CAREER: Enabling Human Situation Awareness across Physical and Cyber Spaces for Vehicle-to-Everything Communication Security and Driving Safety

Vehicle-to-everything (V2X) communication is one of the key pillars upon which connected and automated vehicles rest. In V2X, vehicles communicate directly with other road users and infrastructure, allowing perception of road conditions and driving environment beyond their own sensors, in turn supporting safer and more efficient transportation. However, V2X can also raise security issues in which malicious actors compromise a vehicle to send false or fabricated information that could mislead driving decisions and actions made by both autonomous vehicles and human drivers. This project’s goal is to better understand the risks V2X attacks may pose to drivers' situation awareness -- i.e., their perception of elements in the situation, their comprehension of what is happening, and their ability to respond to road hazards or autonomy failures -- and to develop warnings, explanations, and other interactions with drivers to increase the situation awareness. In particular, the project will focus on cases where the V2X attack sends information that is inconsistent with the physical environment, using those discrepancies to enhance drivers' vigilance for not only situation-specific cues but also the presence of possible attacks. Through this work, the project team will increase understanding and safety in autonomous vehicles that still require drivers to interact and intervene.This work consists of three research thrusts. First, the project will develop usable and trustworthy V2X warning interfaces to enhance drivers' situation awareness of their physical vicinity during automated driving. The project team will leverage a suite of approaches in human-centered design, using comprehensive evaluations that include online studies and driving simulator-based experiments. Second, to mitigate information integrity attacks on V2X communication, this project will systematically generate and assess driver-centered counterfactual explanations that integrate information from the physical environment and cyberspace to identify coherence or discrepancy between the two spaces, across a wide range of driving contexts and dynamic situations. The third thrust seeks to establish shared situation awareness between drivers and automated driving systems through human-in-the-loop reinforcement learning that leverages how drivers respond to and reason about the counterfactual explanations to identify and correct false V2X messages. The research activities will be integrated with educational activities through interdisciplinary curriculum development, hands-on training for graduate students, research experience for undergraduates, and K-12 outreach programs designed to inspire and train the future workforce in the areas of usable security and privacy.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.