-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
First day on the AutoPilot trial.
- Thread starter timk225
- Start date
I lived in Ann Arbor for several years, so I can see where you're coming from when you say it drives too aggressively by MI standards. You should try out "Chill" mode to reduce the acceleration. In CA, however, other drivers are very aggressive, especially in the LA metro area, so I've had the opposite problem. I sometimes get tailgated and passed on a following distance of 2 and I've had braking near misses on 1 when the tailgaters I'm following are forced to slam their brakes, so neither is ideal. To be safe, I nearly always drive on a setting of 2, but what I really need is a setting between 1 and 2, more aggressive acceleration from a stop, and more of the following distance to be used up as the the car slows down.
My complaints are not disapproval of AP. I'd highly recommend purchasing it, as I use it everyday for my commute and it helps me arrive at my destination less fatigued and calmer/happier. Therefore, it's been $5k very well spent. I believe the key to making it more useful is that Tesla should provide acceleration and braking specific configuration for AP, instead of lumping it all in with the following distance.
⚡️ELECTROMAN⚡️
Village Idiot
Too bad you guys didn’t get to experience the first year of AP2’s development. You would be very impressed with where it’s at right now.
I have noticed chill doesn’t affect AP acceleration it seems. I tried using it to smooth out the hard accelerations in stop and go LA traffic, but no joy. I keep mine at 7 follow distance and it still will accelerate too fast and slow down too late for comfortable travel. AP1 is much better in that regard.
I think Tesla will eventually need to give us the option as to how aggressive we want the car to drive. You feel it needs to be more agressive and I feel it needs to be less agressive even though we both travel the So Cal freeways. I think it really comes down to driver preference on that.
Atari2600
Active Member
Have you tried perpendicular parking? I don't have my Tesla yet but I did test drive one and tried everything out successfully except changing lanes because it wouldn't work, I think because traffic was really horrible. The rep entertained all my questions about the system and tried to demonstrate them. All of it seemed stable but one must understand the operational limits of the system to judge it properly such as expecting it to dodge potholes etc.
While true, there is no way anyone would buy AP 2 on a free trial if it acted the way it did when I got the car (which was much improved from the first AP2 release!). That system regularly made passengers yell and would scare the crap out of me. I basically stopped using autosteer until the big 10.4 update.
Even today, it has issues with the phantom braking, lane merges, etc. I had a coworker ask me why he often sees Teslas on his commute suddenly swerve at freeway speeds, and I told him it was probably folks using AP and that he shouldn’t ever tailgate one as the system might brake check him. The freeway he travels has several spots where AP loses the lane lines due to some unusual painting with alternating black and white stripes for lane lines. Perfectly visible to humans, but AP can’t see it and it will catch you off guard if you haven’t used AP in that area before.
So I am kind of surprised they feel the system is stable enough for the free trial now.
Things are going to be very interesting once (if?) FSD is available. Assuming the manufacturer is responsible for any accidents or tickets, it will really be up to them on how the car drives. If they program the standard as speed limit + 5, it would make it pretty easy for a local PD to attempt to fill a budget gap pretty quick.
Tam
Well-Known Member
I heard that you are responsible as an owner with insurance but if it's Tesla FSD technical issues, it will then take responsibility.
Uncle Paul
Well-Known Member
Part of the trial is learning where the autopilot works well and where it does not.
Works great for freeway cruising and in rush hour traffic. Takes a lot of the druggery out of the commute.
Expecting it to be fully autonomous in city driving will result in dissatisfaction.
Some trial people will intentionally give it a "Trial by fire". They will expose it to the most difficult circumstances and try to find situations where it will fail to impress.
Note. It will not yet respond to traffic signals nor stop signs. It does best when you have a lead car that it can follow and learn from. It takes a while to calibrate all its sensors when first installed, so some patience is needed.
Think it is remarkable that Tesla would offer a free trial to those who did not pay for it when ordering.
Like most things, some will decide to pay to activate it permanently, while others will decide it is not worth the $.
If you wish to find benefit from the Autopilot...you will. If you wish to find flaws from the Autopilot...you will.
Works great for freeway cruising and in rush hour traffic. Takes a lot of the druggery out of the commute.
Expecting it to be fully autonomous in city driving will result in dissatisfaction.
Some trial people will intentionally give it a "Trial by fire". They will expose it to the most difficult circumstances and try to find situations where it will fail to impress.
Note. It will not yet respond to traffic signals nor stop signs. It does best when you have a lead car that it can follow and learn from. It takes a while to calibrate all its sensors when first installed, so some patience is needed.
Think it is remarkable that Tesla would offer a free trial to those who did not pay for it when ordering.
Like most things, some will decide to pay to activate it permanently, while others will decide it is not worth the $.
If you wish to find benefit from the Autopilot...you will. If you wish to find flaws from the Autopilot...you will.
I agree with all of the above except that I think it makes perfect sense for Tesla to give the free trial. They lose nothing by it and they avoid the headache of people buying it and then complaining they don't like it. EAP is so great that it will sell itself. And if someone gets to use it for 2 weeks (or a month?) who would never buy it anyway, it costs Tesla nothing but the insignificant bandwidth to push out the switch.
I do think that a month would be a better choice than 2 weeks (not sure which it actually is) because it takes a while to get comfortable with it. It took me a while. You want the trial period to be long enough that people get comfortable with it, or it will have the opposite effect.
