-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
+1 what andrewket said. I would not have upgraded from my 2013 P85 for just the dual motor and 3.2 seconds. Tesla put on a really slick show that night and probably got a large portion of existing P85 and P85+ owners to upgrade. I do think the lane keeping is coming soon. There's a rhythm to the software updates and we're about due for 6.3. They are taking a very slow methodical approach to release the features, one or two at a time then months of data collection in between. The hardware is standard MobilEye used in many other cars with several years of experience so we know it will work. It's just the stakes are so high for Tesla. Any misstep could be devastating for the company and the EV movement.
Short of Tesla saying they've given it to beta testers, none. The beta test community has been scrupulously faithful to the NDA, for good reason.
What is stated on the web site and order page supersedes anything that Elon said or did during the D and Autopilot launch. There would be no way to misinterpret what was said on Tesla's web site - I read it clearly - unless one ordered with their eyes closed.
- - - Updated - - -
Interesting choice of words for someone who owns three Model S cars, no?
- - - Updated - - -
Demo≠available
- - - Updated - - -
Where is the difference? Several means exactly that - more than one and less than some arbitrary, higher number. Your definition of "several" is different than Elon's, clearly, but that's not something you can hang your hat on. Tesla also does not need to prove anything to you. Sorry. They will release it when it's ready, and they've said so all along.
- - - Updated - - -
Satisfaction guarantee applies to leases only, and the owner would still have to incur significant costs to return the vehicle to Tesla. Read the fine print.
So would people have been happier if Tesla had purposely held back the hardware by 10 months so they could announce this summer that Autopilot was coming out in three months for example? They have obviously run into issues with the development and are late with the delivery from their vague promised delivery but I would rather have the hardware and get the software update than have to buy a new car like the first 70,000 or so Model S owners would.
Krugerrand
Meow
No, I think people want their cake and to eat it too, damn everything else. :wink:
I've bought 3, but sold/traded-in the 1st two cars. I've owned 5 different cars in the last 5 years, I could very well be more than halfway thru the life of my current P85D. I bought it for the autopilot that was being pushed in October and still don't have it.
We can talk semantics about what the word "several" means, but only Tesla knew when it could/might be released in the field. Obviously the software for the demo was home built by Tesla and was probably in development for well over a year (I have no idea). I'm sure the demo version wasn't the 1st working version they had, they knew the limitations & problems they needed to cleanup before mass release.
Knowing what we know now (isn't hindsight great), they should have said "released over the next year". HUGE difference between "several months" & "12 months". Obviously, that would never fly with marketing (&/or Elon). Now the question is: Were they that over confident or were they lying? (neither one is a good answer)
I have NO DOUBT that I could make them buy my car back and refund my money (then have me sign a NDA to keep it quiet), thus my conundrum. I love the car, it's just that I feel I was lied to. That's my feeling and it doesn't matter what anyone else thinks: I FEEL LIED TO. You (and many others) on these boards have had MANY problems with Tesla over the years, I have not agreed with most of them because they didn't directly affect me, now I'm on the other side, I'm very passionate about this issue whereas you (any many others) are not.
- - - Updated - - -
I think we all just want honesty.
Andyw2100
Well-Known Member
I, unlike tezzla, don't feel lied to, (though I of course understand why he feels the way he does.) I'm willing to give Tesla and Musk the benefit of the doubt, and as tezzla put it above, believe that they were just being over confident.
That being said, I do feel that Tesla should recognize the fact that this kind of thing, and the cavalier way in which they handle these things is causing them to lose a lot of good will from their most loyal supporters. If they simply said, "we understand you're disappointed, so we're going to do the following for you" and then came up with something that wouldn't cost them much, but that people would appreciate it, it would go a long way towards making things better.
They likely honestly did feel they could hit their 'several' months deadline but we will
never know. I was promised lighted visors as well and that never happened. If they had promised full autopilot with delivery and it wasn't on the car then you were lied to. This is more a grey area. Sure if they never ship it or it slips into 2016 you might have a case but at this point it might be tough to prove.
I do think Tesla should update customers with a more realistic estimate at this point. If Tesla had to give gifts for every missed deadline that might get expensive. Someone always seems to feel they were wronged (missed a new feature by a week and wants a new car or any of the other numerous controversies over the years). A simple statement of facts would go a long way I think.
SomeJoe7777
Marginally-Known Member
I think a lot of people should prepare for some disappointment, because my prediction is fairly pessimistic.
How many Model S owners drive with the Lane Departure Warning feature turned on? I do, but I think I'm in the minority. I think most owners turn it off because it issues a lot of false alarms. This morning, I had it shake the wheel for me because it got fooled by a shadow on the road.
I believe the future Lane Keeping feature of the AutoPilot suite is based on the same sensor input as the Lane Departure Warning. If this is the case, then it's well away from being ready. Right now, if you engaged it with the current software processing that's doing Lane Departure Warning, your Model S wouldn't last 15 minutes before it steered into a concrete barrier.
Here are my predictions/speculation:
The existing sensor suite is the bare minimum required to do lane keeping, and it's going to require extraordinarily sophisticated processing to do it, including camera & sonar input, and even input from the radar tracking things like the lane markers (reflectors/bumps) and side barriers. The on-board CPU for the tech features is already heavily taxed (watch the update rate/frame rate of your kW meter get halved when you engage navigation), and doing this kind of background processing for lane keeping might be more than the existing CPU can deliver.
