Tesla Extends Robotaxi Hours and Tackles Low-Light Pet Perception Issues
Elon Musk announced extending Tesla's Robotaxi operating hours in Austin until 11 PM while highlighting ongoing technical efforts to improve low-light AI vision perception for small animals. Tesla plans to utilize AI-assisted photon-counting analysis on raw camera data to detect low-contrast targets, such as gray cats on dark asphalt at night. Low-light edge cases remain one of the most critical technical hurdles for vision-only autonomous driving systems competing against LiDAR-backed rivals like Waymo. Successfully improving night perception for unpredictable obstacles could enhance confidence in Tesla's Full Self-Driving (FSD) system, especially amid ongoing safety scrutiny by federal regulators. Tesla relies on optical cameras and deep neural networks rather than expensive LiDAR sensors, which the company acknowledges can be limited in poorly illuminated environments. Meanwhile, the National Highway Traffic Safety Administration (NHTSA) is investigating Tesla's FSD system to assess how effectively it detects road hazards in reduced-visibility conditions.
## BACKGROUND
Autonomous driving systems rely on sensors like optical cameras, LiDAR, and radar to perceive their surrounding environment in real time. While competitors like Waymo use LiDAR laser scanners to measure distances regardless of ambient lighting, Tesla relies strictly on vision-based camera feeds processed by computer vision models. Low-light conditions present severe difficulties because optical sensors capture fewer photons, making low-contrast objects difficult for AI algorithms to differentiate.