Hacker News new | past | comments | ask | show | jobs | submit login
Tell HN: 3G sunsetting is remotely killing every Subaru Outback battery
454 points by siftrics 9 months ago | hide | past | favorite | 215 comments
Subaru partnered with a service called STARLINK which shares your location in the event of collisions, among other things.

STARLINK intermittently tries to phone home by hitting 3G towers.

Now that 3G is shutting down, the digital communications module (DCM) gets stuck in an infinite loop of

1. Phone home, expending battery charge 2. Fail, because 3G doesn't work anymore 3. Go back to step 1

This effectively remotely drains the battery of every Subaru Outback built between 2015 and 2020.

Even if you drive your car every day, its battery will die and you won't be able to start it.

Other models are probably affected, too.

There was a class action lawsuit. But this is a pretty egregious engineering oversight, given they were still producing defective cars in 2020, hardly two years before 3G flipped off.

Will a brand ever produce a reliable, mechanical car? Why should 3G towers have anything to do with my car being able to start?

Can confirm. Had 3 or 4 batteries replaced (under warranty) before the local dealership informed me that "some DCMs cause parasitic battery drain". I brought it in, they ran some sort of test, and ordered me a new DCM (also under warranty). It took a while for the part to arrive, so they pulled the fuse. As other posters have noted, it took out one of the speakers (music sounded only mostly normal) and both the mic and speaker for Bluetooth comms were dead.

Previously, there were times that I would drive 100 miles one day, and I couldn't even open the doors using the automatic locks on the following day. It's been about a month on the new DCM and so far I haven't had any issues, so here's hoping...

This is an excellent example of how large corporations can't plan products normally any longer. You could take a college class in embedded systems design and you could ask all the students to come up with examples of edge cases where an SOS system in a car could have issues, and you're pretty much guaranteed that someone will mention both being out of range of cell towers and being old enough to be incompatible with cell towers.

Discussion would then lead to what to do about it, and busy-waiting in a loop trying to contact cell towers again and again would certainly not be a solution.

But can a corporation do this? Apparently not. I honestly wonder if designers are actively discouraged from caring about anything beyond the projected life of a product, because it could negatively impact later sales.

To me this seems more like 10 teams of people worked on the functionality, each isolated internationally and corporately from each other, and had to solve their own contingencies, with no entity really in charge of the system as a whole.

One time I bought a car and the person trying to scare me into purchasing a warrantee told me that there were "60 computers" in like a 2015 Hyundai Elantra, which I understood to mean ICs, but that still seemed like an excessive amount and a product of commoditizing every part of a car.

Or "we'll deal with that issue when it comes up" but then it's no one's job to deal with it when it comes up and everyone either assumes someone else is dealing with it or isn't aware of it at all.

Your comment is funny to me and I agree.

It reminds me of the "Can LLMs think?" comments. Now I'm left wondering what other "classes of reasoning" are impossible for corporate entities, what they may be blind to simply due to their structure.

That makes me wonder at what point does a business lose the ability to think?

You and your friends in the garage are nimble and hard working, thinking of some way to make it, but at some point a company crosses a transition line and stops being capable of autonomous thought and becomes a machine both greater and lesser than any human.

I would hazard a guess that this transition happens at the point where the company would continue to operate even if all of the founders suddenly died.

Second guess: the moment that something starts happening in the business that the founder does not know about or understand.

Corollary: it’s inevitable

Decoder had an interesting interview with the CEO of Ford about how hard it is for car companies to maintain system standardization as vehicles become more computerized: https://www.theverge.com/2021/5/20/22444294/ford-f150-lightn...

> You could take a college class in embedded systems design and you could ask all the students to come up with examples of edge cases where an SOS system in a car could have issues, and you're pretty much guaranteed that someone will mention both being out of range of cell towers and being old enough to be incompatible with cell towers.

Those people are not in management.

A company that uses planned obsolescence probably see's these bugs as more of a feature. Out of warranty, out of mind.

I searched for Subaru Technical Service Bulletins (TSBs) mentioning "STARLINK", and found this one:


It mentions a "3G Sunset Update".

A further search for "3G Sunset" turned up a few TSBs that apply to various Subaru models and years.

You can search for TSBs here: https://www.tsbsearch.com/Subaru

I'll skip ahead and tell you what you'll eventually find:

- Subaru has a fix for this issue.

- Subaru's fix has dealerships use proprietary software to put the DCM module into "factory mode" (suspended).

- It will also have dealerships physically disconnect the Starlink-related buttons (e.g. SOS), because if they're pushed it will "wake up" the DCM.

- There is no assurance the DCM won't wake up again on its own (it has its own backup battery which can fail).

- Getting dealerships to conduct the above fix is a giant pain in the butt, and they'll charge you hundreds of dollars. Few customers have even got that far, most just get an expensive "diagnostics" bill with absolutely nothing to show for it. Most won't do the repair unless you can "prove" a problem, even though it is more work to prove the problem than to proactively conduct the repair.

- Subaru won't release the software, so you can do it yourself.

Condensed further: there’s no recall to fix it during maintenance milestones, and they aren’t paying for the workaround, which is a shitty one.

Therefore, lawyers.

I work in automotive and I am genuinely surprised there isn't a NHTSA helpline of some sort to escalate these kinds of issues to force a recall.

Would NHTSA even care about this? It's not a purely safety issue (I mean, I can see people making the argument it can leave you stranded somewhere but I didn't think that type of thing is where NHTSA gets involved)

If you shut off the car, it's pretty low on the safety issue scale if it doesn't restart.

Maybe Dept of Transportation (DOT)

NHTSA is an agency of the USDOT

Condensed even further: Join the class action lawsuit.

Didn't replace my nice and reliable 2011 Outback with another Outback due to the way Subaru handles this issue.

You are burying the lede - Subaru is reimbursing to dealers

I have a 2016 Subaru Forester with a seemingly parasitic drain issue. Every time I go to start the car, it's like rolling the dice whether it has enough battery power to fully start. I have a portable jump starter that I've had to use at least 20 times or so in the last couple of years. This has been happening even well before the 3G sunset period, so that is not the only source of the battery drain problem, for which there is a class action lawsuit: https://www.subarubatterysettlement.com/.

Everything I've read online about the issue points to the STARLINK system, and the common wisdom in this thread and elsewhere is to pull the DCM fuse. Unfortunately, my Forester has no such fuse, so I'm at a loss what to do with my car.

It's so frustrating, because my wife has a Subaru Crosstrek of the same year that has never had a battery problem, even with the OEM battery. She has a base model without any additional "upgraded" electronics, so that's the likely culprit in my case. I'm currently on my third or fourth higher CCA battery and had to jump it again a few days ago.

