ICASP15 hosts 42 Mini Symposia organised by leading researchers and practitioners from around the world. Click any entry below to read the scope and meet its organisers.
Organisers and Affiliations:
Prof Eun Jeong Cha1, Prof. Abdollah Shafieezadeh2, Prof. Michele Barbato3, Prof. Alexandros Taflanidis4
1University of Illinois at Urbana-Champaign, 2Ohio State University, 3University of California, Davis, 4University of Notre Dame
Summary:
Changing climate conditions have long been recognized as an important source of increasing risk associated with climate-driven hazards. These evolving climate conditions alter the environment to which built infrastructure is exposed, posing a major threat to its safety and functioning. To properly manage the risk to built infrastructure under future climate conditions, it is crucial to accurately assess the impacts of climate change on natural hazards and the physical vulnerability of infrastructure. Additionally, developing appropriate mitigation and adaptation strategies to enhance resilience against such hazards, has emerged as a key national and international priority.
This Mini-Symposium will bring together researchers and practicing engineers to share the latest advances in climate change research for civil infrastructure and building adaptation. Topics of interest include, but are not limited to: (1) modeling methods for nonstationary risk assessment of infrastructure assets and systems, including physics-based, probabilistic, or data-driven modeling approaches for the impact on the patterns (i.e., intensity or frequency) of extreme weather or climate events (e.g., extreme precipitations, winds, heatwaves, droughts, wildfires), structural loads and structural vulnerability (e.g., through accelerated material degradation); (2) assessment of the climate change impact on the risk (failure and damage) and performance (serviceability, energy consumption, etc.) of civil infrastructure in different regions and at different temporal and spatial scales; (3) development of civil infrastructure adaptation strategies and implementation supports, such as the development of sustainable materials, investigation of structural design adaptation, performance-based engineering implementation, energy consumption reduction, and consideration of flexibility; and (4) development of support tools for decision-making under complex and uncertain conditions, including risk communication, consideration of risk perception, acceptability of adaptation strategies, deep uncertainty, and development of decision frameworks. Contributions that address conceptual, theoretical, computational, field-based, experimental, and/or methodological developments are all welcome.
Organisers and Affiliations:
Dr Xizhong Cui1, Dr Chao Feng2, Dr. Katsuichiro Goda3
1Harbin Institute of Technology (Shenzhen), 2Changan University, 3Western University
Summary:
Earthquake risk assessment increasingly requires integrated probabilistic models that carry uncertainty consistently from fault rupture and seismic hazard to site-specific and spatially distributed ground-motion demands. This mini symposium will bring together recent advances in seismic hazard assessment and in the modelling, simulation and validation of earthquake ground motions and surface fault displacement.
Contributions are invited on probabilistic seismic hazard analysis; source, path and site-effect modelling; physics-based, stochastic, data-driven and hybrid ground motion time history models; uncertainty quantification; and the generation, selection and modification of acceleration time histories for engineering analysis.
The symposium will also address spatially varying ground motions for extended structures, infrastructure networks and regional risk analysis. Topics include but not limit to: coherence and correlation models, time-frequency and nonstationary dependence, cross-component and inter-site dependence, conditional simulation, regional ground motion fields, and spectrum-compatible multi-support inputs. A complementary focus will be surface fault displacement caused by fault rupture, including empirical, physics-informed and simulation-based displacement models; principal and distributed faulting; uncertainty treatment; and probabilistic fault displacement hazard analysis (PFDHA).
By connecting seismic hazard, ground-motion time histories, spatial variability and surface fault displacement within a unified probabilistic framework, the symposium will encourage exchange among researchers and practitioners in seismology, geotechnical and structural earthquake engineering, and risk and reliability. It is expected to identify emerging methods, validation needs and pathways for incorporating advanced simulations into performance-based assessment, infrastructure resilience and engineering decision-making.
Organisers and Affiliations:
Prof Tong Zhou1, Prof Xujia Zhu2, Prof Jize Zhang3, Prof Chul-woo Kim4
1Southeast University, 2Université Paris-Saclay, 3The Hong Kong University of Science and Technology, 4Kyoto University
Summary:
Uncertainty is ubiquitous in computational and real-world engineering systems, stemming from incomplete information, inherent variability, measurement noise, and model-form imperfections. Uncertainty quantification (UQ) provides a systematic framework for characterizing, propagating, and reducing these uncertainties and plays a fundamental role in reliability assessment, sensitivity analysis, inverse problems, structural health monitoring, and reliability-based design optimization. However, the increasing complexity and dimensionality of engineering systems, together with computationally expensive simulations and rapidly growing monitoring data, present significant challenges to conventional UQ methodologies.
Recent advances in artificial intelligence (AI), particularly machine learning and deep learning, offer new opportunities to overcome these challenges. AI-enabled surrogate models, data-driven inference, physics-informed learning, and adaptive learning strategies can substantially enhance the efficiency, scalability, and flexibility of UQ while enabling new solutions for complex engineering problems. At the same time, the integration of AI with UQ raises fundamental challenges concerning predictive accuracy, uncertainty representation, interpretability, generalization, physical consistency, and robustness.
This mini-symposium aims to bring together researchers working on the development and application of AI-enabled UQ methods for complex engineering systems. Contributions addressing both methodological advances and engineering applications are particularly welcome. Topics of interest include, but are not limited to, AI-based surrogate modeling, efficient reliability and sensitivity analysis, data-driven and physics-informed UQ, time-dependent and dynamic reliability analysis, reliability-based design optimization, uncertainty-aware structural health monitoring, and large-scale engineering applications.
The symposium will provide a forum for presenting recent advances, discussing emerging challenges, and fostering interdisciplinary collaboration at the intersection of artificial intelligence, uncertainty quantification, and engineering.
Organisers and Affiliations:
Dr Liuyun Xu1, Prof Min Li2, Dr Isabela D. Rodrigues3, Prof André T. Beck3, Prof Bruno Sudret4, Prof Seymour M.J. Spence5
1University of Texas at Arlington, 2Rensselaer Polytechnic Institute, 3University of São Paulo, 4ETH Zürich, 5University of Michigan
Summary:
Estimating rare-event probabilities, such as system failure, structural collapse, and other low-probability, high-consequence events, is essential to risk-informed decision-making and resilience planning in civil engineering. Conventional sampling methods, such as direct Monte Carlo simulation, often become computationally prohibitive when high-fidelity models, high-dimensional uncertainties, complex failure domains, or small failure probabilities are involved. To address these challenges, advanced sampling and uncertainty-reduction strategies have been explored to more efficiently select informative samples and identify critical regions of uncertain parameter spaces. Additionally, rapidly developing artificial intelligence (AI) methods offer new opportunities to accelerate rare-event estimation through adaptive learning, surrogate modeling, reduced-order representations, and data-driven exploration of failure domains.
This mini-symposium will bring together researchers and practitioners working at the intersection of structural reliability, risk analysis, uncertainty quantification, advanced sampling and simulation, and AI. Contributions are encouraged on topics including, but not limited to:
- Risk, reliability, and sensitivity analysis of complex systems subject to high-dimensional uncertainty.
- Advanced variance-reduction and rare-event simulation methods, including importance sampling, subset simulation, stratified sampling, and related techniques.
- Active learning and adaptive design strategies for efficient failure-domain characterization.
- Multi-fidelity and multilevel approaches for efficient probabilistic analysis.
- Linear and nonlinear reduced-order modeling approaches.
- Emerging AI and data-driven surrogate and emulation methods for accelerating uncertainty quantification.
- Physics-informed and physics-guided machine learning for probabilistic engineering analyses.
- Characterization, quantification, and propagation of epistemic and aleatory uncertainty.
- Bayesian inference and updating for uncertainty reduction and model calibration.
- Hybrid methods that integrate AI, advanced sampling, and physics-based computational models.
By connecting methodological advances with demanding engineering applications, the mini-symposium aims to deepen the collective understanding of rare-event probability estimation and uncertainty propagation. Contributions addressing theoretical developments, innovative computational methods, benchmark studies, and real-world applications across civil and infrastructure engineering are particularly welcome.
Organisers and Affiliations:
Prof Iason Papaioannou1, Prof Oindrila Kanjilal2, Prof Costas Papadimitriou3, Prof Daniel Straub4, Prof Jie Zhang5
1National Technical University of Athens, 2Paris-Saclay University, 3University of Thessaly, 4Technical University of Munich, 5Tongji University
Summary:
Design and assessment of engineering systems is often based on numerical models of physical systems and involved processes. The parameters of the numerical models are determined by combining information from different sources such as direct measurements of the parameters or the system behavior, expert knowledge, categorical data and information from literature. In probability theory, the process of combining information to learn model parameters is formalized in the concept of Bayesian updating. Thereby, the prior probability distribution of the model parameters is updated with new data to a posterior distribution. The derived distribution can be further used for forward uncertainty propagation and reliability assessment of the system performance. This mini-symposium aims to attract papers that address either methodological developments or novel applications on Bayesian analysis of numerical models of structural and geotechnical systems. Individual relevant topics include: Markov chain Monte Carlo methods; sequential Montel Carlo methods; Taylor series approximations to the posterior; conjugate priors and Gibbs sampling; approximate Bayesian computation; structural identification; reliability updating; updating in the presence of spatial/time variability; updating of meta-models; applications that investigate the influence of prior considerations on the analysis results; definition of the likelihood function; representation of model errors; optimal experimental design.
