This project is funded through the NSF Translation to Practice (TTP) program, which supports efforts to translate research discoveries into practical tools that benefit communities, industry, and society. For the TTP program, teams advance research results toward real-world deployment and adoption. Flooding is one of the most costly and dangerous natural hazards in the United States. Levees play a critical role in protecting more than 23 million people, millions of acres of farmland, and trillions of dollars in property from dangerous floods. However, many levees are aging, were not originally designed for today’s demands, and are monitored using outdated methods that rely on infrequent inspections and limited data. This research team develops a new, smarter way to monitor and manage levees using a “digital twin” which is a simulation of anticipated levee function that continuously updates using real-world data about levee performance. Combining advanced modeling with real-time information enables earlier detection of flood risks, better maintenance decisions, and improved emergency preparedness. The researchers partner with federal agencies, local levee districts, and industry partners to ensure the tools are practical, affordable, and ready for use. Ultimately, this project helps reduce flood risk, protect communities and infrastructure, and ensure taxpayer investments in flood protection systems are effective. This is the first levee-specific hybrid digital twin framework that integrates physics-based models with data-driven machine learning to provide probabilistic, near real-time predictions of levee performance. The researchers develop multi-scale models that simulate key failure mechanisms such as seepage, slope instability, and overtopping at both structural and regional levels. A central innovation is the integration of these models with monitoring data through recursive updating and machine learning–based error correction, improving predictive accuracy