I recently read that another Toyoto-owned brand Suzuki is selling rusty cars in Turkey. It is sad to see a brand like Toyota is associated with a lot of misdoings.
Japan has a strong culture of "don't rock the boat, be sure, very very sure if you accuse someone", and thus accusations are already sentences practically. Couple this with any kind of loyalty, and you immediately see that speaking up will get folks excommunicated (ronins, woah!)
From reading through some of the reports linked in the CNN article, it sounds like in a lot of cases the test employee was taking shortcuts such as reporting results from the left side of the vehicle as results from the right side of the vehicle under the belief that there was no substantive difference, or not properly adjusting the tire pressure before the test and reporting that it had been adjusted.
It still illustrates that this sort of stuff was either tolerated or not detected, but in all honestly it sounds (to me) like a lot of it was pretty minor stuff. It's not like the car exploded in a raging fireball on the test stand and the reported result was "A++ would buy" like Ford did with the Pinto.
"tests on the airbags of Daihatsu Move/Subaru Stella, Daihatsu Cast/Toyota Pixis Joy, Daihatsu Gran Max/Toyota Town Ace/Mazda Bongo revealed the irregularity that the “airbags deployment computer (ECU)” was not the same as that of mass-produced products."
But what does that mean? Was it a prototype because the production model wasn't ready yet? Was it hand-built as opposed to "mass produced" but presumed and intended to be functionally identical, or was it intentionally different for the purpose of passing the test, knowing that the actual computer would not?
They built ceritifcation specials to pass tests. Which means the devices under test were not representative of mass produced examples.
Specifically they flashed some custom ROM on a airbag computer so it auto-deploys at scheduled T+ time from the beginning of a test run. That's kind of a huge deal because we wouldn't know if airbag is going to deploy in real crashes.
Other tampering includes ignoring requirement that crashes must also unlock front doors, reporting numbers from a note hidden in palm while the test is running instead of from gauges, so on and so forth.
> The question is how much control Toyota had over operational decisions. I'd say not too much, even with Daihatsu being a subsidiary of Toyota.
It could go either way, IMO. It didn't take long for folks to the Toyotafication of Subaru's as Toyota's ownership increased. In general it doesn't take a lot of ownership for a carmaker to start getting some pressure.
Daihatsu CEO right now is ex Toyota regional CTO for APAC, so there's that.
One of contributing factors, as being told in news narratives, is Daihatsu relied on shortened time to market and therefore reduced engineering man-hours as a cost saving measure, and there were barely enough time for scheduled tests, let alone for delaying ones affected by known issues, which had incentivized test engineers to just rig the test and/or forge documents.
Toyota is also known for persistently demanding suppliers to contribute towards $20-30MM yearly cost saving target by reducing price by 1% every half years(~100MM and 5.85% every 3 years). That surely had contributed, and frankly that itself is a culture issue.
Yup, I would like to see a similar third party audit performed at Toyota, too. Toyota and Daihatsu cars share a lot of the same technology and engineering resources, similar to the relationship between Lexus and Toyota. I can’t help but wonder if they also share a similar approach to safety testing.
It’s clear that the problems started long before Toyota had whole or even majority ownership.
Even from dealing with acquisitions at much smaller scales in the tech industry, I was surprised at how frequently we encountered small companies bending the rules or outright flouting the law. Discovering blatant legal or safety violations would have ended the acquisition talks, but these problems are often so deeply embedded in the company that you simply can’t expose everything during due diligence. These problems may have been baked into processes and common knowledge among a small group of people who were more interested in making their own jobs easier and maintaining their position of power than in adhering to the law. They might have had an “everyone does it” attitude after coming from other companies they also broke the law. Once it starts and they don’t have any consequences for a couple years, it becomes the de facto standard in the company. And because it doesn’t show up on a balance sheet or in any of the auditable documentation, it takes a very long time for someone to raise the issue and possibly risk their job in the process.
At one point we talked to some people who had been breaking a regulatory law for years and they genuinely had no idea. They had just been doing things the way they learned at another company and thought it was all above board.
smaller companies often will do 1 of 2 things, both for the same reason.
