Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Still considering advanced autopilot?

This site may earn commission on affiliate links.
With the fatal accidents mounting up because of blatant failure of the basic automatic emergency braking to stop the car, much less the autonomous driving recognizing the obstacle and changing lanes, would you still consider this feature? Automatic emergency braking had been out for a little while, with varying levels of success, but even this simple feature seems to fail miserably on teslas, and if you can't trust the autonomous driving, what's the point? I'd rather skip all the radar guided stuff if it doesn't work reliably and spend less on the car, how about you?
 
Not trolling, asking a serious question. This makes 3 accidents that I've heard of, and even without autonomous driving these cars should have stopped themselves to avoid the accident, and with it on they should have stopped or changed lanes. Not only that, but there are now 2 fatal crashes in supposedly the safest car in the world. Autonomous driving is the only option, autopilot comes whether you like it or not, likely adding at least $1k to the price of the car, that could be my all glass roof or premium sound. I understand buying the car, getting the option, and using the option are all my choice, I'm not stupid, but I'm not sure the draw to a technology you can't rely on, auto stop/lane departure assist or sutonomous driving if it doesn't work 100% of the time. I'm legitimately wondering if the object wasn't recognized or ignored, and why the car didn't at least slow itself enough to at least save the lives of the drivers. To me, these are important questions when considering how to option my m3.
 
Not trolling, asking a serious question. This makes 3 accidents that I've heard of, and even without autonomous driving these cars should have stopped themselves to avoid the accident, and with it on they should have stopped or changed lanes. Not only that, but there are now 2 fatal crashes in supposedly the safest car in the world. Autonomous driving is the only option, autopilot comes whether you like it or not, likely adding at least $1k to the price of the car, that could be my all glass roof or premium sound. I understand buying the car, getting the option, and using the option are all my choice, I'm not stupid, but I'm not sure the draw to a technology you can't rely on, auto stop/lane departure assist or sutonomous driving if it doesn't work 100% of the time. I'm legitimately wondering if the object wasn't recognized or ignored, and why the car didn't at least slow itself enough to at least save the lives of the drivers. To me, these are important questions when considering how to option my m3.

This is like getting worried about the number of ports that will be on the 2018 Macbook Pro.

Maybe you are jumping ahead of yourself a little bit, unless Tesla has been in touch with you regarding optioning your M3..?
 
Not trolling, asking a serious question. This makes 3 accidents that I've heard of, and even without autonomous driving these cars should have stopped themselves to avoid the accident, and with it on they should have stopped or changed lanes. Not only that, but there are now 2 fatal crashes in supposedly the safest car in the world. Autonomous driving is the only option, autopilot comes whether you like it or not, likely adding at least $1k to the price of the car, that could be my all glass roof or premium sound. I understand buying the car, getting the option, and using the option are all my choice, I'm not stupid, but I'm not sure the draw to a technology you can't rely on, auto stop/lane departure assist or sutonomous driving if it doesn't work 100% of the time. I'm legitimately wondering if the object wasn't recognized or ignored, and why the car didn't at least slow itself enough to at least save the lives of the drivers. To me, these are important questions when considering how to option my m3.

Have you considered that not many will write article when AEB actually worked? 3 accidents is very little and those that I've read about were all corner cases with obstacle partially on the road or hard for camera to see.

Also keep in mind that with upcoming version V8.x coming out AEB is still improved. If you buy Tesla you get OTA software updates for free whereas on other cars the only way to upgrade is to buy a new car.
 
3 fatal accidents? We know one for sure. The one in China maybe (IMO most likely). The one in Europe, most likely not.

These are all use cases that I know that TACC will not function properly. So if AP fails on a situation where it is not supposed to, then I will take note.

If you make up your own facts to slant the message, there are a few of you buddies here in this forum who don't own a Tesla or have driven with AP will jump on and keep this thread interesting by piling on. But the rest, who have been driving for tens of thousands of miles will ignore.
 
I'm legitimately wondering if the object wasn't recognized or ignored, and why the car didn't at least slow itself enough to at least save the lives of the drivers. To me, these are important questions when considering how to option my m3.

First, Autopilot is really traffic aware cruise control plus automatic lane keeping. There has not been a fatality related to either of those systems failing so far - cars have not departed their lane, varied wildly in speed or driven into another vehicle that was travelling in the same direction. The fatalities so far have, at best, involved Automatic Emergency Braking, which *every* manufacturer warns cannot detect stationary objects - it is really designed to prevent rear-enders.

