Federal Grant Award Summary

The National Science Foundation (NSF) Directorate for Engineering awarded $224,996 to the New Jersey Institute of Technology (NJIT) on July 1, 2026, for a collaborative research project titled "Reinforcement Learning with High-Probability Safety Constraints: Theory, and Applications" (CFDA 47.041). The three-year project, concluding June 30, 2029, will deliver foundational theory, algorithms, and software tools that enable safe reinforcement learning (RL) for cyber-physical systems such as unmanned aerial vehicles, mobile robots, and networked infrastructures. The core deliverables include novel learning methods that enforce explicit probabilistic safety guarantees under model uncertainty, limited data, and changing operating conditions; open-source software implementations and benchmark scenarios to advance safe autonomy research; and peer-reviewed publications documenting theoretical advances in high-confidence autonomous decision-making.

Beyond technical outputs, the project will produce educational deliverables including curriculum modules and student research experiences designed to broaden participation in STEM education and train the next generation of researchers in safe autonomous systems. These outputs directly address the gap in current RL methods, which typically optimize only average performance and provide insufficient safety assurances for safety-critical deployments. The research is expected to yield practical advances in reliable autonomy relevant to national health, prosperity, defense, and other strategic priorities, while lowering barriers to safe autonomy research through disseminated tools and methodologies.