I haven't taken my car to the dealer for the parasitic drain issue, but previously tried to get dealership service for an unrelated entertainment system issue related to USB media playback freezing up the entertainment system. That fix was unsuccessful, and I have no faith in Subaru mechanics being able to diagnose and fix a potentially more nebulous battery drain issue. It doesn't help that the dealership is now 45 minutes away, so I'm not interested in wasting even more time on a hit and miss solution.

Ironically, I bought Subaru because reliability was my number one concern when purchasing a vehicle, and I'm not sure I would do that again.

I just went through something similar with my vehicle, had a relatively new expensive AGM battery about halfway through its 3 year warranty. Would not start if left for 4-5 days which slowly crept down to 2 days.

Seemed like a parasitic draw but pulling fuses and using a multimeter verified idle draw at 30 milliamps which is within spec. If you find high like a few hundred milliamps you can start pulling fuses to narrow down offending item.

In the end it was a bad a battery, but was difficult to get the warranty replacement because it could be charged and work again so long as you ran it every day.

It's relatively easy to verify parasitic draw and make sure it's in that 30 milliamp range, with even a cheap multimeter by disconnecting negative and running through multimeter. Tricky part with modern cars is getting it to go fully to sleep, anything can trigger wake up (like opening the door) and it can take a few minutes for it to fully sleep to idle draw.

I learned over the pandemic that a Toyota Camry Hybrid would eat its 12V AGM battery very easily. In this case, it is a deep-cycle AGM and isn't even used to start the ICE. It's "just" to run accessories and boot up the power control system that engages the much larger hybrid battery.

One such Camry in our extended family was parked for a few months and the battery failed. We got an expensive OEM replacement and tried to make sure we drove the car every week or two. It failed again in just under three years. Meanwhile, we have an entry-level Audi that also often gets only a short local trip every week or two. It is still on its original AGM battery after 6 years, and this one has to actually start the ICE.

So, I am disappointed to find that Toyota engineers could not be bothered to apply whatever basic power management logic the Audi folks did to protect the battery. Worse, the Camry has this big NiMH traction battery, and I would have imagined some clever backup strategy to infrequently top-up the 12V battery if parked too long, so it would be reliable.

In the end, I gave up and installed an old-school battery cut-out switch, so we can just disconnect the 12V battery if not planning to use the Camry for more than a few days. Every time I operate it, I'm giving the Toyota engineers a rude hand gesture in my mind.

Smart electric drives prior to 2016 used their “[not] clever backup strategy” to cyclically try to keep a failing 12V battery charged and that could result in flattening the traction battery and eventually bricking it.

Ha, that does sound somewhat un-smart. Of course, when I imagined clever, it would include enough state and timer-based behaviors to recognize a bad trend and protect the traction battery too.

It should be feasible to monitor voltage levels and rates of change and signal an error condition rather than silently destroy itself...

There is an ECU flash at the dealership that supposedly fixes the parasitic battery drain. I've replaced my battery a couple of times, but haven't had an issue since the ECU flash.

I would get the battery replaced at a main Subaru dealer.

Then keep coming back to get the battery replaced when it "fails to hold charge".

If every owner did that it would end up like the region restriction in DVD players.

i.e. The retailers would force the manufactures to turn it off because they don't want to deal with the cost of fixing an artificial problem.

Oh and I would start posting to car websites and forums about how unreliable my car was. Marketing teams do care about getting that reputation.

You can’t, the battery will pass the battery tests every time. And who has time to go to a dealership over and over.

There is no money in it for them, and you are wasting valuable time. You can fix the problem yourself by getting the head unit replaced with any CarPlay/androud audio one at a car stereo shop for $1,500 and move on with your life.

Great solution for this problem: Go fix it with $1'500 - which is a lot of money -, for something you paid and someone else is responsible for breaking it.

Life isn’t fair sometimes. This thing even went though the courts and somehow the onus is still on the buyer to prove something was wrong, when it is clearly a design issue.


I guess there is enough noise around the issue these days to get the dealer to properly fix it, but back in 2019/2020, no one believed you had a tiny power drain, and it would have been very costly to prove it (including time), and then who knows if you could litigate successfully.

Hell, if the power drain was only happening while not able to connect to 3G, the issue would be so location specific/sporadic enough, that good luck proving it even to yourself.

Who can even find an appointment at the dealership? In my area, it’s like scheduling for a doctor. Months out.

> region restriction in DVD players [...] force the manufactures to turn it off

Where did this happen? Some players can be unlocked and others are multi-region, but to this day DVD players are still region-locked, are they not?

This was pretty much the norm in the UK.

When the big supermarkets were still selling DVD players they used to sell them region unlocked because people would return them as faulty if they didn't.

If you go to some where like The Range they still sell DVD players and they're all sold as "multi-region" as far as I can see:


Subaru has published a Technical Service Bulletin that appears to offer a free firmware update to switch the Starlink system to use VoLTE. Affected vehicles are:

2016-18MY Legacy/Outback 2016-18MY Impreza 2016-18MY Crosstrek 2016-18MY Forester 2017-18MY WRX

See https://static.nhtsa.gov/odi/tsbs/2021/MC-10189606-0001.pdf for details. And, yes, Subaru is reimbursing dealers for this work.

That's actually pretty amazing, and luckily the modules support VoLTE.

BMW modules only supported 3G (they were older though).

There isn't really such a thing as a module that only supports 3G...

Anything could support the minimum needed to connect to an LTE network with the right software update, but modem chipset makers don't make any money rewriting old firmware for old chipsets, so they refuse to rewrite the firmware to support lte.

Chipset makers, probably correct but there aren't that many companies that make them. Depends on the analogue front end if they support the new frequencies.

I know the 3rd party module makers do tailor their modules to regions and the 'world wide' models can be quite pricey. On the region specific versions they don't include some of the hardware to physically operate on frequencies they don't specify. Sounds crazy, but a few cents here or there is worth running multiple SKU's.

There are probably other issues such as radio compliance and patents that also prevent an upgrade.

The VoLTE upgrade isn't so surprising. For a long time we had 3G/4G modules however the mobile networks didn't support VoLTE so everything would fall back to 3G. Once everyone got their act together, a quick firmware upgrade and VoLTE worked perfectly without adjusting the hardware.

I did not know this.

Now knowing this, could we (theoretically) update the firmware of an iPhone 3G so it works on a 4G network?

Probably not. In general, this sort of thing is buried in the often unchangeable microcode of an RF ASIC. (Which makes sense, after all, the vendor is selling an RF modem for a specific band/modulation capability - they don't expect it to change, certainly not after they've built/configured it.)