Organisers and Affiliations:
Dr Ji-Eun Byun1, Prof Daniel Straub2, Prof Marco Broccardo3
1Imperial College London, 2TU München, 3University of Trento
Summary:
In recent years, Bayesian networks (BNs) have become increasingly popular for probabilistic modeling and computations in applications of engineering reliability, risk, and resilience analysis. They are a powerful model for complex problems that involve many dependent random variables and evolving information. By extending BNs to influence diagrams, they facilitate the optimization of risk mitigation actions.
This minisymposium aims at bringing together researchers that have been utilizing BNs, with the goal of exchanging experiences from the different areas of engineering to which BNs have been applied. These include the probabilistic modeling and risk management, at component or system level, for natural and man-made hazards, infrastructure vulnerability, deterioration, dynamics, and others. Contributions can be on developments of the BN methodology targeted towards risk and reliability analysis, as well as applications of the methodology for probabilistic modeling, inference and decision optimization.
It is expected that the minisymposium will provide an overview of the state-of-the-art and will help to identify areas for further developments and applications of BNs in engineering risk analysis.
Organisers and Affiliations:
Dr Leandro Iannacone1, Prof John Leander2, Dr Ivar Björnsson1, Ms Paulina Chojnicka1
1Lund University, 2KTH Royal Institute of Technology
Summary:
In building and civil engineering, the design and assessment of structures are largely governed by codes and standards. Because no structure is absolutely safe, the goal is to provide guidelines for designers to make sure that the structure attains an acceptable level of safety, i.e., that the probability of failure in a given time span is below a specific level. While fully probabilistic methods aimed at formally quantifying structural reliability are becoming increasingly common, most design methodologies still rely on safety factors and characteristic values, appropriately calibrated based on the results of such probabilistic analyses. The resulting semi-probabilistic formulations are well established in engineering practice and familiar to practitioners, who may apply the prescribed values without necessarily being exposed to the probabilistic considerations and complexities underlying their calibration.
As structural codes continue to evolve, the principles and methods underlying their calibration are receiving renewed attention. Different normative frameworks may adopt different assumptions, target reliability levels, and calibration procedures, while facing the common challenge of translating increasingly sophisticated probabilistic knowledge into provisions that remain transparent and practical for everyday engineering design.
The purpose of this mini-symposium is to bring together experts to discuss and compare current practices and upcoming developments in code calibration. Its international scope will provide an opportunity to investigate different normative contexts and how they approach the delicate balance between mathematical rigour and practical application.
Topics of interest encompass, but are not limited to:
- Novel probabilistic methods for the calibration of safety factors and characteristic values
- Full-blown probabilistic methods and how to incorporate them into norms and design practices
- Probabilistic modeling of loads and resistances on structures
- Real-life experiences and challenges in the implementation of new-generation codes
- Prospects in the development of the next generation of codes
- Target values for reliability and safety requirements
Organisers and Affiliations:
Dr Mattia Anghileri1, Prof Fabio Biondini1, Prof Mark Stewart2
1Politecnico di Milano, 2University of Technology Sydney
Summary:
Structures and infrastructure systems must ensure adequate long-term functionality, safety, reliability, and resilience, while maintaining acceptable levels of risk, despite a continuously changing environment. Global warming and climate change are expected to have substantial impact and potentially catastrophic effects on the long-term durability and performance of structures and infrastructure facilities by exacerbating the occurrence and impacts of hazardous natural events and accelerating structural deterioration processes. Adapting the built environment to climate change is imperative to safeguard people, property, and the economy from its detrimental effects, while fostering sustainable and equitable communities prepared for future challenges. At the same time, there is a strong incentive to design new structures that are resilient to non-stationary conditions and deterioration processes, and to support decision-makers in understanding the associated uncertainties and the implications of climate change on their daily decision-making processes. This Mini-Symposium is intended to discuss emerging research needs and advances associated with incorporating the effects of climate change into the reliability, risk, and resilience assessment of structures and infrastructure systems, while promoting awareness and identifying effective mitigation and adaptation strategies for a climate-resilient built environment.
Organisers and Affiliations:
Prof Emilio Bastidas-Arteaga1, Prof Sebastian Thöns2, Prof Alan O'Connor3, Dr Leandro Iannacone2, Dr Alex Sixie Cao3
1La Rochelle University, 2Lund University, 3Trinity College Dublin, 4Nanyang Technological University
Summary:
The increasing impacts of climate change pose significant challenges to the resilience and functionality of buildings and infrastructure systems worldwide. This mini-symposium focuses on the application of probabilistic approaches and decision analysis methods to address these challenges and assess the effects of climate change on infrastructure performance. Specifically, it explores the evolving impacts of climate-induced changes, such as shifts in loading scenarios, load frequencies, and extreme weather events, on the structural integrity and long-term reliability of infrastructure. It also examines mitigation and adaptation strategies to enhance resilience in the short, medium, and long term, with an emphasis on decision support under evolving uncertainties. This includes the development of cost-benefit analyses, advanced risk assessment methodologies, optimization frameworks, and asset management strategies tailored to a changing climate. Furthermore, the mini-symposium highlights the urgent need to evolve engineering design, assessment, and maintenance standards to reflect new climate realities. By inviting contributions from academics, researchers, and industry professionals, this symposium seeks to foster collaboration and advance knowledge in areas such as probabilistic modeling, vulnerability assessment, resilience evaluation, uncertainty management in standards, and decision support for climate adaptation and mitigation. Together, these efforts aim to provide actionable insights and methodologies for building climate-resilient infrastructure systems that can better withstand future uncertainties.
We encourage submissions in the following areas:
- Probabilistic modeling of climate change impacts: Quantifying the effects of climate change on structural systems and infrastructure networks.
- Criticality, severity, and vulnerability assessments: Identifying and prioritizing risks to infrastructure under changing climate conditions.
- Risk and resilience modeling: Developing frameworks to evaluate and enhance infrastructure resilience.
- Decision support for adaptation: Formulating and optimizing adaptation measures through robust decision-making tools.
- Uncertainty management in design and assessment: Addressing uncertainties in climate projections and their integration into engineering standards.
Organisers and Affiliations:
Dr Cao Wang1, Dr Hao Zhang2, Dr Bilal Ayyub, Dr Michael Beer4
1UTS, 2The University of Sydney, 3University of Maryland, 4Leibniz Universität Hannover
Summary:
This Mini Symposium will address the resilience of civil structures, infrastructure networks, and communities (with interconnected infrastructure networks) in a changing environment. The safety and serviceability of in-service structures and infrastructure networks are often threatened by natural hazards. Owners/decision makers are thus concerned about the resilience of an asset, that is, the ability to be in readiness for, to absorb, recover from, and adapt to disruptive events. This is particularly the case when considering the potential impacts of climate change, which may lead to nonstationary natural hazards in the future (e.g., increasing trends of hazardous events in a changing climate). Moreover, on many occasions, the reduced performance of one network may further negatively impact the serviceability of another independent network. This symposium is themed the climate-resilience of structures, infrastructures and communities, with a focus on the hazard modelling, resilience methods, and case studies.
Organisers and Affiliations:
Prof Alex Sixie Cao1, Prof Yaohan Li2, Prof Marco Broccardo3, Prof Nadieh Meinen4, Prof Pablo G. Morato5, Dr. Orson Tieleman6, Prof. Daniel Straub7
1Nanyang Technological University, Singapore, Singapore, 2Hong Kong Metropolitan University, Hong Kong SAR, China, 3University of Trento, Trento, Italy, 4University of Southern Denmark, Odense, Denmark, 5Aalborg University, Aalborg, Denmark, 6TU Delft, Delft, Netherlands, 7TU Munich, Munich, Germany
Summary:
Climate change challenges established approaches to the design, assessment, and management of structures and infrastructure. Environmental actions are becoming non-stationary, while their future evolution is subject to uncertainties in socioeconomic developments, climate models, and natural climate variability. A scientific basis is thus urgently needed to address these challenges, translate climate information and associated uncertainties into engineering decisions, and support the development of standards, regulations, and guidance.
