P85D Lost power on road, "Pull over safely"

[JohnQ]

Unless there is something happening that we're not aware of, this is poor crisis management by Tesla. Assuming that range mode off solves this problem, send an email to owners. Or pick up the damn phone. Send a postcard. One accident as a result of this--someone unsuccessfully trying to get to the shoulder, or stuck on a blind corner where there is no shoulder--is all it takes to turn an embarrassment into a crisis.

I don't have a "D" variant but this type of SOP does not inspire confidence in the company.

 

[electrish]

... Assuming that range mode off solves this problem, ....

Since my problems, both the reduced speed incident and the complete power loss, happened in cold weather immediately following a somewhat brisk acceleration at highway speeds on range mode, I tried to recreate this situation this morning with range mode disabled and the car performed flawlessly. So to me it looks like range mode is the culprit.

On a related note, some posters mentioned that they were contacted by Tesla after their power loss; I was not.
Doesn't Tesla monitor cars remotely or get pinged when a critical failure like this occurs?

 

[JohnSnowNW]

I don't know what Tesla is able to do, or how they have things organized to do them, but perhaps the people complaining about the response do?

So, can they find out quickly who has .167 on their P85D? When were they positive that "Range Mode" and .167 was the cause?

I mean, it's easy to make a connection when you have multiple people in a thread comment on it, but that doesn't actually make it the issue. I should think that telling people to turn off Range Mode, and that not being the only cause of the issue, wouldn't sit well with people either.

I can understand the gravity of this issue for you all, it could potentially be a very serious. However, I'm not sure that the disappointment some of you feel is justified at this point.

 

[wk057]

I can vouch for range mode OFF working perfectly. I've driven over 200 miles since the incident, some pretty "spirited" driving as well without issue. I won't be enabling range mode until I see new firmware, however, which is unfortunate because efficiency still suffers compared to the P85 with range mode off it seems.

Unfortunately if that fix is not today/tonight I will be leaving the P85D at home instead of taking it on a long trip for the second time in 3 months of owning it, which is pretty disappointing to say the least. (First the lack or torque sleep and decreased range, now the lack of safe use of range mode entirely...)

Suffice it to say I will be, at the very least, probably writing a letter expressing my disappointment. While I understand that s*** happens, all of this seems like it would have been pretty avoidable one way or another from day 1. (Tell me about the range issue before delivery so I don't have to spend time investigating myself and not using the car when I would have liked to; revert .167->.140; etc)

While the car is awesome overall and generally tends to make up for most of the negativity on its own (ie, P85D inspired "Weeeeeee!" moments), the lack of practical use on now a second trip will definitely tip the scales back towards a disappointing experience.

 

[Andyw2100]

I agree with wk that when I completely lost power yesterday morning, it was not really a "life-threatening" situation. Even though I was going around 75 mph in the center lane of a three-lane highway, I was able to safely maneuver to the shoulder. The car was slowing down rapidly, but I still had a lot of momentum. Yes, traffic was light and it could have ended differently, but I just can't bring myself to freak out about it. It was probably more the surprise of losing all power, having red lights flash, and bells ring than any unstable driving situation.

I'm glad you feel your situation wasn't a life-threatening one. But it sounds like the one described below, (and in post number 17 in this thread) may have been:

Same thing happened to me today (P85D with .167 update from last night), and it resulted in a particularly dangerous situation. Was going thru some uphill twisties where you lose sight of cars in front/behind you every 3 to 5 car lengths. Came around a right-hand turn and heard "beep-beep-beep", which I had only heard before when the "early collision warning" alarm went off. A moment later I got "car needs service" and all power disappeared underfoot -- I was left stranded on the uphill side of a blind curve with no power available at all.

Very scary as I sat there trying to figure out what to do -- pulled over as far as I could in the 10 feet of coasting I had remaining before coming to a standstill, but was still sticking out 2/3rds in the road (since there was no shoulder), and put my hazard lights on. Just then I saw a large truck come up behind me emerging from the blind curve -- fortunately he wasn't going that fast -- so I honked to be sure he saw me. After he passed, I decided to put the car in Reverse, then into Drive again, and luckily power was restored and I was able to drive out of there.