To address the stationary object problem as well as the car-you-are-following-suddenly-changes-lanes problem, Tesla is attempting to improve how the radar system works.

I understand it is too tempting for media to do anything but report "Autopilot crashes", but every article I've seen on the subject seems to have been written by people who have an agenda or who have never used Autopilot or both.
 
I understand nobody is going to report the systems working, but these are rather severe in the outcome, thus the cause for my questions/ concern. I also realize these are stationary objects (I've actually done quite a bit of research on these systems in road cars) , but there are systems that work for this case, off center, idk, but stationary, yes. The ota updates are nice, and one reason I'm considering tesla. Even my simple, cheap ice car had several updates, some of which would have cost money since the car was out of warranty, but all of which would have required an out of town trip. It is a but frustrating they advertise it as autonomous driving then they say you still need to babysit it 100%. Call it what it is, advanced cruise control. For example, we have a lot of unmarked, dirt, and snow covered roads up here, are they developing it to see the road and extrapolate lane positions? I would think this would be doable.
 
They have never advertised it as autonomous, it's called autopilot, notice how those words are different. They can call it whatever they want, if you don't intend to use it the way it's intended to be used I would suggest you not get that option. It doesn't matter if you have done research in these systems, Tesla is doing their own thing with autopilot, Mobileye is gone and acting like a bitter ex girlfriend. Do you really think it's necessary to post in multiple forums about autopilot because of 3 accidents in what could be the safest vehicle ever made.
 
flyinghook: What you need to do is, listen to the folks in this forum who have used AP extensively for tens of thousands of miles and read through their hundreds of posts here buried in so many topics. And then you will notice that over 95% of the folks who use it LOVE it.

The folks who complain here and in other new outlets are the ones who have not used AP to any meaningful extent or even sat on a Tesla.

You don't have to baby sit. You have to pay attention. The experience is very relaxing, in fact very liberating. It is hard to explain, unless you experience it. One trip on a busy day with stop and go traffic on a highway is enough to convince a skeptic.
 
I understand nobody is going to report the systems working, but these are rather severe in the outcome, thus the cause for my questions/ concern. I also realize these are stationary objects (I've actually done quite a bit of research on these systems in road cars) , but there are systems that work for this case, off center, idk, but stationary, yes. The ota updates are nice, and one reason I'm considering tesla. Even my simple, cheap ice car had several updates, some of which would have cost money since the car was out of warranty, but all of which would have required an out of town trip. It is a but frustrating they advertise it as autonomous driving then they say you still need to babysit it 100%. Call it what it is, advanced cruise control. For example, we have a lot of unmarked, dirt, and snow covered roads up here, are they developing it to see the road and extrapolate lane positions? I would think this would be doable.

You don't think it is doable - Tesla is already releasing firmware to do it.

The whitelist we'll all be contributing to next week is being promoted as the answer to the accidents you're talking about - and it will be that - but it might be so much more.

Tesla is going to be building high precision radar maps of every route a car encounters with firmware 8, to identify all the things the radar shouldn't react to. One of the interesting twists to that, though: now the system will know where that object should be, if the car is in the lane. If they record enough objects, the car should be able to steer the lane based on radar inputs alone.

At the same time, they have high precision GPS running, too, and have been said to be working on maps for it. That would potentially give them triple redundancy for the routes - does the camera match the radar fix and GPS? If not, which one differs?

The decision to radar map tiles is actually an investment for the future, in addition to the short term benefits with stopped cars.
 
  • Informative
Reactions: Vitold
The decision to radar map tiles is actually an investment for the future, in addition to the short term benefits with stopped cars.
I don't think they're actually storing radar maps of regular locations, just radar representations, along with GPS coordinates, of objects that the car would typically engage automatically emergency braking on so that they can be whitelisted.
 
I don't think they're actually storing radar maps of regular locations, just radar representations, along with GPS coordinates, of objects that the car would typically engage automatically emergency braking on so that they can be whitelisted.

Whitelisting objects is certainly the first goal, yes. I'd be very surprised if they aren't thinking about the rest of it, though - and not at all surprised if the current firmware is starting to collect the rest of the data.
 
I understand nobody is going to report the systems working, but these are rather severe in the outcome, thus the cause for my questions/ concern.
Would they have been any less severe without autopilot? I've yet to hear of an instance where autopilot actually caused an accident, yes, it may not have worked as it should have in some cases but it's not like the car said "ohh look a billboard, maybe it'll be my friend...".