Got my first long drive on 48 yesterday with a round trip out to Vandenberg to see the SpaceX/Iridium launch. In some ways the trip out was great test conditions for the AS vision system because it was near sunset driving straight into the sun with lots of long shadows across the road, high contrast at shadow edges but very flat light otherwise. Very tough conditions visually and just the time of day when the AS on AP1 gives me the most problems. Lane markings on the majority of the route are faded white stripes on white concrete. AP2 48 drove from my hood (San Gabriel Valley) out to the end of the 101 segment with only one intervention needed, which was in a reduced speed construction zone that had a temporary tight curve/lane change. So that's 150 odd miles in varying traffic (mostly moderate traffic with about 20 minutes of stop and go) in very challenging lighting conditions with only a single fairly forgivable intervention. I did this drive on AP1 last June under sunny midday conditions and AP1 did almost as well but I'm pretty confident it couldn't pull that same performance off at dusk.
I wish the story ended there but the drive back was after dark on the same route and required several interventions - I lost count but maybe 8 or 10? It seems like AS for AP2 48 isn't liking conditions where you can't see much of anything except the reflective lane markers and sometimes you lose sight of them due to rises or curves or other vehicles blocking sight lines.
That's an interesting contrast point for me because my sense on AS for AP1 is that nighttime is one of it's strongest performance points. And of course, that fits with my theory that AP1 is more dependent on and more sensitive to lane markers for it's AS capabilities. At night lane markers are super high contrast because they are reflective so that's better for AP1 than it is for AP2 if my conjecture is correct.
I saw a lot of other Tesla's at the launch. Anybody else have notes to compare (on the drive, I mean)?