Elon Musk has perpetually promised that self-driving cars are only a year away. However, today’s relatively rudimentary autonomous systems are still making errors that most humans wouldn’t make behind the wheel. New research found that the flashing lights mounted on emergency vehicles can disorient automated driving systems. Researchers have named the issue “digital epileptic seizure.”
Digital epileptic seizures (or epilepticars) make it impossible for systems trained with AI to identify objects on the road properly. Researchers at Ben-Gurion University of the Negev and the Japanese technology firm Fujitsu Limited ran tests using five off-the-shelf systems. The flashing lights essentially blow out the images captured by the camera, making object detection unreliable. According to Wired, researchers did propose a solution:
The BGU and Fujitsu researchers did come with a software fix to the emergency flasher issue. Called “Caracetamol”—a portmanteau of “car” and the painkiller “Paracetamol”—it’s designed to avoid the “seizure” issue by being specifically trained to identify vehicles with emergency flashing lights. The researchers say it improves object detectors’ accuracy.
While researchers didn’t test Tesla’s Autopilot or the systems mounted on any specific vehicle, digital epileptic seizures could explain why Teslas seemingly crash into emergency vehicles far more often than cars from other automakers. Up to 2023, at least 15 crashes of this nature where Autopilot was involved. After Autopilot’s recall, Full Self-Driving is still causing Tesla vehicles to smash into police cruisers, fire trucks and ambulances. Caracetamol could be the medicine for Tesla’s software woes.
