WCCM-ECCOMAS 2026 — World Congress on Computational Mechanics

A very intense week with the RSUQ Chair at WCCM-ECCOMAS 2026.

The conference offered a unique opportunity to showcase our latest research at the intersection of uncertainty quantification, reliability, stochastic modelling, Bayesian inference, and computational mechanics. We were proud to contribute five presentations spanning several aspects of our work.

Contributions

🔹 S. Marelli, S. Schär, C. Nardin & B. Sudret — Reliability analysis on dynamical systems using manifold-autoregressive surrogate models
The keynote delivered by Stefano explored mNARX+, a data-driven framework for modelling complex dynamical systems with evolving internal states and damage, enabling efficient reliability analysis of systems exhibiting non-stationary behaviour.

🔹 C. Nardin, S. Marelli, M. Broccardo & B. Sudret — A Continuous Markov-Chain Perspective on Joint Seismic Reliability and Resilience Assessment
In this work, I introduced continuous-time Markov chains into the PEER-PBEE framework, providing a unified perspective for modelling both damage and recovery processes and enabling consistent assessment of seismic reliability and resilience.

🔹 A. Herr, S. Marelli & B. Sudret — Introducing Stochastic Emulators for Hierarchical Bayesian Inference of Population Parameters
Anna presented a new approach based on stochastic emulators that directly approximates population-level responses, reducing the computational burden of hierarchical Bayesian inference for expensive and stochastic forward models.

🔹 S. Marelli, S. Schär & B. Sudret — UQLab – An Open-Source Tool for Uncertainty Quantification with Dynamical Systems
Stefano highlighted the latest capabilities of UQLab, including its support for time-series data, dynamical systems, and advanced data-driven surrogate models such as PC-NARX, F-NARX and mNARX+.

🔹 Moustapha — Reliability-Based Design Optimization with Stochastic Emulators
Moustapha’s work on RBDO demonstrated how stochastic emulators can provide an efficient alternative to classical deterministic surrogates, particularly for high-dimensional uncertainty spaces and intrinsically stochastic simulators.

Together, these contributions reflect the broader research vision of the Chair: developing efficient, scalable and reliable computational methods for uncertainty quantification and risk-informed decision making, while bridging fundamental methodological developments with challenging engineering applications.

Kudos to all the authors and presenters for their excellent work, and many thanks to the WCCM-ECCOMAS 2026 organizers for bringing together such a vibrant scientific community. It was a pleasure to meet old and new colleagues, discuss new ideas, and see the UQ and computational mechanics communities come together at such an enormous congress.

Abstract

C. Nardin, S. Marelli, M. Broccardo & B. Sudret —
A Continuous Markov-Chain Perspective on Joint Seismic Reliability and Resilience Assessment
WCCM-ECCOMAS 2026, Munich, July 2026.

⬇ Download Abstract (PDF)