Kevin Pouls

Fraunhofer FFB

Day 3 (November 7, 2024)
12:10 p.m.
Ferrum

Mastering the challenges of ramping up a battery cell factory

The ramp-up of a battery cell factory is a complex and cost-intensive task. High reject rates of up to 50% and a long period of more than a year until key figures such as OEE (Overall Equipment Effectiveness) and yield are achieved represent a major risk for companies, as was recently observed at European cell manufacturers. The challenges during the ramp-up are complex and affect all disciplines and departments of a company. These include, for example, immature process definitions, unclear cause-and-effect relationships and a lack of standards for digitalization.

The main challenges in commissioning the production machines arise from the lack of publicly available expert knowledge, the very complex process chain with many known and unknown dependencies and problems in scaling up pilot lines. Many of these challenges are due to the fact that the battery industry is a relatively new industry with strong growth and many new players. Start-ups with little experience are entering the field and established companies are protecting their intellectual property rights by any means possible.

To overcome these challenges, we aim to use an interactive knowledge graph to store and provide information such as cause-effect relationships (CERs). Initially, CERs are defined by process experts and with each new production run, advanced data analysis can support or refute them and identify previously unknown CERs. In addition, the information stored in the graph is machine-readable and can be used together with process and sensor data from the machine in data-driven models to optimize machine parameters and processes.

In addition, a fully digitized factory will help to quickly identify negative events such as increased defect rates or reduced throughput as well as positive developments such as reduced energy usage or increased cell performance by using automated methods such as anomaly detection. This shortens the time between the occurrence of a significant event and the action taken as a result. As a final measure, we document the progress of our ramp-up in detail to record the knowledge gained.

Curriculum vitae

Kevin Pouls, M.Sc., is a research associate at Fraunhofer FFB and focuses on the digitalization of battery cell production. He joined Fraunhofer FFB in 2021 and brings expertise from his studies in mechanical engineering at TU Darmstadt, with a focus on aerospace engineering. Kevin also participated in exchange studies in Lisbon, Portugal.

His research emphasizes data-driven process analysis in battery cell production, including the investigation and validation of cause-and-effect relationships using historical process data. He is involved in the development of digital monitoring strategies for machine conditions and product quality and works on the development of data-driven quality predictions. Kevin plays a central role in the project management and optimization of the ramp-up phase for the FFB PreFab. His responsibilities include coordinating stakeholders, defining KPIs and milestones, and integrating data-driven methods to optimize ramp-up processes.

He has also worked on predicting energy consumption using machine learning models and statistical approaches. This includes data pre-processing, analyzing historical process data and implementing and evaluating various predictive models. His core competencies include: Data analysis and data-driven optimization, ramp-up optimization, quality management and machine learning in battery manufacturing.