AUSTIN, Texas — Tesla is zeroing in on one of the final frontiers for its autonomous ride service: driving at night. Elon Musk says the main reason robotaxis have stuck largely to daytime operation is the challenge of reliably spotting pets and other small animals in low light, and he has signaled openness to adding lidar as a tool to clear that last hurdle.
The Last Mile Is After Dark
Tesla's robotaxi fleet has racked up paid, driverless miles in daytime conditions, but extending that to full 24-hour service raises the bar on perception. Musk explained that the edge cases keeping the system cautious at night are not cars or pedestrians, which the vision stack handles well, but harder-to-see hazards like a dog darting across a dim street. Solving that last category is what stands between limited hours and true round-the-clock autonomy.
It is a notable sign of pragmatism from a company long associated with a camera-first philosophy. Musk's willingness to consider lidar for specific low-light tasks shows Tesla optimizing for safety and capability rather than ideology. That momentum builds on the robotaxi push Musk showcased alongside policymakers, covered in our report on his turn at the White House AI summit.
Vision First, With Help Where It Counts
Tesla's approach has always centered on cameras and neural networks trained on an enormous fleet of real-world miles. The system excels in daylight and handles complex urban scenes, and the company argues that scale of data is its durable advantage. Adding a targeted sensor to backstop the hardest nighttime cases would complement that strength rather than replace it, giving the car an extra layer of confidence exactly where human drivers also struggle.
