Not so sure about this one. > Even if you record every road with cameras in the world, a computer will make mistakes without radar or lidar, because he can not

Not so sure about this one. > Even if you record every road with cameras in the world, a computer will make mistakes without radar or lidar, because he can not understand certain things without depth data. 2 cameras or more gives the same amount of depth information that the human brain gets. It's the brain that's doing the depth processing, this is an unsolved computer vision problem (but much progress has been made).