Sorry to be so pessimistic, but I think there are major obstacles to overcome that haven't yet been solved.
How many Model S owners drive with the Lane Departure Warning feature turned on? I do, but I think I'm in the minority. I think most owners turn it off because it issues a lot of false alarms. This morning, I had it shake the wheel for me because it got fooled by a shadow on the road.
I believe the future Lane Keeping feature of the AutoPilot suite is based on the same sensor input as the Lane Departure Warning. If this is the case, then it's well away from being ready. Right now, if you engaged it with the current software processing that's doing Lane Departure Warning, your Model S wouldn't last 15 minutes before it steered into a concrete barrier.
Here are my predictions/speculation:
- I think we'll see self-parking features rolled out well before lane keeping.
- Self-parking might be ready by the end of the summer.
- Model X will be released later this year, and it won't have lane keeping either.
- Lane keeping implementation in existing Model S cars may require hardware retrofit -- either additional sensors, a faster CPU, or possibly both.
- Lane keeping in existing Model S cars won't be deployed until Q2 2016.
The existing sensor suite is the bare minimum required to do lane keeping, and it's going to require extraordinarily sophisticated processing to do it, including camera & sonar input, and even input from the radar tracking things like the lane markers (reflectors/bumps) and side barriers. The on-board CPU for the tech features is already heavily taxed (watch the update rate/frame rate of your kW meter get halved when you engage navigation), and doing this kind of background processing for lane keeping might be more than the existing CPU can deliver.
Sorry to be so pessimistic, but I think there are major obstacles to overcome that haven't yet been solved.
GlmnAlyAirCar
Active Member
Has anyone considered why the hardware may have been released before the software? My thought is that this allows Tesla engineering to gather real-world data on the conditions detected by the sensors on thousands of cars. They can run their development software through real-world scenarios by playing back the data. Perhaps they have found a few surprises along the way. Perhaps this accounts for the delay. But I would much rather have them address these issues now, with a bunch of users complaining about the late release, than later with a bunch of users complaining about a feature that doesn't work, and is potentially dangerous.
It's not the same CPU. MobileEye has its own FPGA+CPU. No engineer in their right mind would place safety critical systems on the same computer that runs multimedia and nav. You are too pessimistic. Lane departure warning is also likely twitchier since it's an alert to the driver, I would expect the car to react much more slowly, and to integrate other sensors into it's final trajectory decision (radar following car in front, sonar for cars at the sides, etc.). This entire thread is just arm-chair engineering and (somewhat more legitimate) griping about a late software feature.
Let's remember everything else software related that they've promised a date for an been late on: vampire drain, torque sleep, hover mode (ok maybe not). For better or for worse, this is how they operate right now. This is no different until we have strong evidence otherwise.
Last edited:
Yes. And happier still if they held the entire package (h/w and s/w) until they were ready to sell it. If they felt the need to *announce* it prematurely, they should not have charged customers for the functionality until delivered. Tesla will not succeed by trying to fool its customers. I think they have been deceptive, whether they "thought" it could be done "in several months" or not.
Soolim
Member
I always have my lane departure warning turn ON. Never had any false warning yet, and it could be the road that I travelled are better marked :smile:
- I hope that self parking gets release first. Hence I agree with you.
- I think MX will be released with lane keeping, but MX could have more sensors than current MS. Some videos of MX test driver trying in vain to go off lane was captured in other post.
- I hope Tesla will upgrade existing MS autopilot package with the requisite additional HW to make it work, free of charge as a goodwill to compensate for the extra long delay. Timing of such upgrade will be VIN sequence dependent to make early adopters happy.
I think NHSTA will have more concern on lane keeping because sometimes intentional lane departure might be necessary to avoid more serious accident. I think this is a challenge for software programmers, and lots of tests need to be demonstrated to NHSTA to get the passed. For example, turn off lane keeping if an frontal crush is imminent.
Ditto
9k miles and the only false warnings i got were in construction zones where there were lines everywhere.
SomeJoe7777
Marginally-Known Member
Fair enough, and if so, that's good. It means that at least CPU hardware won't be a roadblock.
I still think they have a ways to go. I'm glad that others haven't noticed as many false warnings from Lane Departure Warning as I have, maybe mine is somehow more sensitive. Or maybe Houston roads are crap. :wink:
But I can't drive more than a few minutes on the freeway without a concrete seam, an oil trail from a previous accident, or something else like that giving me the steering wheel vibration even when I'm smack in the middle of the lane.
I do think that Tesla will indeed deliver the lane keeping eventually, they're not going to abandon it for this sensor suite. They promised it, and legally, I think they have to deliver it one way or another. I just don't think it's going to be "soon".
They were talking about the hardware. Everything I recall from the event was talking about the autopilot hardware being on all cars in the present tense (which was 100% true), but they were also pretty clear about the software and features being several months away.
Google has been demoing self-driving cars for many years now. They had a commercial of sorts with a blind individual "driving" to do chores. I don't see the difference in this situation because blind men cannot currently drive Google's car.
Similar threads
