Candidate ID: DC11
Topic: Uncertainty in multi-fidelity digital twins for infrastructure management
Beneficiary: TUD
My name is Ahmed and I am a PhD candidate at TU Delft within the FOURIER project. I hold a B.Sc. in Civil Engineering from Mansoura University (Egypt) and a joint M.Sc. in Mathematical Modelling from Universität Hamburg and the University of L’Aquila. Before joining TU Delft, I worked as a machine learning researcher in collaboration with INGV and the Gran Sasso Science Institute (GSSI), where I developed deep learning methods for seismic applications using convolutional neural networks and vision transformers. My work focused on domain-specific data augmentation, uncertainty quantification, and explainability, reflecting a broader interest in building models that remain calibrated and trustworthy under noisy, sparse, or biased data.
In my PhD research, I develop AI-based surrogate modelling approaches to enable probabilistic, time- and state-continuous digital twins for infrastructure systems. My work emphasizes representation learning and uncertainty-aware modelling to extract robust information from heterogeneous and incomplete data, integrating real-time monitoring with multi-fidelity information sources and supporting decision-making for infrastructure inspection and maintenance.
I joined the FOURIER project for its strong interdisciplinary environment that connects AI, uncertainty-aware risk modelling, and engineering practice, as well as for its clear focus on translating advanced methods into deployable tools for resilient infrastructure management.