This mini-symposium brings together research contributions that aim to improve engineering design, assessment, and decision-making in the context of climate change and its associated uncertainties. Contributions are particularly welcome on:
- incorporation of climate change into codes, standards, and guidelines;
- stochastic modelling of non-stationary environmental actions and treatment of uncertainties in climate projections;
- decision-making under uncertain futures, including sequential decisions, future learning, robustness, and deep uncertainty;
- benchmark studies, case studies, and practical applications.
The mini-symposium is organized by the Joint Committee on Structural Safety (JCSS) Working Group on Uncertain Climate Futures (https://uncertain-climate-futures.github.io). Contributions and discussions from the mini-symposium will inform the working group’s reports, benchmark cases, and future engineering guidance.
Organisers and Affiliations:
Prof Leonardo Duenas Osorio1, Prof. Ji-Eun Byun2
1Rice University, 2Imperial College London
Summary:
Networks represent systems whose components and process interact under graph constraints. This makes them a versatile modelling scheme, spanning infrastructure systems such as transport and power grids, structural and mechanical systems, and social systems. Yet modelling, inference, and decision-making for such systems under uncertainty remain difficult. Network performance is often expensive to simulate and hard to validate against available data, and the underlying probabilistic problems are high-dimensional, which complicates inference and optimisation alike. Substantial progress has been made on each front, leveraging graph algorithms, Bayesian inference, hierarchical modelling, system reliability methods, surrogate modelling, advanced sampling schemes, and machine learning and AI. Even so, faithful simulation of network-level performance and tractable inference for system risk quantification, Bayesian updating, and optimisation remain open problems.
This mini-symposium aims to bring together researchers working on modelling, inference, and decision-making for complex network systems under uncertainty, across sectors and applications. Contributions addressing theoretical advances and real-world case studies are equally encouraged. Potential topics include, but are not limited to:
- Efficient network reliability methods, such as event decomposition, minimal cuts and paths, surrogate modelling, and advanced sampling
- Accelerated reliability methods through hardware-driven advances, such as parallel computing, GPU- or TPU-accelerated methods, and quantum circuits
- Modelling of dependence, common-cause failures, spatial correlation, cascading failures, and systems-of-systems
- Resilience quantification and restoration modelling of spatially distributed systems
- Bayesian inference and digital twins informed by monitoring and operational data
- Risk-informed decision-making for network systems, including reliability-based optimisation, resource allocation, and retrofit or maintenance prioritisation
- Validation and benchmarking with real-world case studies, open data, and open-source software
The following potential contributors have been contacted: Junho Song (SNU), David Yang (PSU), Paolo University (Lehigh), Min Ouyang (HUST), Negin Alekmazoor (UVA), Iris Tien (GT), Wei-Chang Yeh (NTHU), Radislav Vaisman (UQ), Hector Cancela (Udelar), Pablo Romero (Udelar).
Organisers and Affiliations:
Dr Preetish Kakoty1, Dr Cecilia Nievas2, Dr Chenbo Wang1
1University College London, 2GFZ Helmholtz Centre for Geosciences
Summary:
Communities and infrastructure are increasingly exposed to multiple hazards occurring in close spatial and temporal proximity. The resulting compound risks can produce impacts disproportionately greater than the sum of individual events and complicate response and recovery. Multi-hazard events accounted for only 19% of recorded disasters between 1900 and 2023, but 59% of global economic losses. They are also becoming more likely, driven by rapid urbanisation, growing interdependence of infrastructure, and a changing climate: in Europe, flood events interacting with other hazards increased by 186% between the 1980s and the 2010s. State-of-practice approaches treat hazards independently and superimpose their impacts. This remains tractable and underpins most operational assessments, but it cannot realistically capture interacting hazards and impacts, a gap that matters more as such events grow more frequent and consequential.
This mini-symposium will collect and curate recent developments in probabilistic methods for quantifying compound and multi-hazard risk across the full risk-modelling workflow. We invite contributions on: stochastic and physics-based simulation of joint and successive hazard events; high-fidelity modelling of physical assets and human exposure that illuminates the distributional justice of compound impacts; coupled, state- and path-dependent fragility and vulnerability models; portfolio- and network-level loss and consequence quantification; surrogate and data-driven methods that make high-dimensional assessment tractable; uncertainty quantification and propagation; and decision-relevant metrics for compounding risk. Applications spanning geological, atmospheric, and meteorological hazards are welcome, including case studies that apply probabilistic frameworks to policy decisions.
The theme aligns with the conference's focus on structural reliability, natural hazard risk and resilience, and with the growing role of data-driven methods in uncertainty quantification. By bringing together researchers working at the interfaces of hazard, performance and decision modelling, the session aims to consolidate emerging approaches and identify open problems in probabilistic frameworks for compound risk.
Organisers and Affiliations:
Prof Ketson Dos Santos1, Prof Vasileios Fragkoulis2, Prof Ioannis Kougioumtzoglou3, Prof Antonina Pirrotta4, Prof Athanasios Pantelous5
1University Of Minnesota, 2University of Liverpool, 3Columbia University, 4University of Palermo, 5Monash University
Summary:
Ever-increasing computational capabilities, novel signal processing techniques, and advanced experimental setups have led to highly complex modeling of engineering systems and related excitations. Examples include nonlinear/hysteretic response behaviors and non-Gaussian, non-white, and non-stationary excitation modeling. In many cases, even the deterministic solution of such governing equations remains an open issue and an active research topic. Naturally, solving the stochastic counterparts of these equations becomes even more challenging, particularly for high-dimensional systems.
The objective of this MS is to present recent advances and emerging cross-disciplinary approaches in the broad field of computational methods for stochastic engineering dynamics problems, with a focus on uncertainty modeling and propagation. Of great interest are contributions related to both fundamental research and engineering applications of stochastic dynamics, as well as signal processing approaches in the modeling and solution processes. A non-exhaustive list of topics includes joint time-frequency analysis tools, sparse representation and compressive sampling methodologies, stochastic/fractional calculus modeling and applications, nonlinear stochastic dynamics, stochastic model/dimension reduction techniques, Monte Carlo simulation methods, and risk/reliability assessment applications.
Organisers and Affiliations:
Dr Bowei Li1, Dr Dimitris Giovanis2, Dr Somdatta Goswami2, Dr Michael Shields2, Dr Seymour Spence3, Dr Alexandros Taflanidis4
1Texas Tech University, 2Johns Hopkins University, 3University of Michigan, 4University of Notre Dame
Summary:
Natural hazard engineering increasingly relies on high-fidelity simulations and heterogeneous observations to characterize hazards, structural response, system performance, and community consequences. Yet direct probabilistic analysis of these models remains challenging because of high dimensionality, rare events, limited or biased data, and interacting aleatory and epistemic uncertainties. Addressing these challenges requires advances at the intersection of computational statistics, uncertainty quantification, reliability analysis, and scientific machine learning. This Mini-Symposium will bring together researchers developing computationally efficient methods for natural hazard and performance-based engineering. Topics of particular interest include surrogate and reduced-order modeling for accelerating uncertainty propagation and reliability assessment; dimensionality reduction for high-dimensional hazards, responses, and parameter spaces; Bayesian inference, data assimilation, and model updating; active learning and optimal experimental design; rare-event simulation and reliability methods; and approaches for quantifying predictive, model-form, and data uncertainty in machine-learning-assisted analyses. Contributions integrating physics-based simulations with observational, experimental, or post-event data are especially encouraged. The session will emphasize not only computational acceleration, but also applications spanning wind, earthquake, flood, wildfire, and other natural hazards, from component and structural performance to infrastructure-system and community resilience. By connecting advances in probability and statistics with emerging computational tools, the Mini-Symposium seeks to support credible, risk-informed, and performance-based decisions for civil structures and infrastructure systems.
Organisers and Affiliations:
Dr Jia-Yi Ding1, Dr Yi Luo2, Dr De-Cheng Feng1, Dr Carmine Galasso3, Dr Michael Beer2
1Southeast University, 2Leibniz University Hannover, 3University College London
Summary:
Critical infrastructure systems are increasingly challenged by natural hazards, aging and deterioration, climate-related extremes, and cascading disruptions. At the same time, the growing availability of data from structural health monitoring, remote sensing, inspections, experiments, numerical simulations, and infrastructure operations is creating new opportunities for more reliable and efficient risk and resilience assessment. Translating these heterogeneous, sparse, and uncertain data into actionable insights into infrastructure performance, risk, and resilience, however, remains a major challenge. This Mini Symposium aims to bring together researchers working at the intersection of data science, uncertainty quantification, reliability analysis, and infrastructure resilience. Particular emphasis will be placed on data-driven, probabilistic, and hybrid approaches that quantify and propagate uncertainties associated with hazards and observations through infrastructure response and damage to system-level risk, functionality, recovery, and resilience.