If that's the case (as is common), then there is really nothing to update, short of scraping off that chip or module and replacing the hardware.

Yes. But beware that doing so is probably multiple man-years worth of effort.

Mine is a 2014, guess I narrowly dodged that bullet. Thanks for posting this.

Thanks for the link. I've got a Legacy that's had a 'weak' battery for a while now and it seems like it may coincide with this 3g sunsetting..

Kind of crazy that they push the whole outdoorsy camping image in all their ads and build something so flawed. How does that work? Go explore the great outdoors but don't go too far, or you're battery is gonna die if there's no cellphone service? I don't know about the US but it's not hard to go that far out in Canada!

Yeah this was actually pretty scary for me because I do go car camping in very remote places with no service at all. Luckily I'm fully prepared for my car not being able to start.

It's just sad because I bought a Subaru because it's supposed to be a reliable, outdoor adventure car. To see it fail this way is just sad.

Yeah, I live in Vancouver and I can get to a no-cell-service location in like, 30 minutes. This is blatantly inexcusable.

For real. The real scandal here isn't that this is happening as 3G is phased out -- it's that it was always a problem whenever you drove anywhere without cell coverage.

Which is precisely when keeping a working battery is most important, since you can't call for help, being in an area without cell coverage!

2018 Outback owner here. A truly outdoorsy driver will always carry booster (a.k.a. jumper) cables and/or a boosting-charging pack if the vehicle doesn't have a manual transmission and no other vehicles are likely to be around. I would never go even slightly off road in my Outback without a boosting-charging pack amongst my other gear. A manual transmission has allowed me to bump-start dead vehicles many times over the years, but not this automatic model. Anyways, Subaru needs to get this all cleared up.

I had the battery failing recently. Took it to the service, they did not know what the reason was and said they would test the whole car to find where the drain happens for 500 USD. I was about to drop it off next week and I read this. So is there any official help from Subaru ?

Supposedly just removing the fuse could do it (and recharge the battery completely if it hasn't been drained beyond saving).


Please allow me to suggest engineering fixes:

1. All vehicles with 12v starter batteries should also have deep cycle batteries to power the accessories. Chevy Suburbans had that for decades, but some people laughed at vehicles designed to actually run and drive. That's fine. Enjoy the vehicle you picked.

2. It's my car darn it. Any and all communication with it needs to be easily read by my in plain English. I also need to be able to veto any communication, in an easy-to-use, plain English system. This second option would require legislation. The new law should apply to everything made. If the owner is also the operator, then the owner can read and veto all communications.

3. In the interest of national security, all products made must have a minimum of reliance on the outside world. Those connections that do exist must be as standard as possible. Once against, this requires legislation.Any device that refuses to operate without an Internet connection should be illegal. Phone-home DRam should be illegal.

We get what we deserve. The only downside to my ancient pickup truck is that I have to clean the hybrids out of the wheel wells about twice a month.

We had an older gmc, and it was great. A newer one came from the factory with a bug that would zero the battery in under eight hours sometimes (it left the ABS on). This happened on the first tank of gas. The local chevy dealer refused to fix it and claimed our warranty was void at ~ 125 miles because the dealer we bought it from new had installed some counterfeit electrical components. (There was a factory recall for our exact problem)

That was the least of that truck’s problems. Never buying a GM product again.

Also, the old v8 GMC had a carburetor, but still had better fuel economy than the new v8 GMC with fuel injection and that turned off half the cylinders when possible.

Our Ram is slightly better, but still pretty damn bad vs the old GMC. I’d recommend any other brand over GM or Stellantis at this point.

This seems to be a very common type of software bug. On fail retry, without any limits of how frequent or how many time.

It doesn't need to phone home when the ignition is off and could have been programmed to go into standby. You can hardly crash a car with the ignition off. Even if you do, it could be woken up using an accelerometer and try to phone home just a a few times. This is shitty engineering. I am quite amazed, because the Japanese normally don't do this. Probably another stupid marketing decision.

Most cars I’ve ever had don’t have auto-off headlights because the extra parts cost a few dollars and that eats into margins and growth must continue infinitely thanks to rent-seeking shareholders. Capitalism will corrupt engineering whenever it is in the interests of capitalists to do so.

You don't need extra parts, these are no headlights, but an IoT modem that already has power saving fearures on chip. It just needs an ignition signal and programming to go to sleep when it's not there.

You must have had very crappy cars, because I rented a 2022 Renault Clio this spring and it had very good ergonomics. I was quite amazed. Only an older Hyunday i3 that I know of had annoying electrical quirks. But hey, there are tiktok videos on how to steal current Hyundai and Kia cars.


Communist automotive engineering departments by contrast are notorious for their near-decadent focus on customer satisfaction.

Is a lada really much different in that regard as a cold war era gm car?

Yes. Yes it is. The Lada's engine compartment isn't choked to death with hundreds of vacuum lines trying to reduce/control emissions...

If only there was a simple way to exponentially limit your retries and back off a bit when the other side is constantly failing.

Exponential back-off is fancy. There are a million repeating behaviors in a large automotive codebase. Maybe it's too arduous to code up exponential back-off for all. Even if you have a simple library to do it, it could introduce some complexity.

But it's quite easy to set timers to attempt things at regular intervals, rather than in an infinite loop until they succeed.

One attempt every 5 seconds (assuming a 5-second timeout for the request) is 17,280 attempts a day. One attempt every 5 minutes is 288 attempts a day. One attempt every 15 minutes is 96 attempts a day.

A simple repeating timer is a fantastic solution.

Does the car need to always send telemetry immediately? Can there be a 15-minute latency? If so, just put that on a timer and enjoy 99.4% power savings for offline cars, with basically the same quality of telemetry when they're online.

But what if a car needs to send telemetry at a given moment? Just move up the next timer function call.

But what if I want to get telemetry when the 3G signal is intermittent and briefly available? Just call the function once the 3G signal is available. Do this with a separate cool-down for this behavior.

But what if I need telemetry every second? Just dump all telemetry events into a buffer between successful sends, slap a timestamp on those events, and you can now send data for the past X hours all at once. Even from the black spots.

There are so many options (even less fancy than exp back-off), but this was probably built by a junior programmer without any support network. This feels like a very common lack-of-experience mistake that would be caught in code review if the reviewer cared.

> Exponential back-off is fancy

It's literally just a left bit-shift? It's as easy as a linear back-off. The only thing less fancy is no back-off at all.

Ethernet on coax cables did it already after a collision (late 70s, thats 50 years ago). Funny how basic things are today to much work, but animations everywhere are a must.

