the most resource-intensive stages of battery development: testing. Traditional testing involves numerous test series, consuming large amounts of time, materials and production capacity. By introducing predictive models, BMW aims to significantly reduce the number of required tests while maintaining or even improving quality. At the same time, this frees up equipment and accelerates development cycles.

Another focus is the final stage of production, when battery cells are stored under controlled conditions after initial charging — a phase often referred to as “quarantine.” This step can last for weeks and requires substantial storage capacity. According to BMW, AI-based analysis may allow for early evaluation of cell performance, potentially reducing or eliminating the need for this stage in the future.

The importance of these improvements becomes clearer in the broader industry context. Battery cells account for up to 40% of the added value of an electric vehicle, and their performance directly affects range, charging speed and overall cost. As a result, even incremental gains in production efficiency can have a significant impact on competitiveness.

The project has been underway since 2024 and brings together BMW engineers with researchers, doctoral candidates and students from the University of Zagreb. The academic side focuses on structuring production data and developing AI models, drawing on expertise in mechanical engineering, electrical engineering and computer science. For students, the collaboration offers hands-on industrial experience, while BMW benefits from access to current research methods.

At the same time, the project fits into BMW’s broader strategy of building in-house battery expertise. Alongside the Munich development center, the company operates a pilot production facility in Parsdorf and a recycling center in Salching, where a direct recycling approach is being implemented. This network allows BMW to connect development, production and material reuse within a single system.

The use of artificial intelligence in battery manufacturing is becoming an industry-wide trend, with similar initiatives underway at research institutions such as Fraunhofer and KIT. Against this backdrop, the BMW–Zagreb collaboration stands out for attempting to integrate AI across the entire production chain rather than applying it to isolated steps.

If the projected efficiency gains can be scaled to industrial levels, they could significantly reshape the economics of battery production — one of the most expensive components of electric vehicles — and contribute to the broader advancement of electric mobility.