I've been borrowing a Tesla Model S (w/o autopilot enabled) for a couple of weeks and the weather has been dark, raining and today it was snowing.<p>Every day has the car complained of sensors/cameras not working due to dirt in front of the sensors. It doesn't matter if I clean them before I head out, it takes a couple of kilometers before they are warning again. They do add salt a lot on the roads to prevent ice, which makes it very dirty on the roads here.<p>Today, when there was snow. It warned again, same this time, a lot of dirt due to slush. And it was also black ice, which was really hard to spot myself (looks like dry road).<p>So to my question; How would the Tesla (or any other autopilot enabled vehicle) work on autopilot in these cases?<p>I mean, if the on-board computer doesn't get good input, it must be very hard to take decisions?
How do humans operate in those conditions? Reducing speed and watching carefully: if anything, sensors need to move to a position where they are better shielded, eg. some can go behind a windshield, exactly where human eyes sit.<p>Or they could be in a better self-cleaning container (a la windshield wipers and xenon headlights).