FIRMWARE UPDATE! AP2 Local road driving...and holy crap

[electracity]

I rather suspect that the Tesla AP group is is high pressure place to work right now.

You can't run a company where every day is an emergency.

 

[SailorDave]

I gave AP 2.0 autosteer another go today. I was on a feeder road, new lane markers, three lanes, straight, and stop and go traffic- it worked great! No problem with intersections or turn lanes. It was much more like a highway and stop and go traffic. Looks like it works well in ideal conditions but has a significant issues once you get outside those conditions.

 

[disagree]

Well, initially there were many on the forum that made excuses or claimed "you just don't know how Tesla works!". But for the rest of us, the situation was revealed in the actual presentations and videos released. Some of us sort wheat from chaff for a living, and the clues that hype was the only thing being revealed were pretty clear.

But, I don't think complaints will serve much purpose at this point. That Tesla isn't currently capable of duplicating AP1 performance, let alone "enhanced" AP is now self-evident. What would you like them to do? They are working to recover as quickly as possible. Did they grossly overstate their capability? Yes. Did they hype the situation to cover up the huge setback that Mobileye's departure created? It appears so. Did they knowingly mislead customers? Obviously.

But they are doing what they should to develop their own system as quickly as possible. You can always buy a CPO AP1 car until this situation is resolved. If another maker starts shipping Waymo-equipped cars, and Tesla is still as far from "full self driving" as they appear to be now, buy a Waymo car. Until such an option is actually in the market, my suggesting is relax and enjoy what you have.

Do you want my honest answer? I doubt it. Will I give it to you anyway? Probably. Are my kids ready to head out for school? Not even close.

If Tesla committed fraud, they should be held accountable and penalized. Having a noble goal does not excuse them. Having a charismatic-ish cult leader who also makes rockets does not excuse them.

Tesla needs to stop the bullshit and just be honest about where the software is, whether the hardware is a problem, and how they plan -- with an intelligible roadmap -- to achieve what they showed us on a video in October.

Customers should not have to "buy a different car because you were defrauded." What sense does that make? None.

 

[number12]

The problem with the whole it takes a while to learn and EM tweeting new hardware updates every 12 months is that, for some, the itch for the new version might be here before half the functionality is introducted.

My Sept 2016 X felt "old" in about two weeks time with announcement of AP2.0

I too have realized I can just wait it out till at least FSD is all the way there (in case more hardware changes are needed).

ALSO - almost every car will have self driving in a few years (em words not mine). It's almost cheating to put out these dangerous versions to get real driving simulations when everyone else is waiting till 2019 and will come out with better tech from the jump.

 

[croman]

All I can say is that I'm glad they are at least doing fairly frequent updates. They would be better served delivering functionality that is useable. For example, 35mph local autosteer isn't really something anyone cared about as much as highway autosteer are usable speeds (65mph+). I guess we'll see what they do in another 2 weeks...

 

[josher32]

All I can say is that I'm glad they are at least doing fairly frequent updates. They would be better served delivering functionality that is useable. For example, 35mph local autosteer isn't really something anyone cared about as much as highway autosteer are usable speeds (65mph+). I guess we'll see what they do in another 2 weeks...

I couldn't agree more! Was hoping they would have released functional highway speed AP before giving us half baked versions of both.

 

[OBX John]

I had a somewhat similar reaction - AP2 did great on my first try, but the road looks like this:

After feeling great about the update, giggling like a kid all the way down the beach road, I decided to try a challenge, Colington Road.

It was pretty much unusable. Any curve more than ten degrees, and it would try to run off the road. I'm sure it will get better, but kind of surprised at how poorly it handles curves.

 

[182RG]

t's gathering data all the time

It would be interesting if there was a way to measure the amount of outbound data a HW2 Tesla is transmitting back to the "mother ship", to chart if there are differences between engaged, and the alleged "shadow" mode. That would be very interesting. Is it aggregated and transmitted real time, batched in bursts, overnight, or is it happening at all?

You forgot there are a few that believe that the NN portion of this is snake oil at this point.

 

[Maximapolak]

I had a somewhat similar reaction - AP2 did great on my first try, but the road looks like this:

After feeling great about the update, giggling like a kid all the way down the beach road, I decided to try a challenge, Colington Road.

It was pretty much unusable. Any curve more than ten degrees, and it would try to run off the road. I'm sure it will get better, but kind of surprised at how poorly it handles curves.

In my short experience with the newest update, it corrects itself on "tighter" turns a little late but it still does correct itself.

 

[HX_Guy]

I had a somewhat similar reaction - AP2 did great on my first try, but the road looks like this:


After feeling great about the update, giggling like a kid all the way down the beach road, I decided to try a challenge, Colington Road.

It was pretty much unusable. Any curve more than ten degrees, and it would try to run off the road. I'm sure it will get better, but kind of surprised at how poorly it handles curves.

I can't wait for my update, as in the Phoenix suburbs, we basically do not have roads such as those. Hell, there are barely any two-lane roads with just a center yellow lane, just about every street (aside from) has 2 lanes in each direction and a center left-turn lane. Everything is built as a grid with very few curves, and everything is very new and well marked. I have a feel AP does very well out here.

 

[u00mem9]

I agree. I'm not trying to defend their actions. I'm making suggestions about how customers or potential customers should react as they come to realize the reality behind the hype. The truth is they don't know how to do it yet. If winning a class-action suit and getting your money back will make people happy, by all means, move forward with that. But I suspect what customers really want is a safe, self-driving vehicle. And getting a refund won't provide that.