Topics of interest include, but are not limited to:
- Data quality, data uncertainty, and uncertainty quantification
- Surrogate modeling, reduced-order modeling, and multi-fidelity methods
- Data-driven reliability, vulnerability, and risk analysis
- Spatial and temporal dependence and extreme-event modeling
- System and network reliability of critical infrastructure systems
- Cascading failures, interdependencies, and multi-hazard risk propagation
- Post-disaster functionality, recovery, and resilience assessment
- Data-informed risk mitigation, adaptation, and decision-making under uncertainty
Organisers and Affiliations:
Dr Xinyue Wang1, Prof Carmine Galasso1, Prof Michele Barbato2
1University College London, 2UC Davis
Summary:
Infrastructure systems (e.g., transport, water, and power) must be managed over long lifetimes amid uncertain and evolving hazards, aging, operating conditions, technologies, and societal needs, shaped by factors such as climate change, urbanization, technological innovation, and changing patterns of infrastructure dependence. Decision-making is further complicated by multiple, potentially competing objectives (e.g., technical, economic, societal, and environmental considerations) and diverse stakeholder priorities and preferences, all of which may evolve over time.
Ensuring resilient, sustainable, and just infrastructure therefore requires decision approaches that go beyond optimizing for a single anticipated future and a narrow set of objectives. Instead, decisions should be robust across a wide range of plausible futures while remaining adaptable as new information and changing conditions emerge.
This mini-symposium brings together recent advances in risk and resilience modeling, computational methods, and decision analytics for understanding and managing complex, interdependent, and evolving infrastructure systems. We welcome contributions that demonstrate practical outcomes such as decision-support tools, adaptation pathways, and policy recommendations and impacts. Topics of interest include, but are not limited to:
- Computational tools and frameworks for robust and adaptive decision-making under deep uncertainty and dynamic complexity, such as dynamic decision-making, many-objective robust decision-making, and adaptive pathways;
- Development of infrastructure adaptation strategies under multi-hazard conditions, such as earthquakes and their cascading hazards, floods, wildfires, and heatwaves, among others;
- Risk, resilience, and decision-making challenges associated with emerging infrastructure systems and technologies, including renewable energy (e.g., solar and wind), data centers, electric vehicles, and other rapidly evolving systems;
- Dynamic risk and resilience modeling methods for complex infrastructure systems to assess time-dependent hazards, structural loads, vulnerability, deterioration, and system-level impacts from changing climate, extreme events, and other evolving conditions.
- Conceptual, theoretical, computational, experimental, field-based, and methodological contributions are all welcome, with special focus on practice- and policy-oriented contributions.
Organisers and Affiliations:
Prof Maria Nogal1, Prof Mark G. Stewart2, Dr Erica Arango1
1TU Delft, 2University of Technology Sydney
Summary:
Wildfires at the wildland-urban interface (WUI) have become one of the fastest-growing sources of societal and infrastructure risk, yet the decisions that shape WUI resilience are still taken with limited quantification of the uncertainties involved. Choices such as which structures to retrofit, where to treat fuels, when to evacuate, and how to rebuild, are made under deep uncertainty about ignition, spread, exposure and human response.
This mini-symposium addresses the probabilistic and statistical foundations of that decision problem. It is organised around three couplings. The first is the bi-directional interaction between wildfires and infrastructure: networks and utilities are simultaneously a source of ignition and a system whose disruption governs evacuation, response and recovery. The second is the interaction between fire in vegetation and fire in the built environment, which determines how a landscape fire becomes a community-scale conflagration. The third is the role of communities as both the exposed system and the primary agent of mitigation.
The session invites contributions that develop or apply probabilistic modelling, statistical inference, uncertainty quantification and decision analysis to improve decision-making in WUI communities. Topics to be discussed include, but are not limited to:
- Community-level risk assessment and exposure characterisation
- Utility risk mitigation
- Forensic analysis of the WUI fire problem
- Vegetation and landscape management
- Fire recovery decisions
- Building guides and standards for WUI
- WUI-risk mapping and land-use planning
- Risk-informed prescribed burning
- Human behaviour in WUI fires
By bringing these strands together, the session aims to consolidate a decision-oriented perspective on WUI resilience, currently dispersed across the fire safety engineering, natural hazards and infrastructure resilience communities, and to establish it within the ICASP tradition.
Organisers and Affiliations:
Dr Qindan Huang1, Dr FaBio Biondini2
1Marquette University, 2Politecnico di Milano
Summary:
Civil infrastructure systems deteriorate over time due to aging, environmental exposure, operational demands, and extreme events. Reliable prediction of this deterioration is fundamental to structural reliability, service-life assessment, and life-cycle management. In practice, however, inspections, maintenance, repair, and rehabilitation also strongly influence infrastructure conditions. Accurately representing the effects of these interventions—and their associated uncertainties—remains an important challenge in life-cycle analysis.
This mini symposium will bring together researchers and practitioners developing advanced approaches for deterioration and intervention modeling of civil infrastructure. The symposium will emphasize probabilistic, statistical, and data-driven methods that characterize deterioration processes, quantify uncertainty, and explicitly account for the effects of inspection, maintenance, repair, and rehabilitation. Emerging approaches involving artificial intelligence and machine learning are also encouraged, particularly when integrated with physical knowledge, reliability analysis, or probabilistic frameworks.
Topics of interest include, but are not limited to, probabilistic and stochastic deterioration modeling; data-driven and artificial intelligence-based deterioration prediction; modeling of inspection, maintenance, repair, and rehabilitation effects for life-cycle analysis of infrastructure systems under uncertainty; updating deterioration models using inspection and monitoring data; structural reliability and time-dependent reliability; service-life prediction; validation and calibration of life-cycle models; uncertainty quantification and propagation; and life-cycle risk, cost, and performance assessment.
The mini-symposium aims to foster discussion on integrating deterioration and intervention models to support realistic predictions of infrastructure performance and risk-informed life-cycle decision-making. Applications across bridges, buildings, pipelines, transportation infrastructure, and other civil infrastructure systems are welcome.
Organisers and Affiliations:
Prof Minhyeok Ko1, Prof Pablo G. Morato2, Prof Kostas G. Papakonstantinou3
1The University Of Texas At Tyler, 2Aalborg University, 3The Pennsylvania State University
Summary:
Advances in sensing, computational modeling, and machine learning are expanding capabilities for assessing, predicting, and managing the performance of engineering systems. Probabilistic learning and data assimilation support the updating of system states and model parameters as observations become available, enabling digital twins to inform engineering assessment and decision-making. However, nonlinear dynamics, model discrepancy, nonstationary uncertainty, and changing environmental and operating conditions pose complexities for estimation and uncertainty quantification. For digital twins, a further challenge lies in optimizing information-gathering and intervention decisions. Addressing these challenges motivates the integration of physics-based modeling with data-driven methods to increase predictive accuracy and preserve physical consistency, ultimately leading to enhanced decision-making.
This mini-symposium will bring together researchers developing and applying probabilistic, physics-based, and data-driven methods for prediction, inference, and decision-making in complex systems. We invite contributions on theoretical formulations, computational methods, and applications to engineering systems. Topics of interest thus include, but are not limited to:
- Hybrid physics-based and machine learning models for probabilistic estimation
- Bayesian filtering, data assimilation, and system identification
- Generative world models and probabilistic inference
- Robust inference under model discrepancy, nonstationarity, sparse or heterogeneous data, and distribution shifts
- Digital twins for prediction, assessment, and decision-making
- Value of information, adaptive testing, and structural health monitoring
- Sequential decision-making under uncertainty
- Autonomous and multi-agent systems
Organisers and Affiliations:
Prof Masayuki Kohiyama1, Prof Tatsuya Itoi2, Prof Kazuyoshi Nishijima3
1Keio University, 2The University of Tokyo, 3Disaster Prevention Research Institute, Kyoto University
Summary:
Predictions of more frequent and severe disasters due to climate change are rapidly becoming reality. Beyond environmental and equipment design challenges, such as managing extreme heat, climate change increases the risk that conventional probability models will no longer adequately predict structural loads from wind, snow, temperature fluctuations, flooding, landslides, and tsunamis. As a result, structural design challenges have been brought to the forefront.
Every stage of the urban and architectural life cycle—planning and design, construction, operation and maintenance, and demolition and renewal—must evolve into a new model that harmonizes climate change mitigation with adaptation. The organizers of this mini symposium define methods that enable civil engineering structures to appropriately respond to climate change throughout their service life as “Future-Ready Performance-Based Design.” We are actively working to develop the academic framework and supporting technologies for this concept.
