Tesla Is Being Investigated For a Software Update Meant To Limit Fire Risk (theverge.com) 35
The NHTSA is investigating a software update Tesla shipped earlier this year that was meant to address a few reports of battery fires. "Some owners have said the update, released in May, noticeably decreased the ultimate range of their Teslas," reports The Verge. "A few have even sued the company for alleged fraud over the issue." From the report: The NHTSA received a "defect petition" in September from a lawyer who represents some of these customers. He requested that the NHTSA investigate the software update in order to figure out if Tesla was hiding a defect in its cars that could have caused the reported fires earlier this year. "Tesla is using over-the-air software updates to mask and cover-up a potentially widespread and dangerous issue with the batteries in their vehicles," he wrote. Defect petitions require more information than the typical complaints the NHTSA receives. Crucially, they also force the agency to act more conclusively: following the investigation, the NHTSA has to either issue a recall for cars affected by the alleged defects, or else publish its findings on the federal register explaining the decision not to. The NHTSA will now attempt to put a finer point on this issue by trying to figure out precisely what Tesla did with that May software update. The petition that the NHTSA is reviewing specifically claims that Tesla should have recalled 2,000 Model S and Model X vehicles because of the issue.
At the time when Tesla issued the May update, the company said the update was out of an "abundance of caution" and that it would affect charge and thermal management settings.
At the time when Tesla issued the May update, the company said the update was out of an "abundance of caution" and that it would affect charge and thermal management settings.
Car Updates (Score:4, Insightful)
Having a car update itself without your knowledge or control is fucking ridiculous.
Re:Car Updates (Score:5, Informative)
Re: Car Updates (Score:5, Insightful)
Re: (Score:1)
Re: (Score:2)
You could say the same about an OS.. but here we are.
Re: (Score:3)
Here you are, on slashdot. Most of us are running Linux, and yes, we do choose when we update an OS.
Re: Car Updates (Score:1)
Most Slashdotters are probably not running a linux desktop system. And the incremental updates on a 'distro' can carry you along for awhile, but every year or two you have to shitcan everything except your home directory and plunk down an image from a new iso.
Or are you using git to update a userland source repository and rebuild it every few months? If so, why not get real and run NetBSD-current or something else that is consistently maintained year-over-year?
Re: (Score:2)
the incremental updates on a 'distro' can carry you along for awhile, but every year or two you have to shitcan everything except your home directory and plunk down an image from a new iso.
Up to about 10 years, depending which distro you use.
I just went through this recently, it had been about 5 years. Also, some distros update smoothly. Fedora doesn't require that you redo everything; you update more often, but you can do incremental updates forever. It is mostly the longterm-support distro versions that require a fresh install.
In what way is NetBSD more "real" than Linux? Does mommy give you an extra pastie for being a True Scotsman?
Re: (Score:3)
Having a car update itself without your knowledge or control is fucking ridiculous.
What about having car recalls without your control?
Re: (Score:3)
GMC vehicles with the built-in 4G LTE do similar software updates. The update downloads in the background, and as soon as you try to start the car, or put it in park, it will display an update prompt. The vehicle will not be usable (for up to 20 minutes) if you accept it. There's no indication of what is in the update.
Hopefully you're not in a hurry to get somewhere.
Re: Car Updates (Score:1)
They should learn from Microsoft. The oroper tactic is that the update downloads while you are driving along on your trip. When you arrive and try to shut off the engine the car refuses for about fifteen minutes while it applies the update.
Re:Car Updates (Score:4, Insightful)
Having a car update itself without your knowledge or control is fucking ridiculous.
Do you know how to program the ECU in your car? Know how to do a full tune-up? Would you know how to do it on any modern car? Even if you did updates manually, think you have an ability to read what's in them? Or refuse to update and still obtain support?
You really are rather delusional when it comes to this concept of control, so let me clarify. You have none. Just like the rest of us.
Corrections, as always: (Score:5, Insightful)
This requires more precision: all owners of a specific battery pack found in some old Model S and X vehicles (Tesla has gone through over a dozen different pack varieties over the years). Specifically, it was Tesla's first pack to incorporate silicon in the anodes.
