Alexander Heinlein

Dr. Alexander Heinlein is an Assistant Professor in the Numerical Analysis group at the Delft Institute of Applied Mathematics (DIAM), Faculty of Electrical Engineering, Mathematics & Computer Science (EEMCS), Delft University of Technology (TU Delft). His work bridges scientific computing and machine learning through scientific machine learning (SciML), focusing on domain decomposition methods and multiscale approaches for solving complex partial differential equations on modern hardware like GPUs.

Research interests include:

• Developing high-performance computing algorithms for nonlinear PDEs with applications in fluid-structure interaction and photonic crystals
• Advancing physics-aware machine learning techniques for groundwater heat transport and post-burn contraction prediction
• Creating parallel preconditioners like FROSch for challenging problems in computational mechanics
• Building hybrid numerical-ML frameworks with domain decomposition for multi-physics applications

His recent publications highlight a 128-235x speedup in biomedical simulations through deep operator networks, and keynote presentations on geometric challenges in machine learning-based surrogate models at international conferences like CASML 2024.

Scientific awards include:

• 2025 NWO Open Technology Programme grant for the RAPID-Wind project on offshore wind turbine foundations

Students and collaborations involve:

• Yuhuang Meng (PhD candidate, 2024)
• Jing Zhao (co-supervisor)
• Prof. Jun Zou (Chinese University of Hong Kong collaboration, 2024)

He leads software development for COMSOL and Trilinos extensions while maintaining open-source reproducibility standards.