1. They'll over-adhere to regulatory compliance/security
2. They'll under-adhere to regulatory compliance/security.
It's often done for the same reason: ignorance. It's a question of culture/personality, do they play scared or fast and loose?
often times smaller companies don't have the money or the time to develop the expertise they need in the regulations, hence why they make guesses and over or under adjust.
One of the reasons I dislike "security" people so much is that they often turn into a combination of
1. no real idea what they're doing
2. a tendency to over-adhere, and
3. a belief that security automatically always trumps other concerns.
With the benefits of ownership come everything else. If Toyota didn't know about this issue, that's a major failure of Toyota management. If they did know, it's more of a failure.
unrelated, but can every news site please build a lite version? This opened so fast and was so easy to read. I forgot how nice websites could be with HTML plus some mild CSS.
Oh that fair. I didn't realize they made one now. When I last looked they just had hybrids. I was looking for a car not truck though, so I may have glossed over it then too.
This 2 1/2 year old article has a strange title, compared to content.
Apparently Toyota was lobbying for hybrids and h2(which are 100% electric) to be included in incentives.
That's hardly anti-electric, and there are many, many people that need more range. Hybrids and H2 are answers to that.
An example? Most days a hybrid would get me around just fine, using only electric. Other times I have to drive well beyond electric range, and sometimes for days.
It doesn't matter what "but, you could.." statements someone makes, that's what I need. Hands down. And Toyota provides all options.
This is a good thing. And driving on electric is a win, hybrid or not.
I don’t think anyone cares for the reasons you state.
In Japan they produce mostly kei cars, these are 3 cylinder tin cans on wheels. No one even assumes they’re “safe”. The brakes work fine and that’s about it.
This is a bit of a weasel-ish, click-bait title. Calling it Daihatsu would not generate the same outrage, so they try to virtually inflate the scale of the scandal to a major car manufacturer.
Maybe Toyota has their hands dirty, I don't know, and I'd love to know. As it is, that's bordering a little on the wrong side of journalistic integrity.
The report here (https://www.daihatsu.com/news/2023/report_1_E.pdf) lists Toyota Yaris, Toyota RAIZE HEV, Toyota Agya, Toyota Pixis Joy, Toyota Pixis Truck, Toyota Roomy, and Toyota Tank all affected.
> The scandal is another blow to the automaker, which had admitted in April to violating standards on crash tests on more than 88,000 cars, mostly sold under the Toyota brand in countries such as Malaysia and Thailand.
Seems fair to mention Toyota in the headline, since that was the brand name people thought they were buying when they got one of these cars.
I thought it was clickbait too, until I realized it was Daihatsu, which is all but unknown in the US where CNN is based. But then I realized that "Toyota-owned automaker Daihatsu tampered..." would work and now I think it's clickbait again.
I recently ordered a new car from Daihatsu, and they called and asked if I still want the car or would like to cancel the order in light of this news.
I’m honestly not that concerned about the safety of the vehicle. They’ve been cheating on safety tests for a long time apparently, and in the real world I don’t think Daihatsu cars have fared much worse than other manufacturers in crashes. It’s also got a lot more safety tech than the car it is replacing, so I think it would be a step up in safety regardless.
Instead I’m more just concerned with the future of the company, whether the resale value of the car will drop, etc. I also do find the scale of this fraud pretty shocking, and don’t want to reward this behavior by purchasing from them.
So probably I will cancel and buy a Honda or Suzuki instead.
> in the real world I don’t think Daihatsu cars have fared much worse than other manufacturers in crashes
It's almost impossible to know that without data, given the frequency of crashes involving Daihatsu cars (at least in the US), and the visibility of crashes generally - we don't know about 99.x% of them.
Daihatsu is the market leader for kei car sales in Japan (where I live), with 560,000 cars sold in 2022. The model that I ordered is frequently the top-selling car model in the country in a given month.
They are involved in accidents all the time. If there were issues with e.g. the airbag failing to deploy properly in real-world crashes (one of the areas where they cheated), I believe that would have surfaced by now.