No. It's setting an async timer in a module that might not support that at all. It's establishing a variable for that in a scope the programmer might not normally have access to. It's making decisions about how quickly to back off and is there an eventual minimum floor. It's making decisions about what events ought to reset the timer, such as starting the car. And it's about creating a series of tests to verify this behavior, which is awfully non-obvious when the behavior might take 10 minutes to verify and tests are supposed to complete in 5 min.

It's fancy.

If you cannot mock that controller state machine properly, you should not be writing software that goes in cars.

At least the penalty is you just repeatedly cause every car of the model in the entire country to fail instead of crashing into a wall, but this kind of thing is why the T in IoT is for "shit".

I agree completely.

And it just supports my point, which is that it's not "literally just a left bit-shift".

It's quite a lot of work to do, actually. Which -- of course -- should be done. But there are commenters here acting like it's a single line of code.

That's not work unique to exponential back-off. You'd have to do all that for any back-off, which is why I said it's more complex than no back-off. Exponential back off literally is about a line of code difference (change "x += a" to "x <<= 1" (ok, a different radix isn't so silly easy, but anything is better then no back-off). All the rest, maximums, wrap around protection, reset events, all needs thought in either case. Actually the pathological cases in a linear system are perhaps sneakier as they could creep up on you ("it won't overflow, right? That would take months!")

But I think we all know what actually happened here was a rush to release and then a failure to follow up with "the update that actually does it right". The Jira ticket is probably still "open - should do, version: next" somewhere. And there's a "told you so" email chain linked in there somewhere.

Perhaps you are right. Though it seems easier to reason about a system with a hundred timers running at a set cadence than a hundred exponential back-offs. Maybe I just have more experience with the former.

I guess you could always cap off the exponential interval increases at some max interval, making it basically a repeating timer.

Though there's still the question of why would you need this for telemetry. It doesn't feel like this type of "marketing" telemetry needs to be quite real time.

Most (actually almost all) IoT systems are not well-thought-out in this regard. It requires a bit of paranoiac foresight. Designers far too often think it's really, really important that remote devices stay connected to the mothership, when in reality, most IoT devices would be far better off NEVER requiring a cloud connection, but leveraging one if/when available.

*Rule One of IoT network design: NEVER assume that you even have a network * If at all possible, and to the maximum degree possible, an IoT device must work whether there's a network available or not.

"I guess you could always cap off the exponential interval increases at some max interval, making it basically a repeating timer."

No you MUST have a max interval cap. Exponentials and all, you'll soon be waiting days or months for a retry.

It's ok, if you wait long enough it'll go back down to a small number. Or maybe a negative number and then the car crashes?

Please cap your numbers for both user experience reasons, and safety reasons.

Yes, have dealt with this before. Nothing like doing a fault analysis and telling someone that it'll fix itself in a few months.

A capped fibonacci sequence is a very common back off timer.

It’s a left shift and a non-volatile* variable to store it.

* at least across module sleep/wake

Also good to sprinkle in some randomness.

what's a left bit shift ;)

Not to argue the overall point (retrying on a tight loop is not a good solution), but many of the models support communication the other way—e.g., remotely starting the car, adjusting the HVAC, etc.

So there is some justification for trying to keep a cell signal established or checking in more frequently than every 15 minutes or so. Someone attempting to remote start their car not having it actually respond for 15 minutes is likely to already be outside trying to leave and pissed off that they’re walking out to a cold car with iced up windows and thinking that the “remote start didn’t work”.

When I was involved in certifying new devices to get on a carrier network, the carrier insisted on exponential backoff as well as some randomness so you wouldn't get millions of voic..devices crying out at the same time when power was lost or restored.

It's not like this is a new problem, and it's been solved for decades - it's just good network design.

One example: This is exactly why randomness is required by Ethernet's ARP protocol: to avoid ARP storms when the power comes back on and all nodes try to get back on the network at the same time. (And ARP is still powering most IP connections (v4) today, since IPv6 is painfully fiddly by design, so no one uses v6 unless they have to...)

You mean the concept I learned week three as a junior developer (thanks to great mentorship)? Yeah, probably absolutely impossible.

You're lucky. I've seen entire engineering orgs convinced of the opposite, and highly resistant to improvement.

But I need to unit test each recursion else I won't have completed test coverage!

It sure has been an issue with every cellphone I've ever owned, their radios flip out when I go somewhere without signal (tunnels, the subway, into nature) and will thrash the battery in a mad search for connectivity.

Outback’s of this generation also have an unresolved battery drain issue where if you leave the tailgate up, like you might while tailgating, the battery will drain even if all lights are off.

The solution is to put a carabiner or similar in the latch so it believes it’s closed.

Completely ridiculous issue with an otherwise great vehicle.

My guess, it’s a safety check in a loop because it’s unsafe to drive if that hatch is up.


It’s understandable for them to be checking for the hatch to be down, given that both your brake lights and reverse lights are on the hatch and only seen if down.


The brake lights on a 2016 Outback are not on the hatch. They’re on the portion of the lamp permanently affixed to the vehicle.

Also, why would checking if the hatch is open or closed completely drain the battery in a couple of hours? Not understandable to me.

>>Also, why would checking if the hatch is open or closed completely drain the battery in a couple of hours?

modern shitty engineering approaches. nobody cares about optimisation, about the power consumption, etc.

A compelling reason to avoid “connected cars.”

Better to go with one which pairs with your phone if you even need this functionality.

Connecting an EV means getting better battery efficiency over time.

I work with batteries in the EV industry (not inside cars, though, but lithium-iron-phosphate), and the list of planned improvements is not short.

There’s a gap between when a product is ready to be used and when most of the known optimisations are done. And there’s a big risk factor with the company suddenly deciding to depreciate your thing.

I try to only have only “dumb” appliances in my home: kitchen, washing, lighting. If I couldn’t avoid buying with wifi, it doesn’t get enabled.

> Connecting an EV means getting better battery efficiency over time.

Then ask for wifi credentials, or ask to connect through a phone. If I don't connect a device, it's because the device is not allowed to be connected. Any device that thinks it's important enough to ship with its own SIM card using an account I don't control, rather than putting the user in full control of connectivity, is a device I will never purchase.

there’s another design flaw in the outback where the battery drains if you have the rear lift gate open too long. happened to me the day i was moving my family out of state and loading up the car all day, we were about to leave and the battery was totally drained. later looked it up and its apparently a common issue

I just found out my country also started deprecating 3G last year and will deprecate 2G in 2025. That is a bit sad, as I've just powered on some old 2G phones and they seemed to work just fine, but I guess it allows us to move forward without the tech debt.

