Waymo CEO explains why Tesla’s camera-only self-driving falls short

This is a car story, not a home energy one, but here is the plain summary.

Waymo's co-CEO Dmitri Dolgov argued that self-driving cars relying only on cameras, as Tesla does, can match human driving or power a driver-assist feature, but can't reach the reliability needed for a car to drive itself with no one behind the wheel. His reasoning: cameras struggle in darkness, glare, dust, or fog, and a single leaf or branch covering a lens can stop the car entirely. Waymo instead combines cameras with lidar (a sensor that maps 3D shapes) and radar (which sees through bad weather and reads speed directly), so if one sensor fails, the others still work.

He also pointed to a "nines" problem: getting to good performance is easy, but each added bit of reliability afterward takes roughly ten times more effort, and camera-only systems tend to plateau before reaching the level needed for full autonomy. Tesla dropped radar and other sensors years ago to bet fully on cameras, and its robotaxi service, running in Austin, has logged about 380,000 driverless miles with a crash rate reported to be roughly three times worse than human drivers. Waymo, by contrast, has logged 220 million driverless miles and reports far fewer serious crashes than human drivers over the same distance, while now serving roughly half a million paid rides a week in 15 U.S. cities.