Not necessarily. I'm capable of cooking a meal, but I've never worked in a restaurant nor taken any food health and safety courses or anything.
If tomorrow I started selling my cooking while displaying a fake certificate about my food safety qualifications it wouldn't automatically mean that anyone who eats my cooking will be hurt, it just means I took the lazy way (which could cause hurt, if by not knowing some food safety info I fuck up somehow, but it doesn't automatically mean the food I make must be unhealthy).
But if you serve enough food to enough people, you’ll probably get bit by one of those “this rule is written with blood” fun little gotchas and kill someone.
Will you properly handle celiac or severe allergies?
But any place can get bitten by one of those gotchas. In the example, maybe they are a perfectly safe cook by instincts without formal certification. Then nothing would ever be detected, or it would be on par with those who did pass.
It depends on the variance of their production line. Indeed if the variance is low enough safety testing is a waste of time and money. Just test a handful of random cars off the line and call it good. Of course that sounds shocking, but it is true.
I would also check insurance if those rates aren't going to skyrocket.
And what about governmental regulation? Knowingly driving a car that doesn't meet the safety regulation, in my opinion, should be banned.
Just like people who put a quick-release steering wheel in their car and then get rear-ended. Sure they didn't make a mistake, but because of their own choice they exacerbated the problem.
From light googling, a quick-release steering wheel seems to be a steering wheel that you can simply remove by pressing a button.
So from your post, the implication seems to be that if someone has one, and their car is rear-ended, the steering wheel just comes off into their hands and they can't control their car.
Is that right?
Why would someone purchase a quick-release steering wheel? What does it help them accomplish?
> Why would someone purchase a quick-release steering wheel? What does it help them accomplish?
The “actual” answer is that quick release wheels makes getting in/out of race cars much easier. In some of them it’s nearly impossible to get your knees past the wheel if it’s not removable.
Generally though, it’s just for looks and to feel like a real racer with your momo wheel.
No quick-disconnect steering wheels don't have airbags.
And it makes me wonder, even though I'm at fault for rear-ending someone, should my insurance (or me) have to pay out MORE because the person in front of me removed mandatory safety equipment?
It's easy to copy and paste: "When a measure becomes a target, it ceases to be a good measure"
In application Goodhart's Law means that you must also use another measure that more closely matches the desired result.
In this instance, that means that we should also be looking for injuries per X. Where X may be vehicles sold, distance travelled per vehicle, or per person-vehicle, per time, or some other per capita number. As far as I am aware, numbers like this are not easily available or comparable.
Per accident, normalized over velocity and angle of impact is ideal I'd assume in order to not bias on the kind of driving done in vehicles of a given brand.
I don't think Goodhart's law applies here. They weren't really "cheating" to get better metrics, they just seemed to be taking shortcuts out of laziness.
I think it is hard to find a balance between enforcement, realism and stringency.
If you make enforcement hard and standards stringent, with no realism, you will get cars that are optimized to pass a test that doesn't help in practice.
If you make enforcement hard and tests that are realistic but too easy, you will get an improvement, but maybe not as good as it may be.
And if you loosen up on enforcement and make test that are both realistic and stringent, manufacturers will cheat.
If you make all three hard, you just shift the problem, manufacturers will either optimize for tests, do not enough, or cheat depending on what is the balance between the three.
You make a good point. How to induce people to do the right thing? The carrot? The stick? Having a checklist? Inculcating a safety culture? All of the above? If you are cheating, then typically you know you are cheating. I am going to assume these people knew, even if Toyota at large did not.
> Daihatsu has confirmed that in the side collision tests of the [Toyota Yaris, Perodua Axia, and Toyota Agya], the inside lining of the front seat door was improperly modified, and that there was a violation of the side collision test procedures and methods stipulated by regulations.
Inside lining of the front seat door? That doesn't sound very significant at all.
It assumes the regulations are sensible which is not always the case. Regulations can cause inordinate costs increases relative to their actual value. The fact that the regulations may be well intended is orthogonal. There is no reason to cheat if compliance is trivial.
Here's the longer description of the same thing. It sure sounds more serious.
