NTU-CEE Distinguished Seminar Series: Professor Yu WANG
Learn about machine learning empowered, high-fidelity modelling framework for settlement prediction and updating in large reclamation projects.
About the Seminar
Topic: Machine Learning Empowered High-Fidelity Modelling and Updating of Reclamation-Induced Consolidation Settlement
Consolidation settlements in a large reclamation project evolve with time and vary spatially across the 3D domain. Due to subsurface soil heterogeneity and complex interactions between soils and prefabricated vertical drains (PVDs), settlements develop unevenly (e.g., cluster spatially) which may threaten construction safety and serviceability of facilities constructed subsequently on the reclaimed land. Accurate predictions of time-dependent, differential settlements remain challenging and computationally prohibitive when 3D full-scale numerical simulation is performed with thousands of PVDs and spatially variable soil properties estimated from limited site investigation (e.g., cone penetration test, CPT) data obtained at the site. The task becomes even more demanding when these predictions need to be updated using sparse field monitoring data gathered during construction.
This presentation proposes a machine learning empowered, high-fidelity modelling framework for settlement prediction and updating in large reclamation projects and illustrates the proposed methods using a real reclamation project in Hong Kong. It is shown that integration of physics-based numerical simulations with data-driven machine learning methods achieves computationally efficient and physically consistent prediction of high-fidelity, field-scale spatiotemporal settlements.
About the Speaker
Professor Yu WANG, Hong Kong University of Science and Technology
Dr. Yu Wang is a professor of geotechnical engineering at Hong Kong University of Science and Technology, and an elected Fellow of American Society of Civil Engineers (ASCE). His recent research efforts have focused on digital twin of subsurface geo-structures, machine learning in geotechnical engineering, analytics and simulation of geo-data, geotechnical uncertainty, reliability and risk, and risk and resilience assessment of critical infrastructure systems.
His research earned several prestigious international/national awards, including the GEOSNet Award from the Geotechnical Safety Network (GEOSNet) in 2025, the 2023 Thomas A. Middlebrooks Award from ASCE, the 2020 Higher Education Outstanding Scientific Research Output Awards (the First-class Natural Science Award) from the Ministry of Education, China, the First-class Natural Science Award from the Hubei Provincial Government in 2017, and multiple Best Paper Awards from various international journals, e.g., Georisk, Computers and Geotechnics, Canadian Geotechnical Journal, ASCE-ASME Journal of Risk and Uncertainty in Engineering Systems, Part A: Civil Engineering, and the Highly Cited Research Award from the journal of Engineering Geology. Dr Wang has authored/co-authored over 230 journal papers and three books in English. He currently chairs Technical Committee TC309 Machine Learning under the international Society for Soil Mechanics and Geotechnical Engineering (ISSMGE) and serves in editorial boards of several top journals in geotechnical engineering or risk and uncertainty analysis (e.g., Associate Editor for the ASCE Journal of Geotechnical and Geoenvironmental Engineering).
Venue
CEE Seminar Room A (N1-B1b-06)