Federal Grant Award Summary

The University of Florida Division of Sponsored Research received a $110,359 Project Grant from the National Science Foundation (NSF) Division of Computing and Communication Foundations under the Computer and Information Science and Engineering program (CFDA 47.070) effective October 1, 2025, through September 30, 2026. This award supports continued development of STARV (Spatial-Temporal And Riskiness Verification), a quantitative verification tool designed to address safety and reliability concerns in learning-enabled cyber-physical systems (LE-CPS) that incorporate data-driven machine learning components. The project extends collaborative efforts with industrial partners, particularly in the automotive industry, to develop novel verification methodologies that quantitatively assess temporal properties and safety metrics—such as probability of collision—under real-world uncertainties in sensing, perception, and actuation.

The deliverables include the development of ProbStar Temporal Logic (PSTL) for specifying complex temporal behaviors of LE-CPS, along with qualitative and quantitative verification algorithms operating at the system level based on ProbStar reachability analysis. The project emphasizes practical technology transition through enhanced user interface design, documentation improvements, benchmark development, and community engagement to facilitate broader adoption of the verification technology. This research directly addresses a critical gap in formal methods for machine learning certification, providing industrial stakeholders with quantitatively rigorous assurance tools necessary for autonomous system deployment and regulatory compliance.