This mini symposium will feature presentations and discussions on research contributing to the development of the future-ready performance-based designs, with the goal of establishing a solid academic framework for the field. We welcome presentations on the following topics:
- Load models for various hazards aligned with climate change scenarios
- Evaluation methods for load combinations
- Technologies for assessing the resilience of buildings and cities
- Structural and HVAC (heating, ventilation, and air conditioning) system design
- Design and evaluation technologies that comprehensively address both adaptation to and mitigation of climate change
Organisers and Affiliations:
Prof Quanwang LI1, Prof Chun-Qing Li2, Prof Mark Stewart3
1Tsinghua University, 2RMIT University, 3University of Technology Sydney
Summary:
There is a need to ensure that infrastructure systems be more sustainable and their carbon emissions be reduced over their life-cycle. The durability, reliability and sustainability benefit of sustainable construction materials such as low carbon concretes and recycled materials is a key research challenge to ensure safe and reliable infrastructure during its service life, while also reducing carbon emissions. Key to this is predicting durability, safety and sustainability performance that considers carbon emissions for materials, construction processes, maintenance and repair, demolition, re-use, operational energy, etc. The criteria for decision making may be meeting carbon neutrality, reducing carbon when compared with a benchmark scenario, and/or meeting other performance or regulatory requirements. This Mini-Symposium will allow researchers to discuss these issues, and present their latest work on efforts to decarbonise our infrastructure while maintaining cost-effectiveness and a strong sustainability benefit. This might include durability or safety reliabilities of sustainable construction materials, deterioration model updating, life-cycle cost and carbon assessments, carbon neutrality decision metrics, risk-based decision making under uncertainty, and long-term performance of structures in complex environments, particularly using artificial intelligence methods.
Organisers and Affiliations:
A/Prof Yaohan Li1, Prof You Dong2, Prof Michael Faber3,4
1Hong Kong Metropolitan University, 2The Hong Kong Polytechnic University, 3Harbin Institute of Technology, 4Lusófona University
Summary:
Infrastructure systems today face unprecedented challenges over their life cycles, including aging, environmental degradation, and increasingly severe, climate-induced hazards. Climate change has become a dominant driver of risk, intensifying the frequency and magnitude of extreme events and introducing new patterns of uncertainty that traditional engineering approaches are ill-equipped to manage. These changes challenge conventional assumptions of stationarity in hazard characterization, accelerate the aging and deterioration of critical infrastructure, and increase the likelihood of system failures. From individual assets to infrastructure networks, and communities, climate-induced hazards and associated uncertainties threaten the long-term safety, reliability, resilience, and sustainability of the built environment. Effective adaptation therefore requires probabilistic frameworks that can characterize evolving hazards, quantify uncertainties, assess changing risks, and support robust long-term decision-making.
This Mini-Symposium aims to bring together researchers and practitioners to discuss recent advances in the assessment, adaptation, and management of infrastructure systems under climate change. Particular emphasis will be placed on approaches integrating statistics and probability with climate science, structural and system reliability, resilience analysis, and risk-informed decision-making. Topics of interest include, but are not limited to: (1) probabilistic and nonstationary modeling of climate-related hazards and extremes; (2) vulnerability, reliability, risk, and resilience assessment under extreme, compound, and cascading events; (3) uncertainty quantification for climate projections, hazard models, and infrastructure response; (4) time-dependent deterioration, aging, and performance assessment under changing environmental conditions; (5) climate adaptation, maintenance, and management strategies for infrastructure systems; (6) monitoring and data-driven approaches supporting climate adaptation and decision-making throughout the infrastructure life cycle.
Organisers and Affiliations:
Prof Jinhyun Choo1, Prof. Kok-Kwang Phoon2, Prof. Yat Fai Leung3, Prof. Yu Wang4
1Seoul National University, 2Singapore University of Technology and Design, 3Hong Kong Polytechnic University, 4Hong Kong University of Science and Technology
Summary:
One long-standing challenge in geotechnical risk and reliability is the scarcity of site-specific data for characterizing uncertainty and spatial variability with sufficient confidence. Another is the computational cost of reliability analysis, which can limit its use in routine practice. A third is the important yet largely qualitative role of engineering judgment in geotechnical decision-making. This mini-symposium aims to showcase how recent advances in machine learning and artificial intelligence can offer new perspectives on these challenges and enable more data-informed, efficient, and transparent approaches to geotechnical risk and reliability assessment.
Organisers and Affiliations:
Prof You Dong1, Prof Changhai Zhai2, Prof Mitsuyoshi AKIYAMA3, Prof John van de Lindt4
1The Hong Kong Polytechnic University, 2Harbin Institute of Technology, 3Waseda University, 4Colorado State University
Summary:
Modern critical infrastructure systems spanning energy, water, transportation, and communication face unprecedented threats from intensifying climate change and compounding natural hazards. Traditionally managed as isolated physical networks, these systems are deeply intertwined with the communities they support including natural ecosystems, households, businesses, and the economy. When a hazard strikes, failures cascade across sectors, affecting the economy, and often disproportionately impacting vulnerable populations.
Addressing this complex reality demands a paradigm shift toward multi-disciplinary resilience. This requires bridging quantitative engineering risk analysis with both quantitative and qualitative socio-ecological and economic perspectives to ensure that infrastructure not only withstands shocks but also supports equitable, long-term community recovery.
Core Topics:
- Climate Change & Multi-Hazard Dynamics: Evaluating compounding risks from extreme weather events, sea-level rise, and chronic environmental stressors on aging infrastructure.
- Eco-Social Integration: Embedding socio-ecological dimensions—such as environmental justice, public health, and local ecosystem services—into infrastructure vulnerability assessments.
- Community Resilience: Full systems of system community level modeling and analysis for decision support
- Cascading and compound hazards and their effects on infrastructure systems, community resilience, and decision-making
- Aggregation Modeling: Developing macro-level frameworks to synthesize systemic data for regional planning, economic forecasting, and national policy-making.
- Disaggregation Modeling: Scaling down macro-models to neighborhood and asset levels to identify localized community risks and pinpoint vulnerable structural components.
- Decision-Making Under Uncertainty: Utilizing advanced risk analytics and multi-criteria decision-making
Organisers and Affiliations:
Prof Daniel Straub1, Prof. Kostas Papakonstantinou, Prof. Matteo Pozzi, Prof. Iason Papaioannou, Prof. Charalampos Andriotis
1Technical University Of Munich
Summary:
Results from engineering assessments ultimately serve as decision support. Hence it is often relevant to set the assessments in the context of a formal decision analysis, to ensure an optimal choice of methods and interpretation of results. In addition, increasingly automated and autonomous systems that perform their own decision making require new approaches and methods for quantifying and validating performance, safety, and reliability.
This mini-symposium gathers researchers interested in developing and applying methods for optimal decision making under uncertainty for engineering systems. It focuses on optimization involving stochastic environments, objectives and constraints, and the integration of formal decision analysis with stochastic models, simulators, data analytics, and artificial intelligence methodologies towards long-term planning in complex settings. These include sequential decision processes as well as decisions involving multiple stakeholders, among others. Selection of metrics for stochastic analysis based on decision-theoretic considerations are also of interest, such as the choice of appropriate objective functions and decision-theoretic sensitivity measures.
Areas of interest, both in terms of methodologies and applications, include, but are not limited to:
- Bayesian decision theory
- Markov Decision Processes
- Reinforcement learning / Artificial Intelligence methods
- Value of Information and structural health monitoring
- Stochastic control and dynamic programming
- Stochastic optimization
- Sensitivity analysis
- Multi-agent decentralized systems
- Multi-objective/constrained optimization
- Autonomous systems
- Life-cycle infrastructure management and optimization
- Inspection and maintenance planning
- Early warning, resilient recovery and emergency response
- Risk assessment
- Probabilistic performance-based engineering
Organisers and Affiliations:
Dr Abigail Beck1, Dr Raul Rincon2, Dr Luis Ceferino3, Dr Milad Roohi4, Dr. Jamie Padgett2
1University of Houston, 2Rice University, 3University of California, Berkeley , 4University of Nebraska-Lincoln
Summary:
The philosophy of civil engineering design is expanding beyond traditional life safety and serviceability limit states to encompass broader resilience goals. Engineers and resilience analysts are increasingly tasked with considering how infrastructure disruptions and failures following hazard events affect community services, institutions, and social well-being over space and time. Addressing this broader design approach requires probabilistic methods, scalable algorithms, computational models, and robust metrics that connect physical system performance with the consequences users experience within social systems. This mini-symposium will bring together advances in the modeling, assessment, design, and management of coupled sociotechnical infrastructure systems. The session will emphasize probabilistic methods and modeling techniques that account for the coupled nature of sociotechnical impact and resilience assessment. That is, focus will be placed on methods that advance the modeling of structure-to-system failures, their translation into utility service disruptions and user impacts, how these propagate into broader effects on the functioning of households, institutions, and larger communities, and the co-evolution of socio-physical systems during recovery processes.