Tech debt in this case seems like artificial scarcity

Lack of spectrum and/or tower sites is real scarcity.

LTE and 5G are supposed to be better able to share spectrum, to the point where they can share the control channel, and multiplex the data channels with either encoding, but 2g and 3g can't do that, so you have to dedicate at least minimum sized blocks.

Running one min block on 2G does wonderful things for ensuring access, but it's expensive in spectrum, especially if each network needs to do it. (Because cooperating between networks is really only a thing if mandated, or out in rural areas where there's little demand)

Do you still use dial up?

That's pretty lame. What IP you wanted to drive where there is no cell connectivity rt all, say Mongolia?

Seems that your battery would fail when you need it the most: namely, in the middle of nowhere with no cell coverage.

Ironically, it seems that this only happens when you are within service.

Another commenter speculated that when you're out of service, the software correctly identifies that there is no tower to connect to and stops trying. But when you're in service, the code repeatedly tries to do the 3G handshake because it can see a valid tower.

It may happen with a lot of devices if 3G appears browned out to them and not just plain unavailable.

Or the park in a garage with no coverage?

One can only hope that "smart" will have very bad associations in the future (for the general public).

If it doesn't already (or 5 years ago...), then I don't have much hope for the future.

There are huge swaths of areas with no service in the US that people regularly visit.

Maybe you could disconnect the battery cable every time you stop, and reconnect before you go?

In older vehicles that would reset the clock. No idea what impact it'd have on a more modern car than what I drive though.

It'll reset the engine control unit (ECU), which will cause your car to run rough constantly as the computer has to relearn how to adjust the engine tuning from scratch every drive. You'd be better served using a trickle charger.

Resetting the ECU will also cause the vehicle to fail inspection unless it's driven long enough beforehand. (Here in Upstate NY, anyway)

that is not how ECUs function Some Transmission Controllers will adjust to driver behavior, however engine controller default behavior shouldn't cause 'rough' experience I worked in automotive testing for over 10 years and constantly measured and analyzed vibration data on vehicles. Many prototype vehicle batteries die due to early release software that causes parasitic drain and I have never noticed any rough behavior just because the battery died and got replaced. There are many sensors and closed loop control systems that monitor strange engine behavior and adjust timing. This happens very quickly within few engine rotations. If there was an issue you would notice it for at the most for couple of seconds.

Yes, it's a closed loop. The O2 sensor in the exhaust rapidly adjusts the mixture.

The whole 'learning' the air fuel mixture to adjust the stoichiometric ratio over a longer period of time does not make any sense. You have the air mass, O2, fuel, temperature and sometimes more, you can calculate the mixture near instantly.

Maybe someone way back said you have to warm the car up and that takes a few minutes. Modern fuel maps achieve a fair idle from near dead cold. It's a solved problem.

This is strange. I know a lot of countries are sunsetting 3G and 2G in terms of consumer mobile usage. But they leave IoT alone for a few more years because of contract and money in it.

It’s contract only, mobile carriers are not happy that they have to maintain the 3G equipment just for IoT.

Spectrum is a limited and highly contested resource. There is very likely no way for 3G to coexist just for IoT.

I have a 2017 Subaru with STARLINK. It works just fine. It does run the battery down if the key is close enough to the car to wake it up, but far enough away that the signal is broken and the car goes back to sleep. Keeping the battery 5 feet away from the door instead of next to the door solved that problem.

I have a 2017 Subaru forester touring 2.0, and replacing the head unit with a cheap Sony CarPlay one fixed my problem. We changed nothing with where we place the keys.

Not once has the battery died after that, and it has been ~3 years. With the OEM starlink head unit, we would need to jump start every single week.

Needless to say, Subaru’s lack of willingness to compensate us lost them a customer for life.

If the DCM is in the headunit, that seems plausible. What one did you get?

I got this one:


It works fine, but if I were to do it again, I would have bought a better quality one with a bigger and more responsive touchscreen.

I cheaped out because I did not know if the new head unit would fix the battery drain issue, and if it didn’t fix the issue I was going to get rid of the car.

"It works just fine"? But you have to keep the key the correct distance away to not drain the battery? I mean... that's just ridiculous. I can't believe cars are simultaneously so advanced and yet so terrible.

> Keeping the battery 5 feet away from the door instead of next to the door solved that problem.

Do you mean the key? Otherwise how are you moving the battery?

The key, yes.

"You're holding it wrong"

This isn’t contributing much, especially because it’s not a valid comparison. The iPhone 4 issue only affected people who used their phones without a case and held it unusually tightly or while sweating profusely (I tested this when news came out and it was not easy to reproduce) - not great, and an embarrassing gaffe for a company specializing in hardware, but very far from being equivalent to a problem rendering the device useless for all users.

Not all users, only the users that are holding their cars wrong i.e. parking their cars too close to their keys.

According to Subaru's website & TRB, this only impacts vehicles from 2016-2018



Mine's a 2019 and pulling the DCM fuse resolved my problems. Could be unrelated, but I feel obligated to share.

Did you lose bluetooth or front speakers?

Front speakers and microphone are lost. Bluetooth audio still works. I can still stream music over Bluetooth.

They offered a free upgrade a couple years ago.

I didn't/don't have a "free upgrade?" Are you talking about this battery fault or 3G upgrade to 4G+ for people who PAY for Starlink, that's only a tiny subset of the owners.

Also, that free upgrade for current Starlink subscribers expired two years ago.

Wouldn’t the upgrade fix the battery issue?

Anyway, I didn’t upgrade (have an aftermarket radio), and my battery works just fine. In fact, the car stayed parked for two months in the Texas summer, and started just fine.

I'd rather skip the free upgrade and just have a car that works.


I don't think you know what that word means.

I never got confirmation that it was related to 3G sunsetting, but I had this issue when I stopped driving my 2017 Outback regularly at the beginning of the pandemic. The first trip to the dealership resulted in a "software update" that didn't seem to do anything. On the next visit, with photos I'd taken of my voltmeter showing parasitic drain, they replaced the DCM. I think they charged around $800 for that and a new battery since the original was toast from too many total discharges.

I reached out to Subaru of America and as a one-time "goodwill gesture," they refunded me for the cost of the DCM.

Crazy reading about this here. I just had this problem. Spent a couple hundred bucks for the local mechanic to track it down to parasitic drain from the SOS module (“Starlink”). He just disconnected it since I don’t pay for the subscription anyways.

Somewhat related in 2006 I had a Motorola A1000, very early iPhone sized full screen smart phone with Stylus.

I took it to New Zealand on roaming and despite having a connection it spent so much time looking for the home network (“Three” in Australia) the battery would last half a day.