Whole incident probably lasted 45 seconds, but I was scared sh**less.

One other thing -- at one other point in the drive I felt something funny -- a "lurch" when traveling downhill that caused the car to lose speed for a moment -- then all continued as normal.

I called Tesla Service and they said they pulled the logs and forwarded to a service advisor, who is to call me tomorrow.

For anyone else who this has happened to, was there a "3 beeps" audio alert associated with your power shutdown? Or was what I heard an indication that collision warning system thought that hillside next to me was another car? (Just trying to figure out if "early warning detection" is correlated with "power shutdown" in .167)

Needless to say, this is an awfully bad bug to instantly lose all power when traveling at speed, and could have resulted in serious injury. This incident has shaken my faith in Tesla QA and the pace with which they are updating software for the D.

-- David

 

[qwk]

I'm guessing that Tesla's logic goes like this: if we admit to,the problem then we are potentially open to extra legal liability and worse to an immediate fire storm in the press. So let's get the fix out in a couple days and hope no one is hurt in the meantime. After all, there are only a few hundred cars affected and the probability of a serious accident is pretty low. If we're lucky we can get out of this with minimal damage.

IMHO, that makes a bit of sense but it's an awful risk to be taking. The downside of losing that bet is just too great.

Tesla management has a history of making bad decisions which result in something that is very easy to fix, that gets ignored, or pushed back, resulting in the issue spiraling out of control and becoming a huge issue. Unfortunately this is only getting worse as time goes on.

It is extremely simple to revert back to the last working update until engineering can fix the issue. Sometimes you have to take a step backwards in order to go forward. I guess common sense is not all that common, and they much prefer chaos instead.

 

[Footbag]

So if this is very clearly a software bug and it is very clearly tied to range mode why doesn't Tesla disable range mode effective immediately on all D models? And don't anyone say that they don't have that capability. They do.

I would not appreciate such an action. I needed every bit of range yesterday for a ~600km trip (between -20C and -10C... arrived at a supercharger at one point with just 10KM RR remaining, and that was after adding a bit of extra juice thanks to @3s-a-charm). I used range mode knowing that I may be affected by this issue, but that it would likely be rectifiable. As it turns out, no such error condition during my journey. I've been in range mode almost exclusively since .167 was installed, and have been lucky.

That said, I was able to make a concious decision on testing fate, thanks to this forum. I agree Tesla should communicate with owners in this instance.

 

[electrish]

I'm glad you feel your situation wasn't a life-threatening one. But it sounds like the one described below, (and in post number 17 in this thread) may have been:

True.
All I am saying is that machines do fail, sometimes in the least opportune situation, which can lead to injury. Even with extensive testing, this situation may not have been entirely preventable, so I don't think Tesla is at fault here.

What will be critical though is how quickly they respond to this problem and one can only hope they learned from their previous blunders.

 

[dirkhh]

I would not appreciate such an action. I needed every bit of range yesterday for a ~600km trip (between -20C and -10C... arrived at a supercharger at one point with just 10KM RR remaining, and that was after adding a bit of extra juice thanks to @3s-a-charm). I used range mode knowing that I may be affected by this issue, but that it would likely be rectifiable. As it turns out, no such error condition during my journey. I've been in range mode almost exclusively since .167 was installed, and have been lucky.

That said, I was able to make a concious decision on testing fate, thanks to this forum. I agree Tesla should communicate with owners in this instance.

Yay TMC. But wouldn't you agree that you would have preferred the option to revert to .140?

 

[Brunel]

SpeakEV has reported on this and has a statement from Tesla that this is only affecting a select few cars and they are managing this on a case by case basis. They also state cars continue to have power and there is no risk. That seems to be the only open communique from Tesla.

 

[Andyw2100]

True.
All I am saying is that machines do fail, sometimes in the least opportune situation, which can lead to injury. Even with extensive testing, this situation may not have been entirely preventable, so I don't think Tesla is at fault here.