To return to my earlier computer analogy, I once bought a computer that came with a return policy that said I had to keep it for a month before I could return it. Then if I still didn't like it I had something like two weeks to return it, no questions asked. They knew it would be frustrating at first but that after a month I'd probably want to keep it, which I did.
I have to imagine that when real FSD is here, the car will be programmed to follow the speed limits period. I can't imagine Tesla taking on the liability of programming FSD to speed, despite how everyone actually drives on a particular road.
The liability question is going to get interesting once auto lane change is implemented. Am I going to be responsible for quickly checking everything is clear when the car initiates a lane change on it's own? That would kind of defeat the purpose it seems like.
Last edited:
On my one long annual road trip, in the Prius I drove as fast as I felt I could get away with consistent with safety because driving was so tiring and stressful that I just wanted to get there. With EAP in the Tesla I just drove at the limit or a couple of kph over because it was so relaxing. In an FSD car, I'd be napping in the back seat and in no hurry at all.
But I suspect that FSD cars will have to maximize safety, which will sometimes mean going with the flow of traffic even if that's over the posted limit. Of course, once all cars are FSD they can all go the limit.
Auto lane change now requires you to make sure the lane is clear before changing lanes. This is consistent with Level 2 autonomy, which EAP is today. Once the car is capable of making its own decision to change lanes, without any driver input, the car will be responsible. Liability will be hashed out between auto makers and insurance companies. And since FSD cars will be safer, insurance companies will be happy to negotiate because they'll want more FSD cars on the road. Rates will be set accordingly. The owner will probably pay the premiums, as now, unless some auto makers decide to offer included insurance as a selling point.
Day 2 of trial. It was better today, I knew more what to expect.
I went out on the PA Turnpike, and it was changing lanes well. Other driving on high and low speed roads was better.
Does AP learn new roads if you drive them a few times? I live on a hill with a somewhat curvy road, and it varies from having both paint lines to no lines at all. This evening, the AP made it all the way down the hill on its own, and most of the way back up. I had to step on the electric pedal halfway up the hill, because it was saying "autosteer limited to 30 mph", and I was just getting to the part of the road with good paint lines.
More experimenting tomorrow.
I went out on the PA Turnpike, and it was changing lanes well. Other driving on high and low speed roads was better.
Does AP learn new roads if you drive them a few times? I live on a hill with a somewhat curvy road, and it varies from having both paint lines to no lines at all. This evening, the AP made it all the way down the hill on its own, and most of the way back up. I had to step on the electric pedal halfway up the hill, because it was saying "autosteer limited to 30 mph", and I was just getting to the part of the road with good paint lines.
More experimenting tomorrow.
Tam
Well-Known Member
Not yet but it’s expected to in the future.
For now you can expect consistent behavior within the same patch level.
However, tweaks are made with new software releases that change behavior.
Usually for the better. Not ready to fall asleep with EAP.
Last edited:
I don't believe AP learns particular roads. There are some roads where it thinks the speed limit is higher or lower than it really is, and it behaves accordingly. Sometimes you have to disengage it if you want to go faster than it is willing to go. And sometimes when you engage it it speeds way up beyond the limit and you have to quickly dial it back down. There's no way (yet) to engage it at your present speed if that's less than it thinks the limit is, and there's no way to change the set speed until after you engage it.
Glad you're having a better experience as you get used to it. I expect it to continue to improve with future firmware updates.
I hope one day they give all cars at least TACC. It seems it’s starting to become near standard for many other basic manufacturers these days like Honda isn’t it? Or at least Clarity has it. I mentioned this in another post so won’t try to repeat myself too much but I hope EAP gets really good and maybe offering AP1 or just TACC for $1-2K would be an option. I’d also consider a subscription as others have mentioned.
Also an Ann Arbor to SD person, go blue!
Also an Ann Arbor to SD person, go blue!
TACC has been on the past 3 crappy rental cars I've got while on business travel... I'm really sad that my Tesla doesn't have it.
On another note, the EAP trial started for me today. I found it quite impressive but different than a lot of the rest of you. I don't think it drives like an a$$hole at all, which probably says something about my normal driving :-x I was also way more nervous with it activated than I thought I would be. I'll have to give it a try for a longer period of time to trust it more. As of right now, I definitely won't be buying it. The biggest thing for me is lane positioning. The car in EAP will sit right in the middle of the lane no matter what else is around me (another car, semi truck, etc.). Since I ride a motorcycle this is a little unnerving to me; I tend to naturally hug one side of my own lane when another vehicle is close to me.
Finally, the "wiggle" when disengaging the EAP by turning the steering wheel was also weird, but I got used to it after forcing to do it a few times. Pulling up on the drive selector is a much smoother way to disable it.
On another note, the EAP trial started for me today. I found it quite impressive but different than a lot of the rest of you. I don't think it drives like an a$$hole at all, which probably says something about my normal driving :-x I was also way more nervous with it activated than I thought I would be. I'll have to give it a try for a longer period of time to trust it more. As of right now, I definitely won't be buying it. The biggest thing for me is lane positioning. The car in EAP will sit right in the middle of the lane no matter what else is around me (another car, semi truck, etc.). Since I ride a motorcycle this is a little unnerving to me; I tend to naturally hug one side of my own lane when another vehicle is close to me.
Finally, the "wiggle" when disengaging the EAP by turning the steering wheel was also weird, but I got used to it after forcing to do it a few times. Pulling up on the drive selector is a much smoother way to disable it.
Similar threads