This probably explains recent battery issues on my 2022 Outback. Took it in, they admitted it happens frequently but didn’t explain. Replaced with larger battery for free.

Oh no, I hope it doesn't extend to that model year. I just bought a 2023 and saw posts about battery issues before I bought and was a little worried I might end up with the same thing.

Just bought a 2023 Forester, no problems yet but wondering if this problem is just for the Outback?

2022 Outback does not use 3G. Subaru have had battery drain issues for about a decade now across more than one model and generation.

I have a 2018 Subaru Crosstrek. Just this weekend I stopped for lunch and had XM on until it timed out ~ 10 minutes after key off. I’ve done this since I got the car. But when I went to start the car the battery was low voltage (which really freaks the computer out) and would not start. This story fits the fact pattern!

My daughters’ 2014 Volt that had a bug that would trigger a warning system that would turn off charging until it was reset. Paid a dealer $400 to reprogram (and didn’t do it right the first time and follow the service bulletin. Had to take it back). Write crappy code, get paid over and over.

How is 3G turning off different from people parking their car underground and not driving it for a week?

It seems that the battery only drains when it is within range of a cell tower.

Another commenter speculated that when you're out of service, the software correctly identifies that there is no tower to connect to and stops trying. But when you're in service, the code repeatedly tries to do the 3G handshake because it can see a valid tower.

This would explain why mine's been mostly fine since I moved to a place with an underground garage. Now I park 4 floors underground and the battery has only died once in the last two years.

Can you pull the breaker or cut the wire. Seems like it would be a simple ish hack but maybe access issues

I ended up pulling the DCM fuse, but it comes at the cost of no microphone for phone calls and voice commands and no front speakers (and obviously no STARLINK).

I did the same after having a shop investigate the drain. They didn't put all of the pieces together as the OP did, but they pointed to the fuse and told me I could pull it.

You can probably easily rewire DCM from always-on to ignition-on. No battery drain and working equipment.

Yep, you just need to move the fuse to a portion that's wired through a relay. Some engine bay fuse boxes have spare relay and fuse positions to add these kinds of things; otherwise you can buy a small automotive fusebox and a relay and a couple dollars in cable and crimp connectors and then you just need to find somewhere to mount it, likely above the outside of the glovebox or such.

15 minutes isn't a sufficient amount of time to reliably recharge a drained ICE car battery.

Ask any tow truck driver.

Edit: Cool, apparently you've completely changed your comment. No worries, it happens. The original comment read something like "I don't believe this, a 15 minute drive is more than sufficient to recharge my battery."

Edit: Shit, my bad, thank you @clippy, this is the second time I've made this mistake in as many days.

> Edit: Cool, apparently you've completely changed your comment.

It seems much more likely that you intended to respond to another post[0] and misclicked.

[0]: https://news.ycombinator.com/item?id=37971474

3G sunsetting fking over cars is not exclusive to Subaru. My 2017 lexus was also hit by this.

How difficult is to make a 3g proxy, something like the Stingray?


What's the [sustained] financial incentive to produce an unbreakable product?

The only way you're going to get this is through pride, be that national pride (nationalism) or professional pride. It requires care for the opposite party in the buyer-seller transaction, which the free market is essentially allergic to.

There is a financial incentive to be reliable enough to at least make it to the next sale with your reputation intact. Especially if it’s on a feature that I would guess was paired with a higher end trim which implies higher end customers.

You get to be the largest and most-profitable carmaker, aka Toyota.

I only buy used cars, and past the ~7-year-old range used Toyotas routinely cost about twice as much as any other brand for equivalent model/mileage/condition. Yet I still pay the markup because every time I work on mine I am astonished by the fanatical obsession they have with reliability. I'm sure this value retention is reflected in the profit margin on new Toyotas.

I just paid $98 for a tow last week for a 2015 Outback that went so dead the doors wouldn't open. They said my year-old battery was mysteriously worn down so I bought a new one.

Is this a different Starlink from the Elon Musk one?

Yes, Subaru Starlink is their remote connection platform that uses cellular. No relation to Musk.

I’m pretty certain that the battery lawsuit happened before the 3g shutdown.

So while I totally believe this is killing batteries, I also believe that these cars have more electrical problems than just this.

These days you can buy a battery bank which can jump start a car a few times. They are not so expensive (70USD), at least cheaper than road aide assistance.

Is this just Outback, or other Subaru models with STARLINK?

So instead of complaining on HN, it was a simple Google search to find out that Subaru has already offered a free repair for affected vehicles.

You went through the trouble of searching about it, finding results, and then commenting here, but not including the link(s) you found? o_O

Someone had already posted links. I just wanted to do the grumpy, snarky old man thing…

Irrelevant. You could have provided the link as a decent human being. Instead you decide to get into a petty argument?

It was literally the next comment down. They didn’t even have to Google it. It’s that whole “teach a man to fish”.

If you have a problem is the first thing you do not go on to Google?

> Will a brand ever produce a reliable, mechanical car?

Take a look at Dacia.

Even Dacia has to follow the rules. Here in the EU a cell uplink (eCall) is mandatory since 2018 so even Dacia has about the same integrated radio/satnav/3g unit as every other Renault. Catalog price has also about doubled over the last 10 years too.

"Even if you drive your car every day, its battery will die and you won't be able to start it."

How do you figure this? Seems unlikely to me that a 15 min drive would not bring a car batter drained my occasional failed phone connections back to full

It's a well-documented phenomenon. You can Google it.

The thing drains your battery precipitously.

That still sounds suspect. Surely there are areas that don't have 3g coverage? For instance, underground parking garages or rural areas (especially in western US). These aren't common use cases, but common enough that they surely would have garnered press/media attention back when 3G was working. I vaguely remember a story about how a rental car was trapped in an underground garage because it couldn't get the unlock code. Why are we only hearing about it now? Moreover, surely those incidents would have caused them to fix the issue?

I own a Subaru Crosstrek from this same era, it definitely has the Starlink bits in the cockpit ceiling.

There must be more to this story. I've parked that car in an underground parking garage with absolutely no signal on any carrier for at least 2 weeks straight (likely more) with no issue. I've parked it for days at trailheads and wild camping spots around the High Rockies where there's also no signal.

I 100% believe that the Starlink module is poorly programmed, reliant on false assumptions of 3G always existing, and capable of draining the battery. But I would really love to know why that's never happened to me.

It could be that the modem can still see the phone tower, but its handshake to create a 3G connection fails, which it then retries indefinitely.

That would be different to having no signal whatsoever, where modem wakes up, listens for a tower, concludes there’s nothing there and goes back to sleep again.