> When applying for certification of side collision tests for vehicles developed by Daihatsu for overseas markets (four models), the person in charge of conducting the tests, concerned that it would be problematic if the development and sales schedule was not met due to the failure of the certification tests, made manual modifications on the certified prototype vehicles that differed from those of mass-produced vehicles, such as cutting the back of the plastic front door trims to prevent sharp edges in the event of a collision, in order to ensure that the vehicles
would pass the certification tests.
This is the kind of thing I worry about since finding out the IIHS allow automakers to do their own testing and get certification by providing the data.
You can punish car companies by buying used. It's much more sensible on your pocketbook too. And if your used vehicle's old enough, the automaker won't track your driving, in-car conversations, cell phone use, etc. Really the only possible downside is reliability but a full rebuild for engine and tranny is probably, at most, a sixth the cost of a new car.
One example that made the rounds recently after a Consumer Rwports…report was Honda’s privacy policy which explicitly states they can do whatever they want with the contents of your text messages, including analyzing them a d selling them.
Intel tampered with information security for 30 years?
Yes, they (or at least some of them) had known that speculative execution causes security problems.
But no, they have not been in business that long with technology that could be affected.
And more importantly the whole industry is too young to have any security regulations. Very much like the car industry in the 1930s had nothing comparable to today's safety regulations.
94 comments
[ 3.2 ms ] story [ 165 ms ] threadOr did people speak up, but the pattern wasn't stopped (until it had been going on for 30 years)?
https://www.youtube.com/watch?v=IRn4xzaugbk
So, probably they kept quiet on the first failed test, made it go away to save face, and then there was no going back.
> So, probably they kept quiet on the first failed test, made it go away to save face, and then there was no going back.
That is an essential lesson of integrity: Lie now and you may end up telling the same lie every day going forward.
I'm basing it on the similar Kobe Steel aluminum and copper scandal and report.
https://www.bbc.com/news/business-43298649
It still illustrates that this sort of stuff was either tolerated or not detected, but in all honestly it sounds (to me) like a lot of it was pretty minor stuff. It's not like the car exploded in a raging fireball on the test stand and the reported result was "A++ would buy" like Ford did with the Pinto.
https://www.daihatsu.com/news/2023/20231220-4.html
Specifically they flashed some custom ROM on a airbag computer so it auto-deploys at scheduled T+ time from the beginning of a test run. That's kind of a huge deal because we wouldn't know if airbag is going to deploy in real crashes.
Other tampering includes ignoring requirement that crashes must also unlock front doors, reporting numbers from a note hidden in palm while the test is running instead of from gauges, so on and so forth.
> [Daihatsu] has been a wholly-owned subsidiary of the Toyota Motor Corporation since August 2016.
Did a good company accidentally acquire a company with a very serious secret problem?
Either way, Daihatsu cheating and lying to pass safety tests for 30 years is bad, and hints at a deep seated culture issue.
It could go either way, IMO. It didn't take long for folks to the Toyotafication of Subaru's as Toyota's ownership increased. In general it doesn't take a lot of ownership for a carmaker to start getting some pressure.
One of contributing factors, as being told in news narratives, is Daihatsu relied on shortened time to market and therefore reduced engineering man-hours as a cost saving measure, and there were barely enough time for scheduled tests, let alone for delaying ones affected by known issues, which had incentivized test engineers to just rig the test and/or forge documents.
Toyota is also known for persistently demanding suppliers to contribute towards $20-30MM yearly cost saving target by reducing price by 1% every half years(~100MM and 5.85% every 3 years). That surely had contributed, and frankly that itself is a culture issue.
Even from dealing with acquisitions at much smaller scales in the tech industry, I was surprised at how frequently we encountered small companies bending the rules or outright flouting the law. Discovering blatant legal or safety violations would have ended the acquisition talks, but these problems are often so deeply embedded in the company that you simply can’t expose everything during due diligence. These problems may have been baked into processes and common knowledge among a small group of people who were more interested in making their own jobs easier and maintaining their position of power than in adhering to the law. They might have had an “everyone does it” attitude after coming from other companies they also broke the law. Once it starts and they don’t have any consequences for a couple years, it becomes the de facto standard in the company. And because it doesn’t show up on a balance sheet or in any of the auditable documentation, it takes a very long time for someone to raise the issue and possibly risk their job in the process.
