-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
TACC failed to brake at stop, nearly accident.
- Thread starter Lintrix
- Start date
-
- Tags
- Driving Dynamics Model S
green1
Active Member
This makes a lot of sense, TACC is radar based, Radar can't tell the difference between a car and any other stationary object, it really only knows what a car is because it moves like a car. We don't want TACC "locking on" to street signs, houses, guard rails, etc.
Emergency braking on the other hand needs to lock on to ANYTHING in the path of the car, and we don't care what it is, so it recognized this appropriately.
Sure it would be nice if TACC could recognize a car without it moving, but I suspect we'll have to wait a while for that, unfortunately that means that emergency braking is all that's left, and we know that isn't our preferred method as it's a pretty harsh stop.
Anyone know if any other manufacturer's system would have performed differently? I suspect not, but it would be interesting to know if any of the competitor's systems can do better.
Also, people REALLY need to stop thinking about this as "self driving" Tesla has NEVER claimed that this was anything other than an advanced cruise control, they have always been quite clear that this is cruise control only and that a real driver should at all times be paying attention and ready to take over immediately and without any delay. In fact they advertise it as for freeway use and limited access roads, the fact that it allows you to use it on streets with traffic lights is a bonus.
We seem to be in a situation where we are given an improvement over older systems, but instead of being happy, people are upset that it isn't an even bigger improvement. The bigger improvement will come, but let's wait until they claim it can self drive before we get upset that it can't.
Emergency braking on the other hand needs to lock on to ANYTHING in the path of the car, and we don't care what it is, so it recognized this appropriately.
Sure it would be nice if TACC could recognize a car without it moving, but I suspect we'll have to wait a while for that, unfortunately that means that emergency braking is all that's left, and we know that isn't our preferred method as it's a pretty harsh stop.
Anyone know if any other manufacturer's system would have performed differently? I suspect not, but it would be interesting to know if any of the competitor's systems can do better.
Also, people REALLY need to stop thinking about this as "self driving" Tesla has NEVER claimed that this was anything other than an advanced cruise control, they have always been quite clear that this is cruise control only and that a real driver should at all times be paying attention and ready to take over immediately and without any delay. In fact they advertise it as for freeway use and limited access roads, the fact that it allows you to use it on streets with traffic lights is a bonus.
We seem to be in a situation where we are given an improvement over older systems, but instead of being happy, people are upset that it isn't an even bigger improvement. The bigger improvement will come, but let's wait until they claim it can self drive before we get upset that it can't.
rneugebauer
Member
Distronic wouldn't brake in thinks situation, I think TACC is designed the same
I have no actual experience with TACC, but lots of years with Mercedes' Distronic.
Distronic will ill not react to vehicles that it hasn't seen moving, and Mercedes communicates that very clearly. In other words, it will stop behind a car it sees stopping in front of you, but will not brake for a car that has been stationary all the time since Distronic first saw it. (Emergency braking to reduce (but possibly not prevent) impact is another story, but you don't want to try that!)
Distronic is not designed to react to unmoving objects, there are too many of them all around as we drive. You don't want your car to brake for a car that is stopped in front of you if you intend to turn way before you hit it (the example in the video, but without the car that turned right, and you planning to turn instead of stopping).
Distronic/TACC can be a pleasure to use, but one has to be aware what it is designed to do. It is not able to stop at a red light (with no car in from of you), and it is designed to not react to unmoving objects. When the car in front changes lanes and you don't, you become responsible for making sure no unmoving objects are in your way. Those can be stationary cars, buildings, whatever.
Autopilot is an assistant, but you remain the pilot in control!
i guess Tesla has to communicate more clearly what Autopilot is designed to do, and what the caveats are.
I have no actual experience with TACC, but lots of years with Mercedes' Distronic.
Distronic will ill not react to vehicles that it hasn't seen moving, and Mercedes communicates that very clearly. In other words, it will stop behind a car it sees stopping in front of you, but will not brake for a car that has been stationary all the time since Distronic first saw it. (Emergency braking to reduce (but possibly not prevent) impact is another story, but you don't want to try that!)
Distronic is not designed to react to unmoving objects, there are too many of them all around as we drive. You don't want your car to brake for a car that is stopped in front of you if you intend to turn way before you hit it (the example in the video, but without the car that turned right, and you planning to turn instead of stopping).
Distronic/TACC can be a pleasure to use, but one has to be aware what it is designed to do. It is not able to stop at a red light (with no car in from of you), and it is designed to not react to unmoving objects. When the car in front changes lanes and you don't, you become responsible for making sure no unmoving objects are in your way. Those can be stationary cars, buildings, whatever.
Autopilot is an assistant, but you remain the pilot in control!
i guess Tesla has to communicate more clearly what Autopilot is designed to do, and what the caveats are.