What will be critical though is how quickly they respond to this problem and one can only hope they learned from their previous blunders.

I, and I think many of the others here who have been critical of the way Tesla is handling this situation, aren't particularly upset about the initial bug. At least in my mind, that kind of thing may be somewhat unavoidable. I mean sure, it would be better if it were caught before widespread release, but it wasn't. I'm not interested in beating Tesla up over that. My concern is how they handled the issue, and are still handling the issue, from the point the bug was discovered.

You say "how quickly they respond" as if they could still handle this the right way. In my opinion, that moment has long past.

 

[dirkhh]

SpeakEV has reported on this and has a statement from Tesla that this is only affecting a select few cars and they are managing this on a case by case basis. They also state cars continue to have power and there is no risk. That seems to be the only open communique from Tesla.

(my bold)
Hmm - that's an interesting assessment. It seems we have at least couple of members who have experienced the issue and appeared to disagree with that assessment.
I would also disagree, but haven't had a chance to experience it, yet... (I know this sounds stupid... I'm trying to say that I'm anxiously awaiting delivery of my P85D...)

 

[redox]

It is extremely simple to revert back to the last working update until engineering can fix the issue.

Unless you have very intimate knowledge of the system, I'd be cautious about making such statement. Not all software is both backward AND forward compatible. Backward is a standard requirement for embedded systems (which I consider the car to be), but forward isn't always, as it brings a lot of complications in the code that deals with reading of persistent data (potentially written by a "future version" that you have no knowledge of at the time of writing your code). Those complications are usually better skipped in order to avoid bugs (complexity == bugs).

Lastly, nothing is ever "extremely simple" when it comes to over the air updates of an embedded system. It sounds simple, but the devil is in the details, and a tiny mistake can lead you to a catastrophic bricking of lots and lots of devices. Ask me how I know... :wink:

-- Greg

 

[dirkhh]

Unless you have very intimate knowledge of the system, I'd be cautious about making such statement. Not all software is both backward AND forward compatible. Backward is a standard requirement for embedded systems (which I consider the car to be), but forward isn't always, as it brings a lot of complications in the code that deals with reading of persistent data (potentially written by a "future version" that you have no knowledge of at the time of writing your code). Those complications are usually better skipped in order to avoid bugs (complexity == bugs).

Lastly, nothing is ever "extremely simple" when it comes to over the air updates of an embedded system. It sounds simple, but the devil is in the details, and a tiny mistake can lead you to a catastrophic bricking of lots and lots of devices. Ask me how I know... :wink:

Thanks for making that point. Rollback is something that needs to be designed into the system. If they haven't that set up (and I'm speculating that they don't) then simply sending out an older version is indeed potentially worse.
But then again, in a critical control system they SHOULD have the ability to do rollback.
Here's the ultimate insult to a developer of a critical control system: "hey, you realize that you are missing a safety feature that desktop Windows has had for a decade?"
Ouch.

 

[eepic]

True.
All I am saying is that machines do fail, sometimes in the least opportune situation, which can lead to injury. Even with extensive testing, this situation may not have been entirely preventable, so I don't think Tesla is at fault here.

What will be critical though is how quickly they respond to this problem and one can only hope they learned from their previous blunders.

I think most people here are in agreement that the criticism isn't on the power reduction issue, but rather how Tesla is not being more proactive. I can't imagine why it would be hard to pull up a list of vehicles pushed .167 from their OTA update database, identify the P85D models, then have local service centres give them a call.

Tell them there's a potential issue in rare circumstances under certain driving dynamics coupled with range mode where power gets reduced temporarily, so turning off Range Mode is recommended for the time being. Apologize for the inconvenience.

Customers who do not know about the issue may be a little concerned it made it to customer cars, but that would be offset by knowing Tesla is being proactive about this and prioritizing customer safety over losing face.

 

[PungoteagDave]

I feel it necessary to point out that when a Model S loses all power to the motor it is free spinning (neutral) already. The car would not and could not come "to a complete halt," so that particular statement is incorrect. Shifting into neutral would do nothing at all.

