During a recent presentation at the Network and Distributed System Security Symposium in San Diego, researchers discussed the findings of their study on spoofing attacks targeting nine commercially available LiDAR systems.
The research team, led by the University of California, Irvine, found safety flaws in both first-generation and later-generation LiDAR versions.
The study received financial backing from the U.S. Department of Transportation’s CARMEN+ University Transportation Center and the National Science Foundation.
The researchers demonstrated that LiDAR, a technology used by many autonomous vehicles for navigation, exhibited potentially dangerous vulnerabilities.
Computer scientists and electrical engineers from UCI and Japan’s Keio University showed how lasers could deceive LiDAR into detecting non-existent objects and failing to register actual ones, leading to unwarranted and unsafe braking or collisions.
The researchers highlighted LiDAR as a preferred navigation and sensing technology utilized in robotic taxis operated by Google’s self-driving car brand Waymo, as well as in General Motors’s Cruise.
It is an essential component in consumer-operated models sold by Volvo, Mercedes-Benz, and Huawei.
In their testing of first-generation LiDAR systems, the team carried out an attack known as “fake object injection,” tricking sensors into registering a pedestrian or the front of another vehicle when none were present.
This led the LiDAR system to relay false hazards to the autonomous vehicle’s computer, triggering unsafe behavior such as emergency braking.
Furthermore, the UCI and Keio University researchers found a way to deceive next-generation LiDAR systems.
Using a custom-designed laser and lens apparatus, they were able to conceal five existing cars from the LiDAR system’s sensors.
IBM announced this week that its apprenticeship program has earned…
The U.S. Army Corps of Engineers has been tasked with…
Brown and Caldwell, a leading environmental engineering and construction firm,…
