On Wednesday, October 28, a seminar will be held in Fassò Room (Campus Leonardo, Building 4A, floor 1), starting at 10:15 CET.
The seminar will be given by Prof. Yashar Eftekhar Azam, University of New Hampshire, and is titled “Machine Learning for Bridge Damage Identification under Uncertainty: Contact and Drive-By Sensing”.
Abstract
Developing scalable and practically implementable tools for bridge structural health monitoring requires reliable damage identification under uncertain loading, imperfect modeling, and noisy measurements. This seminar presents a research program that combines machine learning with full-scale experimental validation to advance bridge condition assessment through contact and drive-by sensing.
The presentation covers four complementary components: novelty detection for identifying departures from healthy structural behavior, damage classification for distinguishing among structural condition states, drive-by sensing for assessing bridges through vehicle-mounted measurements, and transfer learning for connecting numerical simulations with field observations. Particular attention is given to traffic load variability, the influence of model uncertainty on damage indices, and the effects of measurement noise on identification performance. Full-scale experiments provide a basis for evaluating these methods under realistic conditions, while transfer learning addresses the limited availability of labeled damage data and discrepancies between simulated and measured responses. Throughout, the emphasis is on translating methodological advances into reliable, implementable tools that can support bridge monitoring at scale.
Speaker’s bio
Yashar Eftekhar Azam is an Associate Professor of Civil and Environmental Engineering and Ocean Engineering at the University of New Hampshire (UNH). He received his Ph.D. with distinction from Politecnico di Milano in 2012 and subsequently held research appointments at ETH Zurich, the University of Thessaly, and the University of Nebraska–Lincoln before joining UNH in 2020. His research integrates structural dynamics, Bayesian estimation, and machine learning to develop practical tools for structural health monitoring and digital twinning of civil and offshore infrastructure. His work combines methodological development with full-scale field experiments and laboratory testing, with particular emphasis on identifying structural states, loads, and damage under uncertainty. Applications include bridge monitoring through contact and drive-by sensing, floating structures, and offshore renewable energy systems. His recent research also explores quantum and hybrid quantum–classical methods for engineering inference. He received the U.S. National Science Foundation CAREER Award in 2025 and UNH’s J. Brent Loy Innovator of the Year Award in 2026.
The seminar can also be followed remotely at the following link: https://politecnicomilano.webex.com/meet/stefano.mariani
