Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Explore the security vulnerabilities of Normal Distributions Transform (NDT) scan matching in autonomous driving localization systems through this 11-minute conference presentation from VehicleSec '25. Examine how NDT-based localization methods, critical for accurate autonomous vehicle positioning, can be compromised through gradual manipulations of LiDAR point cloud structures from pre-built maps. Learn about the research methodology that simulates real-world scenarios using sensor fusion with Extended Kalman Filter (EKF) to evaluate the impact of these security threats. Discover the key factors that influence localization errors, including target object selection and movement patterns, and understand how these vulnerabilities can result in localization errors of up to 23 meters. Analyze the dangerous consequences of these security flaws, including lane departures, missed traffic signals, and unintended sidewalk encroachments that pose significant safety risks. Gain insights into the systematic analysis of NDT-based localization vulnerabilities and the urgent need for developing more robust localization mechanisms to ensure the security and safety of autonomous driving systems.