That would explain why this issue is only cropping up now that 3G is being turned off, and how this scenario was missed by engineers. They assumed there was either a tower they could communicate with, or nothing. They didn’t account for a perfectly good tower being available, advertising itself as a recognisable network, but refusing to handshake.

I feel like this is it. I camped hundreds of days in areas with no service for miles and never had this problem. My battery ended up dying in my driveway in town.

I’m in the same boat, the one time I had issues with the battery on my 2020 crosstrek was when I accidentally left the dome light on for a week. Looking at the class action lawsuit seemingly related to this: https://www.subarubatterysettlement.com/ looks like the Crosstrek and Impreza aren’t included in the class.

Perhaps they are not impacted for some reason.

I can back this up. I regularly go car camping in the middle of nowhere in the Rockies. Absolutely no service for miles.

I hadn't gone camping for ~4-5 weeks, but still drove my car every day. Suddenly had the battery issue, pulled the plug, and the issue is gone.

I don't really know how to debug this. Something recently changed, but I don't know what. You'd think my battery would've died during all the time that I was out of service.

I did google it and while there are many stories about dead batteries all seem to talk about a problem after leaving the car parked for an extended period and no talk about a problem when driven daily.

This is also happening with my (old) Mercedes SUV.

Why are we shutting down 3G coverage?

Wireless spectrum is limited, and older protocols were less efficient so the same spectrum can handle more devices.

This made a massive improvement in places like subway stations, stadiums, airports, etc. where you have a ton of clients in a limited area, often all activating at the same time.

It would be interesting to know how much of the efficiency improvements were only in software – a lot of waste could be avoided if your old Subaru could speak enough 4G to be on the network even if it couldn’t do the higher data rates. Obviously improvements which are just due to clients completing their transfers wouldn’t be an option but I wonder whether any of the more efficient background activity, notification, back off, etc. logic could be backported.

Seems like a monumental waste to me.

Why do people continue to buy poorly-engineered vehicles?

-- Slick marketing.

Always pay attention to Consumers Report ratings.

Consumer report is not that good

Consumers Report still saves us money on reduces auto-repair bills for over 35 years.

No, IoT is here to stay. Makes way too much money.

Just pull the fuse and be done with it, not that hard.

That can have downstream effects, though. When I talked to them about doing it to my more recent Subaru, they told me I'd lose the front speakers and the in-car microphone as well, since all of them went through the same fuse.

Ah yea, make the consumer go through the pain of keeping that in mind and doing that every day and not blame the big corporation just silently breaking their product with seemingly no repercussions until now.

literally pop the hood and remove a fuse - https://youtube.com/shorts/y8xxT3bVf6I?feature=shared less than one minute. and now you aren't trackable. win/win.

Bingo... Possibly losing the in car microphone is another added bonus.

This should be the top comment for this post. I wish there was a way to do that. Or maybe @dang could help.

Not a big deal, you pull the fuse you don't do it everyday - you can also just unplug the module. Also not a big deal.

The solution to the problem is companies not doing such a thing and being made responsible.

Just because it’s easy for some people to do you can’t expect everyone to 1) know about this issue (Remember, there wasn’t an official announcement or recall) and 2) be confident enough to modify something on their car.

Most people never do anything on their care and just bring it to the service for every small thing.

Sod that for a game of soldiers

The SOS module in my car also has the microphone for hands-free calling and voice control.

This is called parasitic battery drain. There are two low tech solutions:

- install small solar panel behind back window. That will trickle charge battery, and will counter ballance drain

- hard switch off on battery, disconnect battery everytime car is not used for couple of days (weekend).

Another option is a device like a "PriorityStart!" shutoff, which I put in my old saturn after it developed a parasitic drain somehow and started killing batteries. In my case, it was easier to put this device in rather than start working through every connection to figure out what changed.

That said, for subaru it seems that pulling the DCM fuse seems to stop the drain: https://news.ycombinator.com/item?id=37973829

> hard switch off on battery

Doesn't this reset the emissions computer settings after some interval?

Not to mention any radio presets.

Yes the classic old anti-theft method plays hell with the increasingly computer-laden infrastructure of a modern car.

Am I wrong in thinking another potential low tech solution might be to just remove the cellular radio?

To be clear, I'm asking more than anything as I have very little car knowledge but to me that seems like a more preferable solution long term than frequently having to physically disconnect your car battery.

Yeah, probably a viable solution to just disconnect its cellular antenna (though it could easily be a PCB antenna, in which case... x-acto knife I guess)

How will that help the issue that the module keeps waking up and trying to send data until it succeeds?

The primary reason why it is an issue is because it causes the cellular antenna to constantly draw power from the battery.

So my thought was "Why not simply remove the component physically drawing all that power and just call it a day?"

Depending on what kind of protection they have in place, that may kill the electronics. You may not care if it’s a separate cellular device, but will if it’s part of the head unit.

Your best bet if you actually want to do this would be replacing the antenna with a properly matched dummy load.

But that (or removing the antenna at all) is going to just leave you with exactly the problem you already have—it can’t make a connection, so it will continue retrying (and probably at its highest power setting) forever and drain your battery.

>But that (or removing the antenna at all) is going to just leave you with exactly the problem you already have—it can’t make a connection, so it will continue retrying (and probably at its highest power setting) forever and drain your battery.

The majority of the power being drawn occurs because the system mistakenly believes the band is still usable and assumes that it probably just located somewhere with poor reception. This causes the system to ask the radio to bump up its radio power and when that doesn't work, wait x amount of time and then try again. However by stripping the cellular antenna, we make it impossible for the system from executing that behavior.

Sure it'd periodically wake up and inevitably fail to initiate a scan but that consume vastly less power. (to the point that it'd likely be a non-issue)

Easier solution: get a 120V trickle charger and plug it in when your car is parked. Batteries love being constantly topped up anyway. NOCO is a good brand I use for my stored vehicles with AGM batteries.

Doesn't solve the problem for anyone who doesn't park in a private garage.

This is because large corporations must provide infinite growth to their investors, or a good number of them will dump the stock and raising investment might become very hard.

For the sake of infinite growth, many lines have been crossed and many decisions that make no sense if not looked through the lens of money have been made.

I think companies will start doing truly unethical things in the pursuit of this in the future. What if the dashboard could show ads before the car starts? It’s just a second of your time. What if a gas station network like Circle K could sponsor your car so it tells you to get a specific type of gasoline? Why not subscription for “optional” safety features that were standard? Why not a way to remotely disable cars with a warning light on the dash “for your own safety” until they visit a certified dealer? Why not coded car parts so the customer gets a “genuine product” and doesn’t “accidentally” install a cheaper generic counterpart? Why not pay as you go features instead of paying upfront? Why not track places you drive to with your car, the conversations you have in it, the people you call, and whether you drive with your wife and/or kids for behavioural advertising elsewhere? What about an eco mode subscription that saves you money when compared to not buying it, despite that it is always free with cars today?

