Another area of construction in North Phoenix. Here #TeslaFSD is turning from one construction lane into another, navigating on the wrong side of the lane, and doglegs through a light into the correct lane. I am sure the cars ahead helped to guide it but it’s still impressive for a computer.

Just a ton of construction in and around the Phoenix area and #TeslaFSD manages it just fine. It makes one little mistake at the end that any driver would, were the exit lane is blocked right before the exit.

Pulling out of the parking lot for Trader Joe’s in Phoenix, there is a lot going on here. There is construction all over the place, people walking besides and in the road, and a lane closure on Shea Blvd. This time, #TeslaFSD was able to make all the right choices.

Look Ma .. no lane lines. Through this section of road, #TeslaFSD does a pretty good job staying within the right areas of the lane and even stops at the light without a “stop line.”

In this obviously sped up video, #TeslaFSD is all over the road. For some reason its latest thing when it gets to this part of the I-10 west of Phoenix, it moves out of the fastest lanes, into the slowest lanes before going back to the fastest lanes.

Without much warning, at 80 mph, #TeslaFSD managed to see the debris in the road and adjust for it without crowding the next lane. It is a little nerve-racking when you’re holding the steering wheel at that speed and it decides to move on its own.

I don’t normally try this because it rarely works. It certainly doesn’t in my building. But it did this time. Leaving the main parking garage in downtown Palm Springs, #TeslaFSD successfully navigates through the crowds and into the street. Fortunately, there was no gate. I still call that luck.

Even in a fairly fast speed zone, #TeslaFSD easily handles this car parked on the side of the road. There is room to move over in the lane and so it does although I would have given it way more latitude in case they opened their door.

I’ve always felt there are some real safety opportunities with automated driving and one of them is to better anticipate the needs of other drivers. Coming up on a cyclist, the driver in the white car is going to have to move over. But #TeslaFSD consistently stays in that driver’s blind spot.

I wasn’t sure what #TeslaFSD would do at the AZ border crossing into CA. But I let the system do its thing and it worked perfectly. I am sort of curious why it didn’t stop though. There was a stop sign (although no line on the pavement) but the guy did wave us through. Is it reading hand gestures?

I’ve never seen resurfacing work like this before or having any road construction take this long. But it does amaze me that #TeslaFSD can successfully navigate with a road surface like this and still find the lane lines.

As near as I can tell, #TeslaFSD mistakes (or is overly cautious?), this lump of cement that’s been there forever, as some obstacle to avoid. What’s worst is, instead of avoiding it, the car runs over it. Had it been something real, there might have been material damage to the object or the car.

I get it. Mistakenly exiting onto the 202 when you wanted I-10, would put you out of your way and would take a lot of time to get back on track. However, watching that Hyundai in front cross three lanes and two exit ramps, makes me think they are the poster child for why we need autonomous driving.

Given a recent post, it appears I am wrong (obviously harsher language was used) thinking that drivers in a merge lane should be a little more accommodating to vehicles entering the highway, at speed. I still don’t think I am wrong to want a little more cooperation and this guy proves me right.

It’s still really hit or miss on which lane #TeslaFSD chooses when making a multi-lane turn. In this case, it used the correct lane, because I’ll be making a left turn down the street. But it should have made a tighter turn into the left lane, just like the white Jeep behind me.

@waymo.bsky.social is really spinning up the testing of their new vehicle today. I saw no less than a half a dozen vehicles in and around the downtown neighborhood on my evening walk. I assumed they were empty, but it was hard to see inside.

Ugh! Drivers in the right lane on a major highway that don’t yield to oncoming traffic totally frustrate me. But #TeslaFSD does handle it by slowing down and moving to the right until they pass. Times like these make me wish for more autonomous vehicles.

While these pedestrians are no where near the crosswalk, #TeslaFSD recognizes them, even though they are almost out of camera range, and anticipates them walking towards it, slowing down to let them cross.

This is a complicated intersection in Phoenix. It's two directions to the right and a single direction to the left, which is further complicated by a bike lane, with the travel lane on the opposite side of the street. So many accidents but #TeslaFSD handles it well. That's downtown Phoenix for you.

I do admit that I sometimes put too much faith in #TeslaFSD, even while being hyper-vigilant behind the wheel. Crossing a double-yellow line, while technically illegal, is unfortunately the only option here. Although, I might have waited until after the oncoming 18-wheeler. It's still impressive.

It is nice to see #TeslaFSD making some improvements. Although, it's typically two steps forward and one step back. It's finally getting into the correct lane off the highway onto 7th St. so it can make a left turn onto Roosevelt. We will see if that holds as they continue to release new versions.

It is fun to watch #TeslaFSD monitor all directions at once before calculating it's safe. Backing up into open traffic is particularly challenging. Cars are coming from all directions, from the side, from behind, all very close to an intersection and through the trees and it handles it flawlessly.

Coming up to this crosswalk, #TeslaFSD appears to stutter but really, it's reacting to the flashing light that turns off as soon as it approaches the crosswalk.

I would have a lot more confidence in #TeslaFSD if they were more transparent about what it's seeing and why it's reacting. I am not sure if the system saw the pedestrian on the right (they were somewhat obstructed), was reacting to the flashing lights (which were not red) or a combination.

Navigation typically avoids downtown because it presents so many complex navigational situations. Here there are emergency vehicles (complete with flashing lights) blocking the road, a Jeep turning from the opposite way and the car in front making a turn. #TeslaFSD handled it flawlessly.

Not knowing what someone is going to do is tricky. #TeslaFSD tends to be overly cautious, which is a good thing. I wasn't sure what this guy on the bike was going to do but once they stopped and wasn't going through the crosswalk, which he had the right, we moved on to navigate this tight turn.

Saw the Rivian R2 in a showroom in Scottsdale. If you're looking for a new EV (cough Tesla), and you don't seriously consider the R2, you are really doing yourself a disservice. The vehicle we saw was stunning. Hard to believe it's the same size as the Model Y. It felt way bigger inside and out.

For all the things that go wrong, and there are many, it's still fun to watch when it does complicated tasks. Here, #TeslaFSD successfully navigates out of a space in Trader Joe's parking lot, crosses a busy six lane roadway and pulls into a space for the Fry's across the street.

Tesla Superchargers in Quartzite, Arizona were very busy on Sunday.

It is impressive that #TeslaFSD can successfully navigate very tight spaces with loads of moving parts like cars, shopping carts and people. It definitely navigates slower than I would.