The Infrastructure Planning group develops high-fidelity digital twins and decision-support tools for water utilities and planners. Our work spans the full infrastructure lifecycle: from capital investment planning and asset management through real-time operational optimization and emergency response. We approach water infrastructure as a sociotechnical system — the engineering model is only as valuable as the governance structures and decision workflows that surround it.
We collaborate directly with municipal water authorities to deploy and validate our tools, ensuring that research outcomes address real operational and planning challenges. Current priority areas include pipe failure prediction, stormwater resilience under climate change, and adaptive management under supply uncertainty.
Real-time hydraulic state estimation and leak detection using ensemble Kalman filters fused with graph neural network surrogates of EPANET models.
Probabilistic performance assessment of aging stormwater networks under future precipitation extremes, coupling hydrological models with asset deterioration curves.
Machine learning models for predicting water main breaks using historical failure records, soil data, and pipe attributes, to guide capital replacement programs.
Attack detection and resilient control algorithms for identifying sensor spoofing and actuator manipulation in water SCADA systems.