And while I agree the ends do not generally justify improper means, if you put yourself in Tesla's position mid 2016...what would you do? I'm all for ideals, but would a company that is losing money on every car sold, bailing out cousin's solar scam op for $2B, and now looking at the very real nightmare of trying to sell vehicles that no longer have Autopilot (one of it's biggest PR boons for the last 12 months and a killer app for many customers) for its vehicles for the next 12+ months, really survive that crisis? Sales would literally drop to zero. Their ability to cover liabilities like our warranties would soon disappear. The. end.

It may still happen. I certainly can't recommend a friend buy any of the current production unless they are strictly interested in EV for moral reasons. If the truth about this situation becomes widely reported, it could be the end of Tesla. I'm not trying to be dramatic, but you have to understand this IS a company trying to change the world, and they are beyond 'heavily leveraged'. I'm inclined to give them any rope I can spare. We can always go back to the cars we had for the last 100 years if they fail.

 

[Vitold]

I rather suspect that the Tesla AP group is is high pressure place to work right now. And I'm quite sure that Tesla would have preferred to continue down the road further with Mobileye, had it not been that ME was trying to blackmail.... sorry... ENCOURAGE, them to sign up with ME for eternity.
[...]

Cut Tesla some slack - they were not deceiving anyone. Director of AP left abruptly, on bad terms, and they recently hired Apple guy to replace him. Such transitions take time and this staff change probably delayed AP2.0 by a few months.

AP visionary has left and top notch software developer is in. This is great development for Autopilot and once Chris Lattner has settled in his new role we can expect AP 2.0 progress to pickup again.

 

[croman]

While we are willing to cut Tesla slack, their lack of communication is what is creating issues, not the deployment of features.

 

[kavyboy]

Any curve more than ten degrees, and it would try to run off the road.

By comparison on AP1, I just tried this one again
View media item 116571Heading west, that's a sharp left turn of more than 90 degrees, followed by a 90 sharp right turn. It's 55 MPH the whole way (+5 MPH) and an undivided, two-lane road with steep canted turns. It's pretty much what you're not supposed to use AP for, but I use this as a test case with my hands firmly on the wheel, foot on the brake.
It actually slowed down to the mid 40 MPH range for the first turn and took it perfectly, though a bit faster than I would have. Lane keeping was perfect. On the second turn, it didn't slow down, and there was no way I felt comfortable approaching a 90 degree turn at 60MPH, so I chickened out early and took over. I think there's a good chance it could have navigated that, one too, though.

 

[OBX John]

In my short experience with the newest update, it corrects itself on "tighter" turns a little late but it still does correct itself.

I was halfway in the other lane before I took control, and didn't want to try the sandy shoulder, but I'll give it another chance today. Just surprised that it would cross a double yellow line instead of following the curve. Pretty good fog line on the other side too.

 

[croman]

Even AP2 will slow down on curves. It does around me when I use tacc on a 4 lane road that winds and turns through valleys. It takes it a bit faster than most but about how quickly I do it.

 

[drshan]

I had a somewhat similar reaction - AP2 did great on my first try, but the road looks like this:

After feeling great about the update, giggling like a kid all the way down the beach road, I decided to try a challenge, Colington Road.

It was pretty much unusable. Any curve more than ten degrees, and it would try to run off the road. I'm sure it will get better, but kind of surprised at how poorly it handles curves.

This is starting to get comical. I hope the joke ends soon.

Just thinking back to my testdrive in late November w AP1. Those were the days...

 

[Cinematechs]

I'm seeing a lot of kids who are throwing tantrums because they are too impatient. Tesla is a tech company that, alike all other tech companies, have shown us a glimpse into it's future software potential if we buy it's new hardware now. The only difference is that they are extremely careful as to what they are releasing since your LIVES depend on it. Just relax and before you know it, they will gradually deliver. If you're unhappy with having to wait longer than your "ETA'd date, then sell your cars and try auto pilot on another vehicle.

 

[NeverEnough]

I can't relate at all to most of this thread (and so many others). I am perfectly satisfied with my purchase, I loved my car from day 1 (even with no AP features enabled), I love TACC, I am excited to get autosteer for my regular driving soon, and I am excited for it's abilities to improve iteratively. Since my opinions seem to be unique, I really don't have any use for TMC except to occasionally get a feel for updates I have not yet received. I just wanted to throw that out there before I disappear. Also, don't wreck Tesla's trajectory with all your complaining! And as my final thought on the subject: don't be so melodramatic that you spread your baseless speculation that Tesla is not capable and the hardware is not capable and self-driving is not possible without intelligent infrastructure.

 

[oktane]

That Tesla isn't currently capable of duplicating AP1 performance, let alone "enhanced" AP is now self-evident. What would you like them to do? They are working to recover as quickly as possible. Did they grossly overstate their capability? Yes. Did they hype the situation to cover up the huge setback that Mobileye's departure created? It appears so. Did they knowingly mislead customers? Obviously.

Firstly, I'm hoping that AP2 is real and simply taking a few months longer than expected to roll out.

If in fact AP2 at least equivalent to AP1 will not be delivered by end of March 2017, I will have officially reached my limit. I think giving them a 3 month grace period on a feature that should have been PERFECTED in October 2016 is quite generous.

What would I like them to do? (in no particular order):
1. Grow up and stop hyping future products.
2. Grow a pair and admit if they made a mistake in overpromising AP2 capabilities.
3. Communicate honestly to people who are victims of this AP2 lie.
4. Man up, and perform full buy-backs of effected cars or cover all payments until AP2 rolls out as advertised.

The morally right thing to do here is self-evident, or at least it should be. The legally right thing to do is also evident, and will bite Tesla if they are indeed found guilty of fraud.

I have one question for Tesla - where is the car that you showed on the FSD video? Can a journalist take it for a little drive and see how it works? If the FSD car doesn't work as show in the video, we know Tesla is guilty of fraud. That would prove a PURPOSEFUL attempt at inducing sales based on a fictitious product (a fake self-driving car).