-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
FSD v12.x (end to end AI)
- Thread starter Buckminster
- Start date
-
- Tags
- #fixphantombraking autopilot chill is not chill elon is an ass fix fsd extreme late deceleration fsd does not commit to go fsd does not maintain proper speed fsd hard and late braking fsd jerky driving fsd late deceleration fsd stops too early too quickly fsd stops too far behind line fsd tailgating in city streets
Seriously, what's your problem with my posts? You've continually been kind of a problem to put it nicely.
Ford sells "Ford BlueCruise Hands-Free Highway Driving" for highway driving only. Don't think it's a $12,000 add on that promises (one day) autonomous driving, does it? It basically does what Autopilot does for free. Yes, but it takes 15 years to get to FSD level at that cost. So moving the goal posts how exactly?
Last edited:
How is my response personal or any indication of a problem. Blue is a highway only option that has a cost as someone posted. Similar to AP which is free (cost) with Tesla. No not hands free nor was that implied. Definitely Not FSDb as Blue would not have the same function off highway. I respond to those that post. You post a lot in threads I have interest in so if you see that as a “problem“ because I don’t always agree that’s not the purpose of a public forum. No hate just a different view.
powertoold
Active Member
I use FSD beta hands-free. I only put a hand on the wheel when there's higher risk or nagged
Even though Blue's clues may be "hands free" it's definitely not nag free, so it's a tomato tomato situation to me when compared to fsd or ap
Even though Blue's clues may be "hands free" it's definitely not nag free, so it's a tomato tomato situation to me when compared to fsd or ap
Hardly. I post why I and others dislike the lack of center horn and you find it "troubling" a moderator of all people is commenting like that. I quoted your own words back to you and you still can't see what you post. Got it. I am glad you like a small horn button that moves as you move the steering wheel since you only use your horn once a year at most. For the tenth time, not everyone agrees with you.
In pretty much all situations where someone is using a horn, using that time to use the accelerator pedal, the brake pedal and the steering wheel will provide better results.
That may be true but if you're in the left lane with a wall to your left and someone behind you, you might not have a lot of options. Yes, this happens. Again, there are better drivers in other places apparently but it's different out West. And in a parking deck you sometimes need to honk briefly to let people know to stop backing up into you. Some people simply don't look and braking or accelerating not an option there. Sure, in that situation you likely have more time but to say the current horn setup is an improvement of center of the steering wheel is a stretch I'd say.
The horn is a last resort, and most often will not work correctly.
The issue is people use it as the first resort.
legendsk
Member
The current horn setup may not be much of an improvement, but surely we can agree that it is more of an improvement than the other touted improvement of a screaming goat noise?
stopcrazypp
Well-Known Member
It's mentioned, but starting in 2019 they no longer promise any concrete features that can be interpreted as L4 on the order page. They only put "Autosteer on City Streets" which is a fully an L2 feature. That seems to be their legal CYA, although that leaves an awkward situation of what to do for cars bought prior to that (although they are slowly being aged out).
Last edited:
S100D
Member
I bought a S90D on December 21ar 2016. Elon said back in 2016 that Full Self Driving (FSD) with Radar, Ultrasonic Sensors, and Cameras had the hardware required to attain level 5 autonomy and could reach level 5 without Lidar. Here we are 7 years later with no Radar and Ultrasonic Sensors, and Tesla Vision sill cannot reliably attain level 2 autonomy.
Elon said that by the end of 2019, a Tesla with FSD would drive from California to New York City without human intervention, this included automated charging. There was a prototype Supercharger cable that could automatically unlock the charge port and charge the Tesla. This never happened.
Elon said in 2017 that if you owned a Tesla with FSD that you could subscribe to their RoboTaxi fleet. You could drive your car to work and while you were working, if your car received a request from RoboTaxi that it would unpark, drive to the location to pick up the passenger, take them to their requested location and you would make money. If additional requests were received, you Tesla would continue driving passengers to their requested locations. Your Tesla would know when you needed to leave work and would return to the parking space and so you could drive yourself home. If it needed to charge the battery to make it home, it would do so before returning to the work parking spot. This never happened.
On August 26th of this year, I traded in my 2016 Tesla S90D for a 2023 Tesla Model S Long Range with HD 4.0. Unfortunately, the Ultrasonic Sensors, turn signal stalk and shift stalk were removed. Tesla Vision sucks.
In the past few weeks when driving with FSD turned on, once driving on a cloudy day (no rain) with a clean car/cameras, I got a message that said Poor Weather Detected. The message below said FSD maybe degraded. On a sunny day with a clean car/cameras, I got the same message.
