October 22, 2025
View Full PaperFrom Physics-Informed Machine Learning to Physics-Informed Machine Intelligence: Quo Vadimus?
Key Points
- Physics-informed neural networks enhance system identification through advanced learning techniques.
- Transformations into bio-inspired solutions like spiking neural networks may improve computational efficiency.
- Applications discussed include digital twins in engineering and materials discovery processes.
- Emerging technologies like neuromorphic computing suggest promising directions for future research.
Abstract
This talk reviews physics-informed neural networks and introduces networks that learn functionals and nonlinear operators for system identification. It discusses applications in digital twins, autonomy, and materials discovery, and introduces bio-inspired solutions like spiking neural networks and neuromorphic computing.
Cite This Study
George Em Karniadakis (Mon,) studied this question.