Tesla Autopilot Failed to Detect Painted Road Wall in YouTuber Test
YouTuber Mark Rober conducted demonstration tests of Tesla Autopilot that reportedly revealed vulnerabilities in the camera-based system compared to LiDAR alternatives. The tests showed the system failed to detect obstacles under adverse weather and when presented with visual illusions, striking a Styrofoam wall at 40 mph while a LiDAR-equipped vehicle halted safely.
What happened
YouTuber Mark Rober conducted tests comparing Tesla Autopilot to a LiDAR-equipped Lexus prototype. In the first two tests with a mannequin on the road, both systems successfully detected and stopped for the obstacle. When fog and rain were added, Tesla Autopilot failed to detect the mannequin while the LiDAR system continued to function.
In a third test, Rober used a Styrofoam wall printed with an image of a road. The LiDAR-equipped vehicle halted before the wall, but the Tesla Model Y running Autopilot struck the wall at 40 miles per hour.
Reported impact
The incident highlighted potential safety vulnerabilities in Tesla's camera-based vision system compared to LiDAR. The test results suggest the system may fail to recognize obstacles under adverse weather conditions or when visual cues are ambiguous.
Index analysis
Tesla discontinued LiDAR and radar sensors starting in 2021, relying instead on camera-based vision. The decision was attributed to cost, weight, and space considerations rather than safety prioritization.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.