Research Overview: The StoStruct Research Group explores topics at the intersection of design, assessment, and monitoring of infrastructure under extreme conditions. We combine risk-based modeling, simulation, and data-driven approaches to understand how structural systems and their constituent components behave, and use that understanding to improve system-level performance.
Figure 1. Research Vision, thrusts, and topics pursued by StoStruct Research Group
Research Thrusts & Directions
1. AI-assisted risk-based design and assessment systems under extreme conditions
We explore how AI models and different learning paradigms can support risk-based design and assessment across scales (i.e., components, structures, spatially dispersed assets, and asset portfolios). The goal is to understand how these models can reason and generalize across broad design spaces and identify solutions that may not be visible to human designers. Key topics include: generalizability, explainability, generative models, and training under limited data and severe imbalance.
2. Digital twins to support monitoring and condition assessment of transportation infrastructure
We examine the development of engineering-oriented, scalable digital twins (i.e., dynamic virtual representations of physical systems) for transportation infrastructure, including bridges, signs, and barriers, to support intelligent maintenance under aging and natural hazards. Topics related to self-learning, uncertainty quantification, and human-AI collaborations are pursued under this research thrust.
3. Performance-based engineering to support community resilience
A primary research thrust of our group is advancing performance-based engineering, particularly performance-based design, to support a broad range of community resilience objectives, including economic loss and recovery, under single- and multi-hazard conditions. Recent topics include investigating the resilience of structural steel systems across the Wasatch Fault and modeling post-earthquake fire performance in tall steel systems. [See our work related to Utah seismic resiliency]
Selected Funded Research
The StoStruct research group is supported by federal and state agencies, nonprofit organizations, and industry partners. Several recent examples include, but are not limited to:
- Principal Investigator, “Developing context-aware computer vision models for robust data-informed condition assessment of bridges”, Center for Transformative Infrastructure Preservation and Sustainability (CTIP), August 2024-Present
- Principal Investigator, “A crowd-sourcing approach to developing visual datasets for safety assets“, Utah Department of Transportation (UDOT), September 2026-Present
- Principal Investigator, “A Numerical Investigation of Erection Bridging in Double-Pitched Joist”, Steel Joist Institute (SJI), August 2024-May 2025
- Principal Investigator, “Application of deep learning method to automate ground motion processing”, Pacific Gas & Electricity (PG&E), August 2023 – July 2024
