Federal Grant Award Summary

The University of Connecticut received a $250,000 Project Grant from the National Science Foundation's Division of Information and Intelligent Systems (Computer and Information Science and Engineering program, CFDA 47.070) awarded January 15, 2026, with completion targeted for December 31, 2028. This collaborative research project develops physics-informed probabilistic forecasting algorithms and tools designed to enable reliable system diagnostics and lifetime prediction for safety-critical engineered systems operating in small-data environments. The project addresses a critical gap in existing forecasting methodologies by combining physics-based modeling with machine learning approaches to handle scenarios characterized by limited experimental data, scarce field failure events, and imperfect physics simulations that diverge from real-world system degradation patterns.

The deliverables include the Modular Analytics for Prognostics with Small Data (MAPS) platform, a comprehensive physics-informed probabilistic prognostics system comprising multiple integrated modules that utilize sparse symbolic regression to derive degradation models and recover unmodeled physics. The platform will support rapid, risk-based decision-making on quality control and maintenance by providing probabilistic lifetime predictions before degradation becomes observable through standard monitoring. The project also commits to advancing scientific dissemination and STEM education through development of open-source tools, modules, and dedicated education and outreach activities, enhancing accessibility of probabilistic forecasting methodologies for both research and practical application communities.