Computational Mathematics for Predictive Digital Twins (PreDiTwin)

In recent years, remarkable progress in mathematics, process-based modeling, data science, and sensor technology has opened up exciting possibilities in creating Digital Twins for real-world systems. These advancements hold great potential for enhancing our predictive accuracy and control over various systems including forest environments, smart industry, building technology, energy, health, and climate change.


Our research

The research group Computational Mathematics for Predictive Digital Twins (PreDiTwin) focuses on the development of the theoretical foundation of digital twins. In particular in the fields:
  • Multi-scale modeling and homogenization of deterministic and stochastic systems
  • Random matrices and mathematical statistics 
  • Process-based and numerical model reduction
  • Reinforcement learning for operations research
  • Numerical methods for deterministic and stochastic partial differential equations 


External members

Stefano Giani, Durham University

Luka Grubisic, University of Zagreb