Investigate the Effects of Laser Attack on Intelligence of the AV Perception

Autonomous Vehicles (AV's) are rapidly advancing in capability and deployment, heavily relying on camera sensors for navigation and environmental perception. This raises significant security concerns, related to sensor attacks, e.g. laser attacks on camera, LiDAR sensors, GPS spoofing, etc. This study investigates how laser attacks impact camera sensor that can in turn affect the perception capability of autonomous vehicle (AV). This called a blinding attack (i.e. AV's cannot perceive these objects properly). Our analysis uncovers critical vulnerabilities, such as susceptibility to misinterpret traffic cues under blinding attacks, and categorizes the disruptions these vulnerabilities cause. The findings highlight substantial risks to vehicle safety and operational reliability and underscore the urgent need for robust defensive strategies.