We invite contributions that develop methods, metrics, decision frameworks, and computational algorithms that link infrastructure and social systems, and that perform infrastructure performance assessment within a joint sociotechnical context. Relevant topics include multiscale coupling of physical and social systems; probabilistic quantification of user and community disruption impacts; differential-based performance metrics; socially-aware optimization- and heuristics-based design of risk mitigation solutions; uncertainties during sociotechnical models’ development (e.g., spatial scale, modelers’ risk perceptions, behavioral decisions of modeled agents); and socially-aware decision-support methods for managing infrastructure. Applications can examine interactions ranging from individuals and local infrastructure components to business, schools, hospitals, and larger regional networks. The session also seeks to span natural, technological, and compound hazards. The ultimate goal is to convene around methods supporting broader sociotechnical considerations needed to advance resilience-based infrastructure design and decision-making.
Organisers and Affiliations:
Dr Pouria Kourehpaz1, Professor Milad Roohi2, Professor Carmine Galasso3, Professor John van de Lindt4
1Swiss Re, 2University of Nebraska–Lincoln, 3University College London, 4Colorado State University
Summary:
Functional recovery, the time for a building or system to regain the functions it was designed to provide, has moved from a research concept to an emerging design objective. Recovery-based criteria are under development for U.S. seismic design standards (ASCE 7-28), with parallel efforts internationally, yet most advances remain earthquake-focused and building-centric. Communities recover as systems of systems: buildings depend on power, water, transportation, and telecommunications, and these networks depend on one another. Recovery trajectories are further altered by multi-hazard interactions, including aftershocks, cascading hydro-meteorological hazards, and independent events that strike while systems are still recovering, conditions that single-hazard frameworks were not built to represent. Probabilistic and computational methods for quantifying recovery across interdependent systems and successive hazards, together with the empirical data needed to validate them, remain comparatively underdeveloped.
This mini-symposium brings together researchers, engineers, catastrophe risk modelers, and social scientists to assess the state of functional recovery assessment and to identify advances needed for interconnected systems and multi-hazard settings. Contributions are invited on topics including, but not limited to:
- Probabilistic and computational methods for functional recovery and downtime quantification
- Uncertainty propagation from inventories, fragilities, and restoration models to recovery metrics
- Functional recovery in design guidelines, standards, and codes
- Systems-of-systems and interdependent network recovery
- Multi-hazard and sequential-hazard recovery analysis
- Empirical evidence, reconnaissance, and longitudinal community recovery studies
- Decision support, mitigation prioritization, and risk communication based on recovery metrics
- Interdependencies and cascading effects on functional recovery
Organisers and Affiliations:
Dr Chao Sheng1, Dr. Hao Qin2, Prof. Paolo Bocchini3, Prof. Mark Stewart4
1Sichuan University, 2Harbin Institute of Technology, 3Lehigh University, 4University of Technology Sydney
Summary:
Civil structures and infrastructure are increasingly exposed to climate-related hazards such as tropical cyclones, extreme waves, sea level rise, storm surges, flooding, and wildfires. Climate change may further alter the frequency, intensity, spatial distribution, and dependence structures of these hazards, posing growing challenges to the safety and resilience of civil infrastructure across transportation, energy, water, communication, housing, and other critical systems. Conventional risk assessments based on stationary hazards, isolated built assets, or direct economic losses alone are therefore insufficient to support long-term infrastructure resilience and climate adaptation.
This mini symposium will bring together researchers and practitioners to share the recent advances in probabilistic modelling of climate hazards, risk and reliability assessment for buildings and infrastructure, community and regional resilience, and climate adaptation. Topics may include, but are not limited to:
- Probabilistic characterization, dependence modelling, and stochastic simulation of single, compound, and sequential climate hazards;
- Nonstationary hazard and extreme value modelling under climate change;
- Physics-based, data-driven, and hybrid approaches including machine learning and artificial intelligence techniques for climate hazard, risk, and resilience assessment;
- Structural and infrastructure reliability, vulnerability, and risk under climate hazards;
- Cascading failures and systemic risks across spatially distributed and interdependent infrastructure and social systems;
- Probabilistic assessment of community-level risk, functionality, recovery, and resilience;
- Uncertainty quantification and risk-informed decision-making for climate adaptation under deep uncertainty;
- Regional-scale applications, case studies, and best practices for climate risk reduction and infrastructure adaptation.
The session aims to connect advances in probabilistic modelling with adaptation strategies for climate-resilient infrastructure planning across asset, system, community, and regional scales.
Organisers and Affiliations:
Prof Taeyong Kim1, Sang-ri Yi2, Ji-Eun Byun3
1Yonsei University, 2Rice University, 3Imperial College London
Summary:
Modern infrastructure systems, including buildings, lifeline networks, and industrial facilities, are increasingly exposed to natural and human-caused/amplified hazards such as earthquakes, floods, urban fires, and wildfires. Ensuring their resilience requires not only accurate assessment of structural and system-level performance but also robust decision-making frameworks that account for the inherent uncertainties in hazard occurrence, system behavior, and recovery processes. This mini-symposium invites contributions that advance the state of the art in resilience assessment of complex infrastructure systems, with particular emphasis on the robust treatment of aleatory and/or epistemic uncertainties and on insights into how these uncertainties affect system resilience and risk-informed decision-making. Topics of interest include, but are not limited to: probabilistic risk and resilience quantification, community and infrastructure-network-level resilience analysis, and system reliability analysis methods. Some examples include novel probabilistic resilience analysis frameworks; Bayesian inference and sensitivity analysis methods for complex systems; regional-scale risk and resilience assessment of structural and lifeline systems; and risk-informed decision-making for the design, retrofit, and operation of infrastructure systems. By bringing together researchers working across these complementary methodological perspectives, this mini-symposium aims to foster discussion on integrated approaches that bridge reliability engineering and decision-making for the next generation of resilient infrastructure systems.
Organisers and Affiliations:
Prof Ricardo Monteiro1, Dr. Gianrocco Mucedero1, Prof. Carmine Galasso2, Associate Professor Carlos Molina Hutt3, Dr. Nicole Paul4
1University IUSS Pavia, 2University College London, 3University of British Columbia, 4ETH Zürich
Summary:
Post-event downtime, rather than collapse prevention alone, is increasingly recognised as a primary driver of household displacement, business interruption, and long-term community disruption, often outweighing direct repair costs in shaping recovery trajectories. Despite growing awareness that downtime governs the true societal impact of natural hazards and human-induced events, current new building or retrofit design practice still largely neglects explicit downtime and household impacts as design objectives, leaving a gap between structural performance-based design and real-world recovery outcomes. This Mini-Symposium brings together researchers, practitioners, and policymakers to discuss how new design concepts and retrofit strategies, for both new and existing buildings, can be conceived, evaluated, and optimised through the lens of downtime household impact reduction, across scales ranging from the single building to urban aggregates. The session welcomes contributions addressing, among others:
a) design methodologies and structural solutions, for new buildings or retrofit interventions, targeting reduced repair time, rapid re-occupancy, and habitability preservation;
b) building-level modelling and quantification of repair time and functional recovery to inform design choices;
c) portfolio- and community-scale design and prioritisation frameworks linking building-level downtime to housing displacement and socio-economic recovery;
d) decision-support and optimisation tools for downtime-informed seismic design and retrofit prioritisation.
Topics of interest include, but are not limited to:
- Downtime-informed design methodologies for new and existing buildings
- Downtime assessment frameworks
- Structural solutions enabling rapid re-occupancy and business continuity
- Displacement and temporary housing needs following catastrophic events
- Multi-scale design and prioritisation frameworks linking building downtime to community recovery
- AI- and data-driven approaches for downtime-informed assessment, design and retrofitting
- Design approaches aligned with performance-based and recovery-based engineering frameworks
- Policy and governance instruments accounting for downtime metrics
Organisers and Affiliations:
Dr Jiashu Yang1, Assoc. Prof. Lili Weng2, Prof. Hector Jensen3, Assoc. Prof. Seymour Spence4, Prof. Jianbing Chen5
1Xi'an University of Architecture and Technology, 2Fujian Agriculture and Forestry University, 3Universidad Técnica Federico Santa María, 4University of Michigan, 5Tongji University
Summary:
Engineering structures and systems are inevitably affected by uncertainties arising from geometric variability, material properties and deterioration, stochastic excitations, environmental and operational conditions, model-form assumptions, and limited data. Reliability-based and robust optimization provide systematic frameworks for integrating uncertainty quantification, reliability assessment, and engineering optimization, with the aims of controlling failure risk, reducing performance variability, and achieving efficient and reliable designs.
With increasing complexity in engineering systems, loading environments, numerical models, and available data, practical optimization under uncertainty faces significant challenges. These include the coupling of nonlinearity and uncertainty, high-dimensional random and design spaces, rare failure events, time-dependent performance, and computationally intensive simulations.