When it is raining and the cameras get wet, I get the same message, Poor Weather Detected, FSD Maybe Degraded. In medium to heavy rain, the cameras cannot see the road and initially FSD will slow your car down by 10 mph. If it starts raining harder, FSD will disengage. With my 2016 S90D with Radar and Ultrasonic sensors, I did not have this issue unless the visibility was almost zero.
Tesla has been charged with False Advertising and I totally agree with it. If the latest HW 4.0 FSD hardware cannot drive in the rain without disengaging, Tesla will never achieve level 2 autonomy much less ever be able to reach level 5.
I love my Tesla Model S and believe it is the best car I have ever owned, but FSD will never achieve Level 5 autonomy that Elon said back in 2017.
It's important to differentiate between current software / hardware limitations and potential. Similar to how your 2016 couldn't get Autosteer on city streets until many years later, there's potential for end-to-end to better handle raining situations without the "degraded" message or disengaging. The initial v12 release probably will still have those messages, but hopefully it won't take years to improve the software for the deployed hardware.
It'll be interesting how regulators and the public assess automated vehicle safety in different scenarios such as if it'll be acceptable or even desirable for say maximum speed to be reduced even more than 10 mph when raining. Would getting to the destination safely but potentially much later still be better than not completing a trip in the rain?
A lot of this will depend on how end-to-end actually behaves in the rain where maybe it already is good enough or can do fine at lower speeds or could need more significant changes to move from driver assist.
Furthermore in most cases the may be degraded message is just that, a message. It doesn’t disable function in most cases. I suspect this is noted as at times it reduces speed due to conditions so the message is a notification that it May need to accommodate road/vision restrictions.
It's also important to differentiate between marketing hopes and reality. The latest rewrite is just that. Another rewrite. Disregard the "e2e" and "nothing but nets" tech bro lingo. Look at actual performance. There is nothing so far indicating that v12 will be vastly better than v11. Perhaps it's twice as good. That's still nowhere near autonomy.
The cameras are still susceptible to rain, dirt, sun et.c. There is no cleaning. Thera are no redundant modalities. The compute is still limited.
Regulators only care about MTBF in the ODD. Either it's safe enough in the rain at max 20 mph (if that's the ODD) or it isn't.
To start with FSD currently has no clue on when it is going to fail. There are no code for ODD-detection and boundaries afaik, as the Head of AP doesn't understand what an ODD is supposedly. There is the hand over protocols, there is certification to for example UNECE R157.
If they got the reliability up by 1000x tomorrow in some limited ODD, it would probably still take 12-18 month just to design the new UX (ODD detection, handover procedure etc) and adapt it to current (non-US) regulation like R157 detection of emergency vehicles, and then validation...
I'm not trying to be a jerk about this. There are major hurdles getting a system that's designed for L2 to autonomy. It's definitely not just about lowering speeds and cutting out operation in the rain.
Last edited:
willow_hiller
Well-Known Member
That's not true. FSD Beta has a dedicated network trained to predict when a driver will intervene:
They are not using it to allow the system to disengage itself when it predicts it's done something that warrants intervention, but they are using it to regression-test new firmware versions in the simulator.
That's not really what I meant, and it's not the gist of my post.
That might be the case, but it still also randomly veers into oncoming traffic with no warning what so ever. Do we know what the reliability of these safety related NN:s are? No.
stopcrazypp
Well-Known Member
I should note, you can achieve L4 and not handle rainy weather, it is not a required prerequisite. The order page previously suggested some features that can be interpreted as L4, but never mentioned specific levels (much less a specific definition like SAE's) or that it would handle all weather.
As for Elon, he have said a lot of things about potential features over the years (not just AP/FSD) that didn't come to fruition and I'm not sure how liable Tesla will be as long as it is a forward looking statement (not as it relates to how things operate currently) and when it is not tied to any specific order (it's just a general statement).
willow_hiller
Well-Known Member
You're moving the goal-posts. You first said "FSD currently has no clue on when it is going to fail," and now you're saying: "Okay fine, FSD might have some clue about when it's going to fail, but that clue may be unreliable."
What about reading the post and not single out a sentence and use it out of context? Did you have any other comments or thoughts about evolving an ADAS to autonomy?
I’m merely pointing out the hurdles that are not related to performance thresholds and ODD limitations.
willow_hiller
Well-Known Member
I'm not using your sentence out of context at all. The entire premise of your post was that Tesla is a long way away from higher levels of autonomy because they have not yet developed any of the requisite features.
I'm telling you that you're verifiably wrong, because one of the key features you point to is in development. And that's just what we have some insight into through internal sources.
Similar threads
- Article
