ESPINN

Numerical simulations and later also data-driven methods such as neural networks (NNs) have developed into indispensable tools in engineering sciences and beyond in recent decades. However, the respective methodological weaknesses (high computing time or enormous amounts of simulation or measurement data required) limit their applicability, the solution efficiency and thus the product improvement capabilities for multi-billion markets such as micro-, nano- and power electronics.

Within the framework of ESPINN, a hybrid approach, so-called Physics-Informed Neural Networks (PINNs), will be used to develop simulators that solve physical and chemical processes - supported by a few measurement data - on an atomistic to continuous scale several orders of magnitude faster than previously used numerical methods. The methodological focus is on the implementation of PINN-based interaction potentials including their quantum mechanical effects, the resolution of 4D tensors (3D geometry and time) for parameterized diffusion processes, as well as the development of a concept-based understanding of PINNs with different NN architectures. AI and application knowledge is thus combined with explainability and transferred into four industry-relevant software solutions. These include a methodologically largely generic molecular dynamics solver, two specific PINN simulators for the semiconductor processes silicidation and photoresist processing, and a PINN analysis tool for estimating learning content and accuracy.

The ESPINN project is funded by the Fraunhofer PREPARE program and is being carried out in close cooperation with Fraunhofer IISB and Fraunhofer SCAI. In particular, HHI is contributing its extensive expertise in the field of explainable artificial intelligence (XAI) to this innovative project. In this way, PINNs can be improved in terms of their reliability and robustness, which on the one hand comprehensively secures the AI-generated simulations and on the other hand also enables reliable use in critical industrial applications.

• Fraunhofer IISB (Project Lead)
• Fraunhofer SCAI