I think decision making in driving is highly contextual, but LiDAR doesn’t help there either. Purely visual field extraction is something even very simple animals can do (presumably which much weaker abstract context processing capabilities).