SMM News
MMDG-Bench Accepted at NeurIPS 2026
Our new benchmark study on multimodal domain generalization has been accepted at NeurIPS 2026, offering a systematic testbed for assessing progress in multimodal learning under realistic conditions.
Self-supervised learning for robust structural damage identification
Our new study, in collaboration with Kistler, introduces a label-free self-supervised learning framework that distinguishes damage-related changes in structural vibrations from variations caused by changing operational and environmental conditions.
Welcome to our new members
The Chair of Structural Mechanics and Monitoring is pleased to welcome five doctoral researchers and two postdoctoral researchers. Their work spans structural health monitoring, digital twins, scientific machine learning, reduced-order modelling, and computational materials science.
SMM at ISMA2026 in Leuven
Members of the Chair of Structural Mechanics and Monitoring and their collaborators are presenting contributions at ISMA2026 this week, spanning structural health monitoring, wind energy, fatigue assessment, graph-based learning and reduced-order modelling.
New Publication on Utility-Driven Model Selection for Digital Twins
A new study by Konstantinos Vlachas, Antonios Kamariotis and Eleni Chatzi introduces an adaptive framework that enables structural digital twins to select the most suitable computational model according to evolving operational requirements.
PHLieNet: Hypernetwork-Based Forecasting
Led by Pantelis Vlachas, in collaboration with Konstantinos Vlachas and Eleni Chatzi, this work forming part of the TURING Horizon project, presents a new hypernetwork-based approach for adaptive and generalisable forecasting in complex parametric dynamical systems.
Survey on Unsupervised Vision–Language Model Adaptation Accepted in IJCV
A comprehensive survey co-authored by Hao Dong, a member of the Chair of Structural Mechanics and Monitoring at ETHZ and the IMOS Lab at EPFL, has been accepted for publication in the International Journal of Computer Vision. The work introduces a unified taxonomy of methods for adapting vision–language models without labelled data.
BIM2Struct: Connecting lifecycle building data with seismic assessment
A new web-based implementation of BIM2Struct enables the automatic conversion of updated BIM information into analysis-ready nonlinear OpenSees models. Developed by Huangbin Liang as part of the EU-funded INBLANC project, the framework supports lifecycle seismic assessment by linking inspection and monitoring data with structural analysis.
SMM at the 9th World Conference on Structural Control and Monitoring
The Chair of Structural Mechanics and Monitoring at ETH Zürich is pleased to participate in the 9th World Conference on Structural Control and Monitoring (9WCSCM), which opened today at the historic Sheffield City Hall. The conference brings together the international structural engineering community to discuss the latest advances in structural control, structural health monitoring, digital twins, and intelligent infrastructure.
SMM at WCCM–ECCOMAS 2026 in Munich
The Chair of Structural Mechanics and Monitoring is participating in WCCM–ECCOMAS 2026, taking place in Munich from 20 to 24 July 2026. Through a series of presentations, a jointly organised minisymposium and a plenary lecture, members of the Chair and their collaborators will contribute to discussions spanning computational mechanics, scientific machine learning, model calibration, reduced-order modelling and engineering decision support.