Project Grant 2137753

This three-year, $309,811 project grant from the National Science Foundation's Engineering program (CFDA 47.041) will fund research at The Trustees of the Stevens Institute of Technology to develop techniques for improving the resilience of vision-guided unmanned aerial vehicles and other mobile robotic technologies against cyberattacks. The research aims to advance a framework for detecting and responding to stealthy attacks that simultaneously target mission planning, control, perception, and sensor data for these autonomous systems. It will characterize stealthy attacks based on unobservable subspaces of physical system dynamics and neural networks used for perception. Evaluation will rely on experiments with quadcopters and advanced simulation. Broader impacts include integrating research into graduate courses and engaging undergraduate and K-12 students in STEM education opportunities using mobile robotic platforms. The goal is to make rigorous contributions to using control theory and machine learning tools to characterize and defend against stealthy attacks exploiting vulnerabilities in networked autonomous systems dependent on visual data.