Maybe we could also disable some people’s cars to align with certain political demographics and identities, like we cancel, deplatform, and get people fired for slights online now? Could that paint us in a positive light in certain demographics? One client lost for a thousand gained!

Maybe an auto-911 crash detection subscription? You wouldn’t want to be the mother whose kid dies in a car crash so you’d save $19.99 a month, would you? Anything if it moves more cars or at a higher marginal revenue, right?

And let’s not forget marginal costs - what could be cheaper? Maybe some parts are too durable. A lot of things could be digital - break lines, steering, the accelerator, door locks, seat belt mechanisms, mirrors and so on. And we could save doubly by outsourcing all this work to the cheapest bidder, whether they specialize in automotive safety or web design.

Anything for the sake of growth. And it’s coming. Sooner or later, there won’t be a company in the auto industry providing appealing returns unless they charge their users more, exploit them more, or save up for a nicer dividend by cutting some corners. Not today, not this year, maybe not in the immediate future. But this is the natural conclusion of the flavor of capitalism we do.

I don’t blame capitalism for everything. It’s good for a lot of things. But it totally fails humanity when large corporations are involved. We should change it before a bunch of our industries collapse under their self-destructive behaviors driven by greed. Because right now, a good investment strategy is to constantly keep investing in companies who irreparably harm themselves and their customers long-term for a quick buck. You just need to build your portfolio of these and dump them once their sell-out is over. And this makes money. An investor would be unwise not to pick a money-making strategy.

It is in everyone’s best capitalistic interest to exploit others to the max. And when all that could be exploited is, well… no one managing companies thinks about this. The investors want returns now, all the time. The executives know how to sell-out for this and golden-parachute into the next company to do it all again. The employees are laid off and customers are mistreated, but hey - capitalism is to acquire capital. It’s not called moralism.

>I think companies will start doing truly unethical things in the pursuit of this in the future. What if the dashboard could show ads before the car starts? It’s just a second of your time. What if a gas station network like Circle K could sponsor your car so it tells you to get a specific type of gasoline? Why not subscription for “optional” safety features that were standard? Why not a way to remotely disable cars with a warning light on the dash “for your own safety” until they visit a certified dealer? Why not coded car parts so the customer gets a “genuine product” and doesn’t “accidentally” install a cheaper generic counterpart? Why not pay as you go features instead of paying upfront? Why not track places you drive to with your car, the conversations you have in it, the people you call, and whether you drive with your wife and/or kids for behavioural advertising elsewhere? What about an eco mode subscription that saves you money when compared to not buying it, despite that it is always free with cars today?

>Maybe we could also disable some people’s cars to align with certain political demographics and identities, like we cancel, deplatform, and get people fired for slights online now? Could that paint us in a positive light in certain demographics? One client lost for a thousand gained!

All them can be theoretically done today. If companies are really chasing after "infinite growth", why are they leaving money on the table by not implementing those things?

The answer is obvious. There's nothing stopping them from making things arbitrarily shitty, but they don't because consumers will hate it and switch to a competitor that doesn't have those things.

>Maybe an auto-911 crash detection subscription? You wouldn’t want to be the mother whose kid dies in a car crash so you’d save $19.99 a month, would you? Anything if it moves more cars or at a higher marginal revenue, right?

Isn't this already a thing today? I think GM pioneered it with OnStar. I don't think there's even a car that does auto-911 today for free.

It is required in the EU. The mobile device only connects to the network in case of a crash. Of course, manufactures having to add a mobile device prefer to keep it connected continuously to use it for other things than required.

LTE-M/NB-IoT will solve this once it's activated on cell towers. Much lower consumption, it isn't paging the network at all times. They just have to shut it off when the ignition is off.


> I don't think there's even a car that does auto-911 today for free.

Ford! The SYNC system has had 911 Assist since the beginning of SYNC in the late 2000s. It uses your paired Bluetooth phone to call 911 in the event of an accident and plays an audio announcement to 911 based on vehicle information, then allows them to open the mic to the cabin.


> why are they leaving money on the table by not implementing those things?

> The answer is obvious. There's nothing stopping them from making things arbitrarily shitty

Two things to say about this: one, the laws of physics are stopping them, because developer time is limited. Two - there may be better ways to deliver growth right now, but over time there will be fewer such options.

> Isn't this already a thing today? I think GM pioneered it with OnStar. I don't think there's even a car that does auto-911 today for free.

So I guess then consumers aren't really stopping them.

> they don't because consumers will hate it and switch to a competitor that doesn't have those things.

Investors are not emotionally attached to a company, they will switch over to the ones delivering financial results this quarter. If that comes with consumer relations being irreparably damaged, does it concern an investor? They'll just hop to another company in the industry next. I said that this is an industry problem, and I meant it. It's not just one car manufacturer doing this, it's all being pushed towards doing it by the same force.

I think a lot of people have never traded even stocks in their lives and they don't see this part of it. Market-dominating companies about to enshittify their products are cash cows. And if they enshittify so much that they fall off the market, in a healthy market with many companies such as automotive, there will always be a company forced to enshittify to prop up its stock. It's not even just about sourcing investment - the net worth of the c-suite if often tied to the stock and they aren't looking to cash out when it's down.

Regardless, the overall effect is slowly making the whole industry hostile to the consumer. This is like private health insurance or life insurance - go ahead, switch to one that doesn't charge you a premium for your preexisting condition. And get your employer to switch. Not so easy when the entire market is guided by the same incentives that work against you, is it?

Gas station tried adverts near me. Oddly, though, there was a live image of me looking at the screen = camera behind. They have since removed the ads, and the camera doesn't show: they still get my middle finger and no face (i am slightly off to the side) for my every fill up.

Until car companies are sued to the tunes of hundred of millions for something like this, their attitude will never change. Sigh.

Subaru did settle a class action lawsuit:


The settlement is pathetic. I could never prove the parasitic battery drain on my 2017 forester, so I just had the OEM audio/starlink system replaced with a cheap Sony CarPlay head unit and the problem went away.

But I cannot get reimbursed since I have no proof, even though surely every car manufactured with those components has a parasitic battery drain.

Are you sure giving even more money and power to trial lawyers is going to solve anything? There are already these people we already pay taxes for in DC at the DOT who have the power to require a recall.

Call your representative, tell them what's going on, and ask them what the DOT is going to do about it. It'll take 30m of your time.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact