Federal Project Grant Award Summary

Rutgers, The State University received a $395,821 Project Grant award from the National Science Foundation's Division of Information and Intelligent Systems under the Computer and Information Science and Engineering program (CFDA 47.070), effective May 1, 2026 through April 30, 2031. This CAREER award supports research on structured learning and verification of control policies for linear temporal logic (LTL) objectives, with the primary objective of enhancing the reliability, safety, and interpretability of intelligent robotic systems performing complex long-horizon tasks. The research deliverables include the development of a principled framework that embeds temporal logic constraints directly into reinforcement learning objectives, mechanisms for automatic inference of temporal task specifications, and compositional verification methods for validating agent behaviors in continuous, stochastic environments.

Beyond core research outputs, the award generates reusable algorithmic capabilities and software tools for specification inference and compositional verification that are designed for integration into existing artificial intelligence and machine learning development pipelines. The project also includes an educational component that brings formal methods into K-12 and undergraduate classrooms through an interactive robot simulation tool, preparing a workforce trained in formal methods for trustworthy systems development. These deliverables collectively address the need for verification components in learning-based systems while advancing the accessibility of formal methods for non-expert users.