In my case when the car said it needed service I did lose power, car was in drive, pressing the accelerator had no effect. I had ~40 MPH worth of momentum, and the car did not slow without braking. I could have gone a decent distance if the light in front of me wasn't red.

I am sure you are well-meaning but you are wrong in saying it would not come to a complete halt. What is your definition of no power and stopping? It ends in a complete halt, which for me, and probably others, occurred on a very busy highway, in the left lane with no shoulder, requiring crossing several lanes onto the right shoulder while weaving through traffic as the car lost forward momentum. The car did come to a 100% complete stop. I never said that it applied the brakes or locked up. It lost total motive power, did NOT freewheel, and there WAS a drag from the motor(s) that went away and allowed the car to coast more freely when I put it into Neutral. That is the only way I was able to dodge a car and two trucks while getting over to the side while it was in the process of coming to a complete halt. There is no question that the car had negative drag until I put it into neutral. I did not have time to look and see if it was regenning at the time, but it is clear that taking it out of drive and into neutral bought me time and distance to get off the road. I do not actually know that the car lost all power - just that applying the accelerator had no effect - it was like the throttle was disconnected. Shifting into neutral released it and allowed coasting much further than if I had allowed the car to come to the complete stop that it was determined to accomplish.

​Still no call from Tesla. I have registered and reported the issue to the NHSTA.

 

[yobigd20]

​Still no call from Tesla. I have registered and reported the issue to the NHSTA.

I hope everybody that has experienced this issue has reported it to the NHSTA as it is a major safety flaw with the vehicle. Thank you, this was the right thing to do.

 

[MarcG]

Able to resume driving after asked to pull over *without stopping*

Alright, I've been driving for almost a week with 2.2.167 and RANGE MODE set to ON at all times, and until this afternoon I've only every gotten the "Car needs service - power reduced" message a couple of times (first one was actually on 2.2.140).

Well, on my 45-minute commute home this afternoon, I finally got the dreaded "Car needs service - pull over safely" message not once, but twice.
However, and I believe no one has reported this yet, I was able to resume driving *without stopping* by simply pressing the accelerator pedal again!

Here's the sequence of events as I remember them:

1. Three short & quick beeps to indicate a warning
2. Message stating "Car needs service - pull over safely"
3. P R N D indicators turned red, with the D still highlighted in large bold font
4. Motor power at 0 kW
5. Stepping on the accelerator pedal caused both yellow dashed lines to rapidly go from 0 kW to max power (top half) and 0 kW to max regen (bottom half) then disappeared immediately
6. Pull over safely message also disappeared
7. Power was restored immediately and I was able to resume driving & accelerated away


Both times, I was driving at about a constant 50 mph (80 km/h).
The first time, I had TACC on and was tracking a vehicle. The second time, TACC was off.

I wish I had time to take a video in slow-mo of the whole event, but it happened so quickly that I didn't even get a chance to take a photo of the message before it disappeared.
After the second time, I had my phone in video mode ready to fire - alas, the only message I got as I approached home was "power reduced".

At this point I have a theory: could it be that this error occurs because torque sleep is on but unable to "fire up" the rear motor again?
In both case I'm pretty sure torque sleep was active, and for whatever reason the rear motor needed to get engaged (perhaps for more power) but couldn't.

Anyway, I was happily surprised that I was able to resume driving without pulling over and stopping. I lost maybe 10 mph at most before power was resumed.

 

[fadkar]

I find it funny when people try to downplay this (along with any problems related to Tesla) as if it's no big deal. Even though this is a VERY serious and potentially fatal error on Tesla's part, they could have easily regained the consumer confidence by handling the issue better. I sure hope they at least learn from this lesson...