This Mini-Symposium aims to bring together recent theoretical, computational, and engineering advances in reliability-based and robust optimization. Topics of interest include, but are not limited to:
- Reliability-based and robust optimization and related formulations, including worst-case, risk-informed, multi-objective, and distributionally robust approaches;
- Static, dynamic, and time-dependent optimization involving reliability and statistical performance measures;
- Reliability-based and robust topology optimization of discrete and continuum structures;
- Efficient reliability analysis, rare-event simulation, and sensitivity analysis for optimization;
- Surrogate modeling, reduced-order modeling, multi-fidelity methods, and machine-learning-assisted approaches for computationally intensive optimization;
- Data-informed optimization under limited or incomplete probabilistic information, including prior and posterior information;
- AI empowered and AI stimulated new methodologies and approaches, intelligent approaches from physics-constraint to physics-guided, for reliability-based optimization and systems.
Particular emphasis will be placed on methods that improve computational efficiency, scalability, robustness, and practical applicability. The Mini-Symposium welcomes both methodological advances and real-world applications to structures, utility and infrastructure networks, and related engineering systems, and seeks to facilitate discussion of common challenges and emerging directions in reliable and robust engineering design.
Organisers and Affiliations:
Prof Hussam Mahmoud1, Prof Hiba Baroud1
1Vanderbilt University
Summary:
Modern society relies on the functionality of deeply interconnected infrastructure networks. The physical performance of power grids, water networks, and transportation systems directly dictates the operational capacity of vital social institutions. This session focuses on novel probabilistic frameworks, network analysis, reliability models, and simulation methods designed to quantify and enhance the resilience of these complex, interdependent systems against extreme hazards.
A critical frontier in structural and system reliability is mapping physical infrastructure disruptions to societal impacts and vice versa. For instance, the functionality of hospitals and schools depends not only on their structural integrity but also on the continuous supply of electricity, water, and accessible roadways. If a power grid fails, a hospital’s ability to provide life-saving care is severely compromised, even if the medical building itself suffers no structural damage. Similarly, the closure of schools due to prolonged utility outages disrupts community stability and delays regional economic recovery. Understanding the cascading effects, including failures and benefits, from the built environmental to social and economic systems is crucial to inform resilience planning.
This session invites contributions that bridge the gap between engineering reliability and social vulnerability. We welcome papers addressing:
- Multi-network modeling: Formulations capturing cascading failures across physical, cyber, and social systems.
- Dynamic system reliability: Methods evaluating infrastructure degradation and community recovery trajectories over time.
- Surrogate and simulation methods: Computational advances (e.g., Subset Simulation, machine learning surrogates) for rare-event quantification in high-dimensional system networks.
- Decision-support frameworks: Risk-informed optimization strategies for pre-disaster mitigation and post-disaster resource allocation.
By exploring the intersection of physical network reliability and institutional functionality, this session aims to advance actionable, risk-informed strategies that safeguard both built environments and the communities they support.
Organisers and Affiliations:
Dr Rui Teixeira1, Prof. Alan O'Connor1, Dr. Beatriz Martinez Pastor2, Dr. Vikram Pakrashi2, Dr. Maria Nogal3
1Trinity College Dublin, 2University College Dublin, 3Delft University of Technology
Summary:
Modern infrastructure systems, transport networks, energy grids, water systems, and the built environment, are increasingly instrumented, interconnected, and exposed to compounding climate and operational hazards. The full potential of the connection between the digital, instrumented and the physical world is yet to be realised (e.g. digital twins), while the complexity of such systems also put forward many challenges. Rapid or real-time uncertainty quantification, risk and resilience assessments are not yet systematically embedded in updating or modelling approaches and architectures.
This mini-symposium aims to bring together researchers working at the intersection of digital twin technologies, complex systems, and probabilistic/statistical methods for infrastructure assessment. It is expected to foment discussions on the challenges imposed be the sheer complexity of our large-scale infrastructure systems, and welcomes contributions on the updating of models using uncertain data; digital twins and reliability analysis within twinning frameworks; surrogate or reduced-order models for efficient computation; network-based and systems-engineering approaches to capturing interdependencies in infrastructure systems in real-time; or data-driven techniques for real-time risk and resilience assessment in large-scale systems.
Particular emphasis will be placed on applications relevant to critical infrastructure resilience against some of the key grand challenges of our times (e.g. climate change). With this, the symposium purposes the creation of a dialogue between engineers working on risk and reliability, data scientists, and infrastructure managers addressing methodological gaps between traditional probabilistic approaches and the dynamic nature of the modern approaches linking modelling, instrumentation and data to system level needs.
Organisers and Affiliations:
Dr Gaurav Datta1, Prof. Giuseppe Carlo Marano2
1ArtIStE Research Group, 2Politecnico di Torino
Summary:
Engineering systems and civil infrastructure are increasingly exposed to extreme hazards — earthquakes, hurricanes, floods, technological failures, and cascading events. Though infrequent, such events can be devastating, causing significant human, economic, and environmental losses. Amid growing uncertainties from climate change, urbanization, and aging infrastructure, the community must move beyond deterministic design toward probabilistic, robust, and resilient performance-based frameworks — firmly rooted in engineering mechanics — that capture the complex interactions between loads, materials, and structural systems under uncertain and evolving conditions.
This Mini Symposium brings together researchers and practitioners at the intersection of mechanics, optimization, and uncertainty quantification to advance methods that enable systems to resist, absorb, and recover from extreme events. Robustness denotes a system's ability to sustain functionality despite disturbances and modeling imperfections; resilience denotes its capacity to recover, adapt, and maintain performance after disruptions. Together they form the cornerstone of sustainable, adaptive infrastructure.
The session will focus on mechanics-based principles and computational strategies that integrate probabilistic mechanics, multi-objective optimization, and data-driven modeling to quantify uncertainty and improve risk-informed decision-making in complex engineering environments.
Topics of interest include, but are not limited to:
- Mechanics-driven formulations for robust and resilient design
- Multi-objective and reliability-based optimization under uncertainty
- Probabilistic and stochastic mechanics of structures and materials
- Performance-based and risk-informed design frameworks
- Surrogate modeling, reduced-order modeling, and machine learning techniques
- Data-driven and hybrid computational approaches for resilience assessment
- Decision-making and adaptive control strategies under uncertainty
Organisers and Affiliations:
Prof. Xiao-Hui YU1, Prof. Fulvio PARISI2, Prof. Jun YU3, Prof. Andre BECK4, Prof. De-Cheng FENG3, Prof. Yi-Ping GAN5, Asis/Prof. Lu-Chuan DING6, Prof. Jian-Bing CHEN6, Prof. Robby CASPEELE7
1Guilin University of Technology, 2University of Naples Federico II, 3Southeast University, 4University of São Paulo, 5Putian University, 6Tongji University, 7Ghent University
Summary:
Disproportional or progressive collapse is a relatively rare event, as it requires both an abnormal loading (e.g., explosions, vehicle impacts, fire and terrorist attacks) or an unforeseen event to initiate local damage and a structure that lacks adequate continuity, ductility, and redundancy to resist the spread of damage. However, such events can cause catastrophic casualties and property loss, exemplified by the 1968 Ronan Point collapse, the 2001 WTC attacks, and the 2021 Florida condominium collapse. Therefore, designing structures for sufficient robustness to mitigate this risk is paramount. Furthermore, optimal decision-making necessitates the quantitative assessment of structural robustness.
This Mini Symposia (MS) invites contributions from scientists, academics, and engineers addressing theoretical aspects, experimental results, numerical modeling, and practical design recommendations for both new and existing structures. Both theoretical developments and applications involving different structural systems are particularly welcomed in this MS. The scope includes, but is not limited to:
- Alternate load path analysis;
- Structural robustness quantification;
- Fragility modeling;
- Aging and deterioration effects in relation to structural robustness;
- System reliability;
- Risk-based approaches;
- Design optimization for structural robustness or reliability;
- Performance-based design under progressive collapse.
Organisers and Affiliations:
Dr Jungho Kim1, Dr Taeyong Kim2, Dr Ziqi Wang3, Dr Junho Song4
1Sejong University, 2Yonsei University, 3University of California Berkeley, 4Seoul National University
Summary:
Accurate assessment of structural and system-level performance under seismic hazards requires repeated evaluation of probabilistic response quantities, which is often computationally prohibitive when relying solely on high-fidelity numerical simulations. Recent advances in data-driven surrogate modeling, including Gaussian process models, polynomial chaos expansions, neural operators, and other statistical and machine-learning-based response predictors, offer promising pathways to accelerate probabilistic seismic response analysis while preserving the accuracy needed for reliability and resilience assessment. This mini-symposium invites contributions on the development and application of surrogate models for predicting structural and system responses under seismic hazards, with emphasis on uncertainty quantification, learning strategies for data-deficient conditions, and integration with downstream system reliability analysis. Topics of interest include, but are not limited to: surrogate models for nonlinear structural response prediction under seismic hazards; probabilistic and stochastic approaches to response prediction under uncertainty; data-driven strategies for structural performance and failure mode prediction, including in data-limited settings; efficient training, adaptive sampling, and multi-fidelity learning frameworks for structural response prediction; and surrogate-accelerated methods for system-reliability-based resilience analysis. By bringing together methodological advances in surrogate modeling with their application to probabilistic seismic risk and resilience assessment, this mini-symposium aims to foster discussion on scalable and computationally efficient tools for the next generation of resilience-informed engineering practice.
