KIBa

The field of artificial intelligence is experiencing a boom in recent years due to recent breakthroughs in computing power, AI techniques and software architectures. Among the many areas affected by this paradigm shift, are applications in battery process engineering are which are diverse and hold a lot of untapped potential, e.g., in predictive modeling, process and formulation optimization, fault detection and uncertainty assessment, intelligent process control, scale transfer, and mechanistic modeling. Given the importance of mechanistic models for design, optimization and transfer of materials and processes along the entire value chain of electrochemical energy storage, reverse engineering strategies (in combination with black-box AI modeling) and hybrid modeling based on coupling mechanistic short-cut models with machine learning methods will play a crucial role in the future.

The goal of the joint project KIBa is to collect reproducible data for the production of battery materials via a fast and simple process design and to use it as a basis for intelligent hybrid modeling by applying machine learning methods. The battery material synthesis process as well as the conditioning by deagglomeration/shredding of particulate calcined battery materials will be focused in order to consider the manufacturing process chain in a product-optimizing way. The hybrid modeling itself is performed using a combination of neural networks and genetic programming. The consortium consists of BASF, Malvern Panalytical, the Fraunhofer Institute, and the Institute for Particle Technology at the Technical University of Braunschweig.

Contact:

Ahmed Eisa
ahmed.eisa(at)tu-braunschweig.de