Digital Twin-Enabled Operation Time Analyses
This project focuses on developing novel methods to build digital twins of cyber-physical systems to support advanced analyses (e.g., anomaly detection) at the operation time. To this end, we do research in learning digital twins automatically from historical and live data streams with advanced machine learning techniques. Moreover, to develop analyses, we employ various machine learning techniques such as Generative adversarial networks, transfer learning, and curriculum learning.
Funding source
- Internal strategic project. Funding is approximately 1 Million NOK per year.