At one point we talked to some people who had been breaking a regulatory law for years and they genuinely had no idea. They had just been doing things the way they learned at another company and thought it was all above board.
1. They'll over-adhere to regulatory compliance/security
2. They'll under-adhere to regulatory compliance/security.
It's often done for the same reason: ignorance. It's a question of culture/personality, do they play scared or fast and loose?
often times smaller companies don't have the money or the time to develop the expertise they need in the regulations, hence why they make guesses and over or under adjust.
One of the reasons I dislike "security" people so much is that they often turn into a combination of
1. no real idea what they're doing
2. a tendency to over-adhere, and
3. a belief that security automatically always trumps other concerns.
Toyota's Daihatsu to halt all vehicle shipments as safety scandal widens - https://news.ycombinator.com/item?id=38756044 - Dec 2023 (61 comments)
This, along with their resistance to EVs is going to bite them.
https://www.toyota.com/bz4x/
Are they not sold where you are?
I know of people who had to wait more than a year.
Apparently Toyota was lobbying for hybrids and h2(which are 100% electric) to be included in incentives.
That's hardly anti-electric, and there are many, many people that need more range. Hybrids and H2 are answers to that.
An example? Most days a hybrid would get me around just fine, using only electric. Other times I have to drive well beyond electric range, and sometimes for days.
It doesn't matter what "but, you could.." statements someone makes, that's what I need. Hands down. And Toyota provides all options.
This is a good thing. And driving on electric is a win, hybrid or not.
And so this Verge article is quite odd.
After all, I'm not sure those who are looking for the safest car would put Toyota near the top of their lists anyway. Volvo comes to mind instead.
In Japan they produce mostly kei cars, these are 3 cylinder tin cans on wheels. No one even assumes they’re “safe”. The brakes work fine and that’s about it.
Once we needed a safer car we sold it.
People love to glamorize and empower corruption, for some reason. That is something people care about and act on, in all cultures, at all times.
All of their model names (for cars and vans sold in Japan) look like English speling misteaks. It was absolutely hilarious.
I can't tell if this is intentional. Is "speling misteaks" a meme, like people writing "no ragrets"?
1: https://en.wikipedia.org/wiki/List_of_Toyota_vehicles
This is a bit of a weasel-ish, click-bait title. Calling it Daihatsu would not generate the same outrage, so they try to virtually inflate the scale of the scandal to a major car manufacturer.
Maybe Toyota has their hands dirty, I don't know, and I'd love to know. As it is, that's bordering a little on the wrong side of journalistic integrity.
Seems fair to mention Toyota in the headline, since that was the brand name people thought they were buying when they got one of these cars.
I think this is probably because most news readers don't know the name Daihatsu, not because they're trying to inflate the scandal.
Seems plenty fair to me.
You'd have more of a point if the acquisition had happened this year, but Toyota took majority ownership in 1998.
I’m honestly not that concerned about the safety of the vehicle. They’ve been cheating on safety tests for a long time apparently, and in the real world I don’t think Daihatsu cars have fared much worse than other manufacturers in crashes. It’s also got a lot more safety tech than the car it is replacing, so I think it would be a step up in safety regardless.
Instead I’m more just concerned with the future of the company, whether the resale value of the car will drop, etc. I also do find the scale of this fraud pretty shocking, and don’t want to reward this behavior by purchasing from them.
So probably I will cancel and buy a Honda or Suzuki instead.
It's almost impossible to know that without data, given the frequency of crashes involving Daihatsu cars (at least in the US), and the visibility of crashes generally - we don't know about 99.x% of them.
They are involved in accidents all the time. If there were issues with e.g. the airbag failing to deploy properly in real-world crashes (one of the areas where they cheated), I believe that would have surfaced by now.
The statistics don't matter if your child dies because your vehicle is one of those that wouldn't have passed the safety tests.