Organisers and Affiliations:
Prof Bruno Sudret1, Prof Jean-Marc Bourinet2, Prof Michael Shields3, Prof Alexandros Taflanidis4
1ETH Zurich, 2SIGMA Clermont, 3Johns Hopkins University, 4University of Notre Dame
Summary:
Computational models have become indispensable tools in engineering and applied sciences for the analysis, design and decision-making of complex systems. High-fidelity simulators, including finite element and multiphysics models, are now routinely employed for uncertainty propagation, reliability and robustness assessment, optimization under uncertainty, Bayesian inference and global sensitivity analysis. However, their repeated evaluation often remains computationally prohibitive, even with modern high-performance computing resources. Additional challenges arise for stochastic simulators, whose outputs are random for fixed input parameters due to intrinsic randomness or unresolved latent variables.
Surrogate models provide fast-to-evaluate approximations that can substantially reduce these computational costs. Over the last years, the field has rapidly expanded with new developments in machine learning, operator learning and probabilistic generative modelling, as well as dedicated approaches for stochastic simulators.
This mini-symposium aims to highlight recent advances and emerging trends in surrogate modelling for uncertainty quantification and reliability analysis. Contributions are invited on established methods such as sparse polynomial chaos expansions, Gaussian process regression/Kriging, support vector regression and sparse-grid techniques, as well as modern approaches including deep neural networks, neural operators, ensemble method, diffusion models, and flow matching / normalizing flows. Particular topics of interest include stochastic emulators, active learning and adaptive experimental designs, dimensionality reduction, multi-fidelity and multi-level modelling, and the quantification of surrogate-induced errors. Methodological developments as well as challenging applications in engineering and applied sciences are particularly welcome.
Organisers and Affiliations:
Dr Jin Wang1, Dr Yongxu Liu1
1Western University
Summary:
Tornadoes and other severe convective winds, including downbursts, derechos, and gust fronts, pose significant threats to buildings, infrastructure, and communities. Their short duration, rapid evolution, diverse spatial scales, complex flow structures, and limited observational records make their probabilistic characterization particularly challenging, with substantial aleatory variability and epistemic uncertainty.
This mini symposium will bring together researchers and practitioners working across meteorology, wind engineering, structural engineering, statistics, reliability, and risk assessment to discuss recent advances in characterizing these hazards and their associated risks. Topics of interest include probabilistic characterization of event occurrence, intensity, spatial extent, and temporal evolution; climate variability and change; wind loading and debris effects; structural response and vulnerability; loss estimation; and risk-informed design and mitigation. These topics may be addressed through statistical and probabilistic modeling, uncertainty quantification, and integration of field observations, remote sensing, post-event investigations, laboratory experiments, and numerical simulations.
The mini symposium will provide a forum for exchange aimed at advancing understanding of tornadoes and other severe convective wind hazards, their climatology, and the risks they pose to the built environment. The insights gained through these discussions are expected to help improve engineering design and risk-informed decision-making, while also highlighting remaining research needs.
Organisers and Affiliations:
Dr Aikaterini (Katerina) Kyprioti1, Dr. Jize Zhang2
1The University Of Oklahoma, 2Hong Kong University of Science and Technology
Summary:
Machine learning is rapidly transforming how natural hazards are simulated, forecasted, and translated into risk for civil infrastructure. Surrogate models, neural networks, and emerging machine learning and artificial intelligence models now replace computationally expensive numerical simulators for storm surge, flooding, wind, wildfire, seismic, and tsunami hazards, enabling probabilistic analyses at regional scales and, in some cases, near real time. Yet their value for risk-informed decision making depends on whether their predictions, and the uncertainty around them, can be trusted. This mini symposium brings together researchers developing and applying machine learning methods for hazard and risk assessment with a focus on rigorous uncertainty quantification, reliability, and validation.
Contributions are invited on topics including, but not limited to: Gaussian process, polynomial chaos, and deep learning surrogates for hazard simulators; physics-informed and physics-guided machine learning; generative and diffusion models for stochastic hazard scenarios; multi-fidelity and transfer learning with limited data; calibration and quantification of epistemic and aleatoric uncertainty in data-driven models; weather and climate foundation models for hazard forecasting; surrogate-based fragility, reliability, and life-cycle risk analysis of infrastructure systems; compound and cascading multi-hazard modeling; and applications supporting emergency management, design, and resilience planning. Studies that critically compare data-driven and physics-based approaches, or that address out-of-distribution performance and extrapolation under a changing climate, are especially welcome.
The session aims to connect the probabilistic modeling tradition of the ICASP community with the fast-moving machine learning community, and to identify shared benchmarks, open datasets, and best practices for trustworthy surrogate modeling in hazard engineering. By bringing together contributions internationally, the symposium will foster collaboration across different countries and highlight how uncertainty-aware machine learning can support safer and more resilient infrastructure.
Organisers and Affiliations:
Dr Meng-Ze Lyu1, Prof Matthias Faes2, Prof Ioannis Kougioumtzoglou3, Prof Jian-Bing Chen4, Prof Michael Beer1, Prof Jie Li4
1Leibniz University Hannover, 2Technische Universität Dortmund, 3Columbia University, 4Tongji University
Summary:
Stochastic dynamics provides a powerful framework for understanding the probabilistic behavior of civil engineering systems subjected to uncertain and time-varying excitations. Recent developments in computational probability and stochastic analysis have enabled new approaches for addressing high-dimensional, nonlinear, and nonstationary stochastic systems, while efficiently quantifying uncertainties and assessing their reliability. This Mini-Symposium aims to bring together researchers to discuss recent theoretical and computational advances in uncertainty quantification and reliability analysis of high-dimensional nonlinear stochastic systems, with emphasis on probabilistic characterization, efficient computation, and engineering applications. Topics of interest include, but are not limited to:
- Nonlinear and high-dimensional stochastic dynamical systems;
- Uncertainty quantification and propagation;
- Probability density evolution methods;
- Transient and nonstationary stochastic responses;
- Stochastic processes and dynamic response analysis;
- Dynamic and time-dependent reliability;
- Advanced stochastic simulation and computational methods;
- Data-driven approaches to uncertainty quantification and reliability;
- Multi-hazard stochastic dynamics and reliability;
- Applications in structural and infrastructure systems.
The symposium will provide a platform for exchanging new ideas, methodologies, and applications, and for identifying emerging directions in uncertainty quantification and reliability of high-dimensional nonlinear stochastic systems.
Organisers and Affiliations:
Dr Jaehwan Jeon1, Prof Wanting (Lisa) Wang2, Dr Isabela Durci Rodrigues3, Mr João Gabriel Duarte4, Prof. André T. Beck3
1University Of Toronto, 2Old Dominion University, 3University of São Paulo, 4University of Minnesota
Summary:
Structure-specific reliability analysis is the most faithful way to predict how a structure will perform under a natural hazard, but its cost has confined it to individual structures. At the scale of a community, assessments of risk and resilience therefore still rest on fragility curves assigned to building archetypes, even though two structures of the same archetype can respond very differently. Recent advances are narrowing this gap from several directions: analytical and semi-analytical methods that evaluate probabilistic response under stochastic excitation at a fraction of the cost of simulation, stochastic emulators and advanced sampling strategies that estimate rare failure events with few model evaluations, and learning-based surrogate models that transfer results across many structures and levels of fidelity. How these methods can be combined, how structure-level uncertainty propagates and aggregates into the community scale, and how the resulting predictions can inform regional models of loss, functionality, recovery, and mitigation decisions remain open questions that span the structural and community scales.
This mini-symposium brings together research on uncertainty quantification and stochastic emulation that makes structure-specific reliability assessment affordable at community scale, as well as research that translates these structure-level predictions at community-scale risk, resilience, and decision-making frameworks. Contributions on methods and on applications are equally welcome. Pertinent topics include, but are not limited to: analytical and numerical methods for the reliability of structures under stochastic excitation; stochastic emulators, surrogate models, and advanced sampling for rare-event estimation; multi-fidelity and learning-based approaches that scale reliability analysis across structures; propagation of structure-level uncertainty into regional loss, functionality, and recovery models; risk-based design and optimization of structures under natural hazards; and uncertainty-aware community resilience assessment, mitigation, and recovery planning.