This particular pack has long been known to be problematic - specifically, with lithium intercalation in the anodes. An earlier update had limited its max supercharging charge rate after 100 supercharging sessions. The latest update limits the max rate further, as well as limiting the max voltage. It's the latter that causes a range drop (generally under 10%).
Specifically, the person who filed the lawsuit is also the same person who filed the petition with the NHTSA.
Software cannot "cover up a fire". E.g. you can't have your house burn down and you not realize it because your computer had some "don't notice a housefire" program running. Software can however cause or prevent a fire. An update that prevents a fire is ameliorating a fire risk.
The software in question is the BMS. The BMS monitors battery health and controls the charge rate. If charging is conducted faster than lithium can intercalate into the anode, it can plate out as metallic lithium, which can be hazardous. Poorly managed charging can also adversely effect the lifespan of a battery. The BMS, however, can only work on incomplete knowledge; it has no molecular-level insight into what's going inside its cells. Developers have to create models for how it to make its decisions on what to allow in what conditions.
This would of course be an odd finding because it's the plaintiff's own claim that the software was a fix to prevent fires (this is their own assertion, which they're trying to prove, and part of the reason why they're getting the NHTSA involved). If the software fix actually does what they're claiming, then there would be no grounds to recall the ~2000 batteries in question. The NHTSA case is primarily a fishing expedition to help their lawsuit - to get the NHTSA to do discovery for their case.
The plaintiff in question is active on the Tesla forums and is fond of making a particular line of argument.
* The warranty on the original Model S, unlike the new warranty for the Model 3 (I think they've since updated the S warranty, though I'd have to check) - does not guarantee a particular minimum level of acceptable degradation.
* The vehicle was "not software limited" before, but now Tesla has imposed an "artificial software limit" on the batteries, mislabeling it as degradation in order to avoid liability, since they don't have a warranty degradation limit.
This is, however - as I'm sure Tesla will argue - a gross misunderstanding how batteries are managed. All EV battery packs, even when they're new, are software limited. That's the purpose of the BMS. The BMS has to guess at the state of health of the battery packs. If it's wrong, that's a problem. Updating the software of the BMS to more accurately understand the underlying health of a degradating battery pack is a perfectly reasonable thing to do. The underlying problem is that the anodes - which were perfectly capable of safely intercalating lithiu
Re: (Score:2)
Not sure what that link has to do with the current topic. We weren't talking about a factory fire.
Re: (Score:2)
Huh? Why are you linking to a 2018 article by (snicker) Lora Kolodny about smouldering that was extinguished in a matter of seconds? You do realize that fires at car plants are not exactly rare, right? Here's a GM fire on July 15th at Bedford [wbiw.com]. Now here's one at Saginaw [nbc25news.com] just a week later. Heck, a GM fire back in April was actually useful to GM [wkbn.com], as it delayed UAW from filing a lawsuit against them ;)
Re:Corrections, as always: (Score:4, Insightful)
In general:
Most people feel bad for the people who got these packs. The degradation, it should be added, is high by Tesla standards, but not by EV standards in general - indeed, it doesn't even remotely resemble the degradation in, say, the early Leafs. But Tesla has a reputation for very low degradation rates, and these packs most definitely have not lived up to it.
Tesla's refusal to replace them under warranty - as they don't violate the original S/X warranty, which does not warranty against going below a given capacity level - is understandable. But from a moral and brand loyal standpoint, I do wish they would have done something. I don't support the precedent of covering anything that comes up that's not warrantied (because people will exploit that to no end), but in this particular case, IMHO, should have at least done something like offering an upgrade to the latest battery at-cost, or offering some partial credit relative to the amount of range lost, or whatnot. That said, this lawsuit and NHTSA filing have pretty much slammed that door shut. Now Tesla couldn't even if they wanted to, as them offering to do so could now be used against them by the plaintiff. Which is unfortunate for the majority of owners who are not taking part in the lawsuit.