 

[wk057]

I am sure you are well-meaning but you are wrong in saying it would not come to a complete halt. What is your definition of no power and stopping? It ends in a complete halt, which for me, and probably others, occurred on a very busy highway, in the left lane with no shoulder, requiring crossing several lanes onto the right shoulder while weaving through traffic as the car lost forward momentum. The car did come to a 100% complete stop. I never said that it applied the brakes or locked up. It lost total motive power, did NOT freewheel, and there WAS a drag from the motor(s) that went away and allowed the car to coast more freely when I put it into Neutral. That is the only way I was able to dodge a car and two trucks while getting over to the side while it was in the process of coming to a complete halt. There is no question that the car had negative drag until I put it into neutral. I did not have time to look and see if it was regenning at the time, but it is clear that taking it out of drive and into neutral bought me time and distance to get off the road. I do not actually know that the car lost all power - just that applying the accelerator had no effect - it was like the throttle was disconnected. Shifting into neutral released it and allowed coasting much further than if I had allowed the car to come to the complete stop that it was determined to accomplish.

​Still no call from Tesla. I have registered and reported the issue to the NHSTA.

The it was phrased in the original post I was responding too it sounded like the problem came up and the car stopped immediately. I was clarifying that this wasn't the case. Of course eventually with a power loss the car is going to stop. I thought this part was common sense, which is another reason I figured the mention of a "complete halt" was in reference to an immediate event that would not accompany the normal events following a power loss (a gradual coast to a stop). Sorry for confusion on my part.

In any case, if the car was able to regen it wasn't a complete loss of power (like in my situation). If shifting had any effect (aside from shifting to park and engaging the parking brake) it simply wasn't a complete loss of power. When the motor has no power it can not regen. Regen requires actively maintaining the magnetic field that allows it. Impossible without power. Very little power, but it is required.

So either there wasn't actually a complete loss of power and the car was still able to regen, or there was a complete loss of power and it was not able to regen (free wheel). There really is no in between in this case.

In my case, when the error popped up (image in the first post of this thread) the car immediately behaved as if I had shifted to neutral. No regen drag, no response to the accelerator. As mentioned previously, had the light in front of me not been red I probably could have coasted for quite a ways.

- - - Updated - - -

Alright, I've been driving for almost a week with 2.2.167 and RANGE MODE set to ON at all times, and until this afternoon I've only every gotten the "Car needs service - power reduced" message a couple of times (first one was actually on 2.2.140).

Well, on my 45-minute commute home this afternoon, I finally got the dreaded "Car needs service - pull over safely" message not once, but twice.
However, and I believe no one has reported this yet, I was able to resume driving *without stopping* by simply pressing the accelerator pedal again!

Here's the sequence of events as I remember them:

1. Three short & quick beeps to indicate a warning
2. Message stating "Car needs service - pull over safely"
3. P R N D indicators turned red, with the D still highlighted in large bold font
4. Motor power at 0 kW
5. Stepping on the accelerator pedal caused both yellow dashed lines to rapidly go from 0 kW to max power (top half) and 0 kW to max regen (bottom half) then disappeared immediately
6. Pull over safely message also disappeared
7. Power was restored immediately and I was able to resume driving & accelerated away


Both times, I was driving at about a constant 50 mph (80 km/h).
The first time, I had TACC on and was tracking a vehicle. The second time, TACC was off.

I wish I had time to take a video in slow-mo of the whole event, but it happened so quickly that I didn't even get a chance to take a photo of the message before it disappeared.
After the second time, I had my phone in video mode ready to fire - alas, the only message I got as I approached home was "power reduced".

At this point I have a theory: could it be that this error occurs because torque sleep is on but unable to "fire up" the rear motor again?
In both case I'm pretty sure torque sleep was active, and for whatever reason the rear motor needed to get engaged (perhaps for more power) but couldn't.

Anyway, I was happily surprised that I was able to resume driving without pulling over and stopping. I lost maybe 10 mph at most before power was resumed.

Definitely was not able to restore power in this fashion when it happened to me. Pressing the accelerator had no effect at all.

I wonder if this could indicate Tesla actively monitoring and automatically responding to the error OTA somehow to clear it? I theorize this because I've noticed my P85D sending a lot more data over Tesla's VPN since the incident. It could be actively sending data to them to alert them to the issue if/when it happens, then some automated script jumps in and tells the car to clear the error remotely. Just a theory.