Assured Neuro Symbolic Components and Systems (ANSCS) - DAINS

Dependable AI for Autonomous Systems

This project focuses on making artificial intelligence (AI) safer, more reliable, and more trustworthy for use in high-stakes environments such as national security and autonomous systems. Today’s AI-powered systems—like drones or other autonomous vehicles—often still rely on human oversight, especially because current AI models can be vulnerable to errors, cyberattacks, or unexpected situations. This project aims to change that by developing new ways to test, evaluate, and strengthen AI so it can operate more independently and with greater confidence.

To achieve this, the team is building tools, simulations, and real-world prototypes that allow researchers to better understand how AI systems behave under both normal and challenging conditions. This includes creating a testing framework, developing a small-scale autonomous drone platform, and designing methods to identify weaknesses in AI systems before they are deployed. By combining research, practical demonstrations, and collaboration with government and industry partners, the project helps pave the way for more dependable AI systems that can be safely used in critical, real-world scenarios.