Project Grant 2530256

Project Grant Summary

The University of Miami received a $400,000 project grant from the National Science Foundation's Division of Information and Intelligent Systems (CISE Program, CFDA 47.070) awarded February 1, 2026, with completion targeted for January 31, 2029. This collaborative research initiative develops artificial intelligence (AI) methods and computational frameworks to enhance the management of urban water and wastewater networks. The primary deliverables include machine learning models, graph-based network representations, and physics-informed algorithms designed to integrate fragmented infrastructure data, identify missing network connections, and detect anomalies such as leaks, blockages, and structural deficiencies in water systems. The tools leverage attention-based neural networks and multiresolution uncertainty quantification to provide interpretable confidence assessments for inferred network characteristics.

The project addresses critical operational challenges in water infrastructure management by transforming unstructured inspection data and incomplete geographic information system records into coherent, actionable network models. By automating data analysis and reducing manual review requirements, the developed technologies enable early detection of system failures, support data-driven decision-making, and help utilities reduce operational costs and service disruptions. The research advances computational methods for completing and repairing directed network representations while improving overall water system reliability and community environmental protection outcomes.