Why Do Teslas Keep Smashing Emergency Vehicles? (Warning: Graphic images in article)

[HaulingAss]

I really want FSD to win the autonomous driving race but its looking less and less likely that it will.

Name one company that's closer to solving autonomy in the general case (without small geo-fenced boundaries) than Tesla is.

I'm still waiting...

Sponsored

 

[JCERRN]

Actually that means it’s infrequent. ICE fires and non Tesla crashes are so common they stopped reporting them.

again, more the “happening at all” piece related to fsd/ap, not the frequency of incidence.

How do you know it's happening at a higher rate than the general fleet at large? And how do you know that each Tesla even had access to FSD, let alone had it engaged, every reported instance? Because people run into emergency vehicles parked on the side of the road all the time.

To be fair- we do not know if fsd or ap was activated or or inactive during or prior to the incident. Related to frequency- see above post.

 

[JCERRN]

Name one company that's closer to solving autonomy in the general case (without small geo-fenced boundaries) than Tesla is.

I'm still waiting...

I guess that depends on the metric you are using to define “closer” to “autonomous driving” if you go by legally authorized by the people in charge to do things in the US, even if in a limited capacity, Merc is winning.
If you go by “how much road can it drive on with the driver only kind of paying attention and hopefully nothing happens” then i suppose tesla is in the lead.
AGAIN, I'm rooting FOR tesla here. I was just seeking other’s input/experiences ex: “ive experienced fsd disengaging with flashing lights” “yeah its designed to do that, not a flaw” - the kind of discourse i was seeking.

neither fsd nor tesla are perfect, just like the humans that use them… but if tesla wants fsd to be fully autonomous, its gotta be better than a human in every potential road situation. Eventually there will be a driving system that can do that, i hope tesla is first, but others are fighting for that goal too, trying many other algorithms and systems

 

[Ogre]

again, more the “happening at all” piece related to fsd/ap, not the frequency of incidence.


To be fair- we do not know if fsd or ap was activated or or inactive during or prior to the incident. Related to frequency- see above post.

Media assumption is usually that it is active at the time of the incident.

Reality is usually that it isn’t activated or often is crippled (foot on gas pedal).

 

[jerhenderson]

You can have my steering wheel weight when you pry it from my cold dead fingers (after I crash). It doesn't mean I'm not paying attention, just like having a useless hand on the wheel doesn't mean you are paying attention.



I'm also enjoying the battle of wits with Elon. They figured out how to detect it, on the "normal" side which is the left. So I moved it to the right. Fixed.

when someone is trying defeat a safety device there isn't much of a battle of wits to be had.

 

[jerhenderson]

It looks like passenger may survive so they may be able to clarify if FSD or autopilot was engaged.















In this case, regardless of if it FSD was engaged, I think it was very likely the driver was doing something irresponsible (hard to miss those flashing lights especially if it was at night). That said, in general I don’t quite understand the argument that driver should have payed attention when FSD was engaged. Isn’t the whole point of FSD is being able to pay less attention? Even if you have signed some sort of release or agreement that you will pay attention and are responsible, wouldn’t Paying less attention be a natural byproduct of FSD? If any of you have been selected to test FSD beta, and can be honest here (since you are not using your real name), please share how much attention you paid throughout (the beginning, middle and end of) testing period in comparison to when you drove a vehicle with no driver assistance.















For me it has always been a binary thing. Either I can trust it or I can’t. Anything in between, I rather it is someone else testing it on a road I am not using.

" I don’t quite understand the argument that driver should have payed attention when FSD was engaged... " As FSD is only in Beta with specific wording about driver requirements for safety? what's not to understand? this is entirely the driver's fault we can conclude already.

 

[Ogre]

again, more the “happening at all” piece related to fsd/ap, not the frequency of incidence.

You’ve assumed that this is indeed a recurring thing which needs to be fixed but we really don’t know what happened.

There is not actually any evidence this is FSD… or Autopilot. So what exactly is “Happening”? A car driven by a human smashed into the back of another vehicle. Odds are this is almost identical to the incident last year where the media jumped all over a Tesla “On Autopilot” accelerating and flying off the road with no one in the drivers seat… when the truth was there was a driver with a BAC 3x the legal limit. Or the “Tesla” that smashed into a garage that turned out to be not a Tesla at all. Or the Tesla that caught on fire when nobody was driving it that turned out to be insurance fraud.

The list of crappy reporting on this is longer than the list of actual incidents. So what exactly needs to get fixed here?

 

[Diehard]

" I don’t quite understand the argument that driver should have payed attention when FSD was engaged... " As FSD is only in Beta with specific wording about driver requirements for safety? what's not to understand? this is entirely the driver's fault we can conclude already.

