Love itWe’ll know as soon as Omar posts the obligatory “OMG this version is so great! ZERO disengagements!” video.
Since every version is perfect with zero disengagements it’s unclear to me how the next version can be any better.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
FSD Beta 10.12
- Thread starter EVNow
- Start date
You forgot "smooth".We’ll know as soon as Omar posts the obligatory “OMG this version is so great! ZERO disengagements!” video.
Since every version is perfect with zero disengagements it’s unclear to me how the next version can be any better.
Is it me or is this release progressively getting worse as the next one is approaching. Mines been doing a lot of harder brakes instead of utilizing regen as much and for the last couple of days it’s been taking super wide turns crossing yellow lines, left hand and right hand.
Maybe it’s all in my head, maybe it’s just too smooth for me.
Maybe it’s all in my head, maybe it’s just too smooth for me.
VanFriscia
Member
I agree with others who say it doesn't really make sense for the same version of FSD to improve or degrade over time, but I also have to agree that 10.11 at least SEEMS to be getting worse for me too. Maybe it's just variance in how the car happens to perceive things each time, but if so, there's too much variance! At some point in the future I hope things will at least become more predictable. Right now, I feel like I have no clue how my car's going to handle anything on FSD.
One example off the top of my head - last week, a car was parked on the side of the road with its blinkers on, sticking out a bit into my lane. There was no oncoming traffic in the other lane. As expected, FSD began to hesitate and applied the brakes as it approached. I expected it to go around the vehicle, as it's done several times in the past. Instead, after the initial hesitation, it barreled full speed ahead in a straight line. Like, how could it possibly have come to THAT conclusion based on what it was seeing? On the same version of FSD that's successfully handled similar situations before? Maybe it would've avoided a collision had I let it go, but I wasn't about to find out...
As stated before 10.11 for me is like a petulant 2 year old. One never knows what's coming next. A recent example that just started 2 weeks. After taking a normal right hand turn at an intersection I've taken many times FSD now stops in the road after taking the turn. First time it got me by surprise but now I have my foot ready to push the accelerator pedal before it comes to a full stop especially with someone right behind me. Happens every time now.
Other recent oddities make me wonder what is going on since we know FSD isn't actively learning.
Yea I have a couple of those strange instances where it goes through a traffic light and then just regen stops in the middle of the road in a 55mph zone. Pressing the accelerator gets it going, no warnings or messages in either screens.
I’m starting to think it might not be “learning” but there is some data somewhere getting messed with. Maybe map data?
Could be map related. I know several of my most common routes to my home now take different exits off the interstates. While it's a different exit the time difference to my home is insignificant. Map version number hasn't changed in a couple of months either. Of course we can all speculate but we will never find out what the answer is unless in a few years a Tesla FSD engineer decides to write a tell all book on the effort.
ChaoticNeutral
Member
Map issues are a very much a thing. I’ve run into multiple instances where FSD will think the speed is different than what is posted for a short amount of time. Suddenly increases by 5 mph or drops by 10mph even though there’s no speed sign. Sometimes it will read a sign across a median, but that’s not what’s happening in most cases, from what I’ve observed.
I have a mysterious section in a city where the speed limit increases 5mph in one direction but is the correct 25mph traveling the opposite direction. I carefully watched to see if it was corrected with the map update but was not.
And recently, it changed to the incorrect 30mph in both directions! (clear signs it is 25mph previously in the good section; the bad stretch has none). So something new happened. I will have to see if the sign visualizations have disappeared on the previously good section.
And recently, it changed to the incorrect 30mph in both directions! (clear signs it is 25mph previously in the good section; the bad stretch has none). So something new happened. I will have to see if the sign visualizations have disappeared on the previously good section.
Last edited:
beachmiles
Member
Wish we knew exactly which maps Tesla is using at this point for navigation. I have extensively updated problem streets in openstreetmaps as well as reporting map issues with Google maps, tomtom, Mapbox, and here maps. In my FSD snapshot companion emails I include map coordinates of issues that I have been having the entire 6 months I've had FSD. Arg.
I hope it’s not the Apple Maps. I have reported my house address being wrong for a year and 3 months now, at least once or twice a month, and it’s still not fixed. It’s a new built home. The ironic part is that the street name on the map is correct, but my house has a number that’s wrong and a street name that’s also wrong. That street is like a mile away from my house and doesn’t intersect my street at any point
Todd Burch
14-Year Member
Pretty sure it’s OpenStreetMaps, but there is a delay between when it gets fixed to when that updated database gets pushed to the car.
The FSD system may very well degenerate overtime, just like your windows pc and has to be fixed by a reboot. Small memory leak or lack of optimization here and there will slow down the perception which gives you larger lag and as a result, worse performance.
Speed limit data in my neighborhood is most likely coming from TomTom. They are absurdly incorrect (labeled 50mph but actually 25mph), and the car shows the limit at 50mph. I've changed it a while back, but it hasn't been picked up yet.
I have yet to find any other mapping source that even labels the speed limit on these roads, including OSM.
Rusty1
Member
Elon needs to clarify the 100,000 number. He said 60,000 at the end of December. No one has been added since then in the US and I don’t think there were 40,000 Canadians added. BTW I am one of those waiting since the second week of January.
Rusty1
Member
sleepydoc
Well-Known Member
I’ve seen some other posts indicating Tom Tom. There’s a road I travel that’s actually 55 but labeled 35. Changed it a year ago I. Open Street Maps and haven’t seen any change in Tesla. Submitted a change in Tom Tom recently that’s still pending so we’ll see.
jabloomf1230
Minister of Silly Walks
All these recent posts show is that no one outside of Tesla knows what source data are used for the navigation.
The routing is good enough for getting from Chicago to Des Moines, but it is a hot mess locally. The routing algorithm appears to try to shave a minute or two off each trip by taking back roads and side streets. At least for where I live, it rarely picks the route that I would have chosen. And Tesla's implementation of waypoints is terrible.
But all that could be a criticism of any of the existing consumer GPS navigation systems. They are better than nothing, but far from perfect. IMO, navigation data and routing algorithms combined are the Achilles heel of autonomous driving. They don't pose a safety risk per se, but they are a barrier to robotaxi adoption.
The routing is good enough for getting from Chicago to Des Moines, but it is a hot mess locally. The routing algorithm appears to try to shave a minute or two off each trip by taking back roads and side streets. At least for where I live, it rarely picks the route that I would have chosen. And Tesla's implementation of waypoints is terrible.
But all that could be a criticism of any of the existing consumer GPS navigation systems. They are better than nothing, but far from perfect. IMO, navigation data and routing algorithms combined are the Achilles heel of autonomous driving. They don't pose a safety risk per se, but they are a barrier to robotaxi adoption.
Similar threads
