About

The Risk-Based Prioritization of Critical Valves Using Limited Utility Data and Machine Learning project seeks to advance critical valve management through innovative, data-driven risk assessment. The program will develop a data-efficient machine learning model to estimate valve failure likelihood using available utility data such as asset attributes, condition indicators, environmental factors, and geohazards. It will also integrate consequence-of-failure analysis through hydraulic modeling and network connectivity to evaluate service disruption, affected populations, outage duration, and community impacts. By incorporating uncertainty in asset condition and operability, the project will establish a practical risk-based framework for ranking critical assets to guide maintenance, rehabilitation, and capital investment planning. Additionally, it will define minimum data requirements for reliable prioritization. Applicants may request up to USD 300,000 in funding from WRF. The project duration will follow the schedule requirements established for WRF research projects.

Eligibility

Proposals are open to U.S.-based and non-U.S.-based entities, including educational institutions, research organizations, governmental agencies, consultants, and other for-profit entities. All researchers must comply with WRF’s Timeliness Policy regarding adherence to project schedules. Applicants with delayed ongoing WRF-sponsored studies lacking approved no-cost extensions are not eligible to participate in proposals.

Post Date: July 29, 2026