Welcome to the Engineering Risk, Uncertainty, and Resilience Lab at Rice CEE!
We contribute to building resilient urban infrastructure by developing new computational uncertainty quantification (UQ) methods tailored for complex, large-scale, multi-hazard simulation models, advancing both methodological innovations and applied practices.
Our research interest covers the full range of engineering UQ applications including probabilistic modeling, efficient uncertainty propagation, and effective reduction of uncertainties. Special focus is on developing scalable and robust statistical tools that can be readily deployed to real-world engineering problems.
Modern natural hazards risk and resilience assessment paradigms are characterized by multi-scale, multi-fidelity, and complex infrastructural systems, presenting challenges such as:
- High-dimensionality in representation of system input-output
- Treatment of aleatoric system uncertainties
- Need for fail-safe UQ algorithms for high-risk systems
- Complex inter- and intra-system interdependencies
- Decision-making under uncertainty and conflicting decision objectives
- Analysis of spatiotemporal correlations
- Modeling and inference for scarce/imperfect/indirect datasets
- Trade-offs between computational scalability, efficiency, and robustness
- High computational memory and processor demands
Research Areas
We develop computational UQ methods to overcome the challenges and better understand the uncertainty, risk, and resilience of infrastructural systems under various hazards. Specific topics of interest include:
- Stochastic and deterministic surrogate modeling
- Adaptive design of computer experiment
- Statistical machine learning
- Smart sampling techniques
- Random vibration analysis
- Random process and field representations
- Bayesian inference
- Dimensionality reduction and data compression
- Multi-fidelity UQ
- Probabilistic decision theories
See Research page for more!