Tesla has to communicate more clearly what TACC can and cannot do.
Tesla Autopilot HAS NOT been released yet.
TACC is not Autopilot. They are two different things.
Tesla Autopilot HAS NOT been released yet.
TACC is not Autopilot. They are two different things.
You guys brought up some important points that I definitely did not consider while using TACC. I was not aware that the sensors do not detect stationary objects as I could have sworn from previous experiences that the TACC would lock onto a stopped vehicle in front of me and stop on time - albeit it was a vehicle that was directly in front of me and not two vehicles ahead.
It appears that in the video the vehicle stopped just short of collision, but it was so close that it may have missed collision by centimeters. I don't know, from what I recall, even after avoided to the right, after the vehicle stopped, the distance between me and the vehicle ahead really felt like had I not swerved, we would have hit. It was close. The emergency braking did kick in though, that's for sure!
It appears that in the video the vehicle stopped just short of collision, but it was so close that it may have missed collision by centimeters. I don't know, from what I recall, even after avoided to the right, after the vehicle stopped, the distance between me and the vehicle ahead really felt like had I not swerved, we would have hit. It was close. The emergency braking did kick in though, that's for sure!
Todd Burch
14-Year Member
Tesla makes it clear that TACC in its current state is for highway use. Just be aware that using it on surface streets means that you are assuming risk for this type of thing. Always be ready to step in.
Kbra
Member
+1
It's much more useful on the highway when you are already tracking a car and pacing with it. If you have a larger gap due to your max speed it can be slow to react to a sudden drop in speed. It seemed to improve a bit with 6.2 but has room to improve. When it can read further than the car in front and predict an upcoming slow/stop it will be better for street use.
Oh bull. What is the system right now? TACC. Traffic. Aware. CRUISE. CONTROL. Do you regularly use cruise control on surface streets?
jgs
Active Member
I do on certain streets, but I'm aware of the issues and maintain control of the car.
green1
Active Member
That seems a bit harsh when as far as we know it is performing exactly as advertised, and every bit as well as similar competing systems...
- - - Updated - - -
So why would this be any different? Tesla doesn't claim that you don't need to maintain control of the car, so why do people on here assume they shouldn't have to?
When we got our A6 the dealer tried to show the adaptive cruise control off to us as well, and it failed miserably in similar fashion (approaching a light where there was a stationary car). So, while I think Tesla should try harder to differentiate TACC from all the "AutoPilot" stuff they keep saying (which may give the wrong impression), I think the system performed as a system of this kind should have -- just maybe not how folks would expect it to.
Though, the situation with the Audi means that Tesla may not be alone in needing to do this.
Though, the situation with the Audi means that Tesla may not be alone in needing to do this.
Cyclone
Cyclonic Member ((.oO))
I would imagine that as the autopilot system improves, it's deployment will also bring enhancements to TACC, emergency braking, etc. I would imagine that TACC would eventually use the camera to scan ahead for brake lights for this very reason. If the first car moves out of the picture and the next car is stationary, the camera may be able to pick up the brake lights and "assume" the object ahead is a stopped car and "lock on" to then brake.
well, to be clear, I would say it's not really the TACC system that failed, but rather the "automatic emergency braking" system that failed.
whatever you want to call it, it's brand new in 6.2, and it failed to stop. that's pretty damn clear to me that it's a massive failure of the reliability of the system, especially when they go around touting the feature in the release notes.
actually, before I even say any more about this feature failing, OP - are you on 6.2 or are you on 6.1?
green1
Active Member
The video shows emergency braking working perfectly. it automatically engaged to reduce the impact of an unavoidable frontal collision.
In fact, it did BETTER than advertised, the advertising says it will only reduce the impact of an unavoidable collision, and in this case it appears that it fully stopped short of the collision (by a hair)
Please tell me again how that's a "failure"?
I thought the OP said it did not brake at all and he had to manually brake at the end?
green1
Active Member
Nope. he did not brake.
So emergency braking worked better than advertised, and TACC worked exactly as advertised.
The problem here isn't in the systems, or in the advertising, it's in people having expectations that are ahead of what's currently being offered.
If you're upset that the vehicle doesn't yet have self-driving, wait until they say it does before complaining the features don't work! this whole thing (I don't mean to single out the OP, however it has come up many times now) reminds me of the urban legend of a guy who puts a camper on cruise control and then goes in to the back to make a sandwich. Everyone laughs at that guy for being so stupid as to expect the cruise control to keep him safe, and yet people here blame the car instead of the driver.
I'm all for pressuring Tesla to hurry up with auto-pilot, and I'm all for pointing out legitimate flaws in a system. But we shouldn't be expecting things that haven't yet been released to already be working.
Similar threads