You have to forgive me. Some of us are just not that bright. It seems you have already gone through all the forensic evidence of this case. To me it seems like a Robot and a human were sharing the driving Task. Any of them could be at fault regardless of which one signed which agreement.

Edit: That is if FSD was on

 

[HaulingAss]

I guess that depends on the metric you are using to define “closer” to “autonomous driving” if you go by legally authorized by the people in charge to do things in the US, even if in a limited capacity, Merc is winning.

I don't use any one metric to determine who is closer to autonomy, I go by the totality of the evidence. You are the one who claimed it was looking increasingly unlikely that Tesla would get there first so I find your answer lacking in any substance that could back up your assertaition that Tesla is losing their lead.

But this kind of talk is nothing new, three years ago there were numerous people who were rating Tesla nearly dead last out of a field of 10-15 companies working on autonomy. Some of those companies have closed shop and given up. If you think Mercedes is now in the lead, I can't take you seriously.

 

[Ogre]

I don't use any one metric to determine who is closer to autonomy, I go by the totality of the evidence. You are the one who claimed it was looking increasingly unlikely that Tesla would get there first so I find your answer lacking in any substance that could back up your assertaition that Tesla is losing their lead.

But this kind of talk is nothing new, three years ago there were numerous people who were rating Tesla nearly dead last out of a field of 10-15 companies working on autonomy. Some of those companies have closed shop and given up. If you think Mercedes is now in the lead, I can't take you seriously.

Dude… it’s so obvious that Cruise is way ahead. They are at Level 17 already… they would be at Level 22 but apparently stopping in the middle of intersections is considered bad?

 

[JCERRN]

I don't use any one metric to determine who is closer to autonomy, I go by the totality of the evidence. You are the one who claimed it was looking increasingly unlikely that Tesla would get there first so I find your answer lacking in any substance that could back up your assertaition that Tesla is losing their lead.

But this kind of talk is nothing new, three years ago there were numerous people who were rating Tesla nearly dead last out of a field of 10-15 companies working on autonomy. Some of those companies have closed shop and given up. If you think Mercedes is now in the lead, I can't take you seriously.

Ok.

 

[JCERRN]

You’ve assumed that this is indeed a recurring thing which needs to be fixed but we really don’t know what happened.

There is not actually any evidence this is FSD… or Autopilot. So what exactly is “Happening”? A car driven by a human smashed into the back of another vehicle. Odds are this is almost identical to the incident last year where the media jumped all over a Tesla “On Autopilot” accelerating and flying off the road with no one in the drivers seat… when the truth was there was a driver with a BAC 3x the legal limit. Or the “Tesla” that smashed into a garage that turned out to be not a Tesla at all. Or the Tesla that caught on fire when nobody was driving it that turned out to be insurance fraud.

The list of crappy reporting on this is longer than the list of actual incidents. So what exactly needs to get fixed here?

I havent assumed anything. Ive already acknowledged that there was a chance ap/ fsd was not engaged.

 

[Dids]

So people hit emergency vehicles all the time. Then they are usually identified as incapacitated in some way.... But some people spend their whole life in an incapacitated state. In fact everybody is in a spectrum of incapacitated at any given time... What even is the metric for someone that is "capacitated"?
If something keeps happening the fault isn't just the user it is also the system's fault.

The purpose of emergency lights is to cause drivers to identify emergency vehicles and avoid them. They don't work....

 

[Diehard]

So people hit emergency vehicles all the time. Then they are usually identified as incapacitated in some way.... But some people spend their whole life in an incapacitated state. In fact everybody is in a spectrum of incapacitated at any given time... What even is the metric for someone that is "capacitated"?
If something keeps happening the fault isn't just the user it is also the system's fault.

The purpose of emergency lights is to cause drivers to identify emergency vehicles and avoid them. They don't work....

If the measure of system working is 100% safety, it does not but if it is +1% Safer it MAY be working ( I don’t know how to measure that). The idea with emergency lights is to replace a highly unexpected situation with a somewhat unexpected situation. Chances are regardless of what level of incapacitated a driver is at, they would recognize a flashing light as something unexpected faster than seeing the bottom of a semi in middle of the road one time and a car pile up the next. It is entirely possible in a specific situation emergency responders make the situation less safe but if a study shows that is in fact the case for all incidents and you were running the show, would you shut down their operation and stop taking calls? Or keep running it and wait until someone figures out how to improve the system?

Now that our cars are quipped with incapacitometers and watching us all the time, can we trust them not to blame us for their own screwup?

 

[HaulingAss]

I havent assumed anything. Ive already acknowledged that there was a chance ap/ fsd was not engaged.

Maybe I misunderstood your original question. What was it again?

Sponsored