If they're fudging to pass safety tests, then there's most definitely people whose lives have been affected by it.
If tomorrow I started selling my cooking while displaying a fake certificate about my food safety qualifications it wouldn't automatically mean that anyone who eats my cooking will be hurt, it just means I took the lazy way (which could cause hurt, if by not knowing some food safety info I fuck up somehow, but it doesn't automatically mean the food I make must be unhealthy).
Will you properly handle celiac or severe allergies?
the chances being statistically low don't help the person who was hospitalized due to a peanut allergy.
And what about governmental regulation? Knowingly driving a car that doesn't meet the safety regulation, in my opinion, should be banned.
Just like people who put a quick-release steering wheel in their car and then get rear-ended. Sure they didn't make a mistake, but because of their own choice they exacerbated the problem.
From light googling, a quick-release steering wheel seems to be a steering wheel that you can simply remove by pressing a button.
So from your post, the implication seems to be that if someone has one, and their car is rear-ended, the steering wheel just comes off into their hands and they can't control their car.
Is that right?
Why would someone purchase a quick-release steering wheel? What does it help them accomplish?
The “actual” answer is that quick release wheels makes getting in/out of race cars much easier. In some of them it’s nearly impossible to get your knees past the wheel if it’s not removable.
Generally though, it’s just for looks and to feel like a real racer with your momo wheel.
And it makes me wonder, even though I'm at fault for rear-ending someone, should my insurance (or me) have to pay out MORE because the person in front of me removed mandatory safety equipment?
The emission cheaters are still years after being caught trying not to take their responsibilities against all end customers they caused damage.
1995: Toyota ownership increased from 16.8% -> 33.4%
1998: 51.2%
2016: 100%
https://en.wikipedia.org/wiki/Daihatsu#:~:text=Daihatsu%20wa....
It's easy to copy and paste: "When a measure becomes a target, it ceases to be a good measure"
In application Goodhart's Law means that you must also use another measure that more closely matches the desired result.
In this instance, that means that we should also be looking for injuries per X. Where X may be vehicles sold, distance travelled per vehicle, or per person-vehicle, per time, or some other per capita number. As far as I am aware, numbers like this are not easily available or comparable.
Not because cheating exists and they are caught, but because it lasted so long before being caught.
If you make enforcement hard and standards stringent, with no realism, you will get cars that are optimized to pass a test that doesn't help in practice.
If you make enforcement hard and tests that are realistic but too easy, you will get an improvement, but maybe not as good as it may be.
And if you loosen up on enforcement and make test that are both realistic and stringent, manufacturers will cheat.
If you make all three hard, you just shift the problem, manufacturers will either optimize for tests, do not enough, or cheat depending on what is the balance between the three.
> Daihatsu has confirmed that in the side collision tests of the [Toyota Yaris, Perodua Axia, and Toyota Agya], the inside lining of the front seat door was improperly modified, and that there was a violation of the side collision test procedures and methods stipulated by regulations.
Inside lining of the front seat door? That doesn't sound very significant at all.
These are almost always significant things, at the very least worth looking into.
https://www.daihatsu.com/news/2023/report_1_E.pdf
Here's the longer description of the same thing. It sure sounds more serious.
> When applying for certification of side collision tests for vehicles developed by Daihatsu for overseas markets (four models), the person in charge of conducting the tests, concerned that it would be problematic if the development and sales schedule was not met due to the failure of the certification tests, made manual modifications on the certified prototype vehicles that differed from those of mass-produced vehicles, such as cutting the back of the plastic front door trims to prevent sharp edges in the event of a collision, in order to ensure that the vehicles would pass the certification tests.
https://www.iihs.org/ratings/about-our-tests#verification
IIHS is non-government but they have more rigorous standards than the government and they’re what I would rely on evaluating a vehicle.
At this point you should assume anything with a mic is spying on you, because you're right. We long since left behind trust.
Yes, they (or at least some of them) had known that speculative execution causes security problems.
But no, they have not been in business that long with technology that could be affected.
And more importantly the whole industry is too young to have any security regulations. Very much like the car industry in the 1930s had nothing comparable to today's safety regulations.