That said: while people throw a fit if anything negative happens, funny enough, you never hear anyone offering to hand Tesla more money when Tesla improves their vehicles after the fact. I've lost count of how many times updates have added range, performance, charge speeds, etc etc. Some pretty significantly, like the Model 3 V3 Supercharging, which more than doubled the max charge rate, or the early Model S/X "Torque Sleep" update for idling the less efficient motor in any given situation whenever torque demands were low enough in order to increase range and decrease consumption.
All this said... I'm quite curious as to what the NHTSA will say. There's not many packs in question, so there's not really any meaningful total liability in question. But I am curious as to why the degradation didn't show up in the accelerated aging tests for these packs in the first place. Tesla always prefers the approach of, "If you discover that anything's wrong, just fix it fast, for everyone, before it becomes a big deal", rather than waiting to be forced to fix something by the NHTSA (they've had exceedingly few recalls of things that were their own hardware). Which is great in general, but the downside is, if there's a negative consequence to a fix, people may be left in the dark as to why. Off the top of my head, I can mainly think of three examples of this:
* The original C-rate downgrade for the packs described above.
* The early "launch mode" limits (which were ultimately lifted)
* The new C-rate and voltage limit for these packs
There've been some other repair items of note like the original Model S drive unit longevity issues (they didn't leave enough slack between the parts, which led to excessive wear and metal shavings in the gear oil), but communication was usually pretty good about that one.
Re: (Score:2)
1) Tesla has turned a profit in two of the past four quarters, and strongly positive FCF in three of the past four.
2) "Survival" is based on FCF, not profits.
3) "At cost" does not harm any of the above.
Re: (Score:3)
Generally speaking the principal with products is that they should perform as advertised and the manufacturer should not downgrade them after purchase without compensation.
"As advertised" includes what the vehicle was like when you test drove it, or in Tesla's case when you received it since they don't do test drives any more. For example some Hyundai Konas in the UK have had a free upgrade to read speed signs because the demo models people test drove had that feature even though Hyundai didn't put it on th
Re: (Score:3)
All EV warranties and purchase agreements discuss the fact that the battery will degrade over time. No EV is advertised as maintaining the purchase range indefinitely.
Re: (Score:2)
Yes, but this is not degradation. This is a change to the original specification, i.e. the maximum voltage and charge rate.
It would be extremely odd if manufacturers could remotely gimp your battery after a year or two, perhaps to encourage you to buy a new car, and pass it off as "degradation".
Re: (Score:3)
Yes, it literally is degradation. These packs charged up to the nominal voltage perfectly fine and safetly and at high C-rates when new. They no longer physically can. The physical degradation going on at the cell level is what it is.
Attacking the notion of having the BMS properly understand the degradation going on in the pack is shooting the messenger.
Re: (Score:2)
That's an interesting way of interpreting it. The fact that they had to release a firmware update kind of undermines it though. Clearly it wasn't what they originally designed or specified, hence the need for an engineering change.
Re: (Score:2)
Man, you really love Tesla. Two treatises in a row in support of a corporation! Fascinating.
Re: (Score:2)
Would it be better if he somehow felt irrational hate toward them and posted a lengthy response against Tesla?
There’s nothing wrong with articulating cogent thoughts on any topic. If it happens to be positive towards Tesla, does that somehow make it wrong?
Re: Corrections, as always: (Score:1)
Likely, it makes for a lucrative paycheck. Don't kid yourself that the kind of personality that just luvves the Elon Musk business model would so cravenly support the company for free.
Re: (Score:2)
I think the bottom line is that there is a "beta" element to electric cars of Tesla's sophistication and range. The fact that they have gone through a couple of different battery pack designs seems to signify they are kind of operating on the bleeding edge to some degree of battery design and engineering.
If battery design and engineering was kind of settled, there would be a lot more electric cars.
Tesa being Shorted again (Score:1)
By any chance would any of these customers work on Wall Street and put down a big short on Tesla?
Re:Tesa being Shorted again (Score:4, Insightful)
The NHTSA discovered... (Score:3)