Hacker News Reader

Comment by jacobush | Hacker News Reader

Weirdest thing, a colleague had unintended acceleration in a Toyota 4WD of much older vintage, 80s or early 90s. Must have been something different entirely, those things were much more mechanical in nature.

View on HN

Most cases of unintended acceleration, in all brands of cars, are due to the foot coming down on the wrong pedal. And then in a panic, your brain refuses to acknowledge the error and doubles down. Your memory gets based on what your brain thought, and not what actually happened.

US government estimates are that this happens an average of around 16k times per year in the USA. The press release that Wikipedia cites for that has disappeared but you can find it on the wayback machine: https://web.archive.org/web/20180423205652/https://www.nhtsa...

Toyota's acknowledged problem was a floor mat that could get jammed and cause the accelerator to stick. However most of their cases of unintended acceleration were still likely to be human error.

I'm sorry but you are wrong regarding the Toyota problem.

There was (is?) a firmware issue that caused the vehicle to attempt to speed up, regardless of if the accelerator was pressed, or even if the brake was pressed.

In my case, I had the brake fully depressed with around 100kg of force, and the engine was revving to attempt to overcome the brake (VERY scary situation, believe me!). The system only recovered when I bumped a taxi in front of me.

Now you may say I was pushing the accelerator. If I did that I would have hit the car in front of me a lot faster than 5kph. If I was somehow pushing both the accelerator and the brake, the computer system should ignore the acceleration anyway (due to brake taking precedence).

But you don't have to take my word for it. Go read up on the Toyota firmware issue, starting with link I put in the prior post.

EDIT: here's another article, summarizing the faults as determined by court: https://www.edn.com/design/automotive/4423428/Toyota-s-kille...

- Toyota’s electronic throttle control system (ETCS) source code is of unreasonable quality.

- Toyota’s source code is defective and contains bugs, including bugs that can cause unintended acceleration (UA).

- Code-quality metrics predict presence of additional bugs.

- Toyota’s fail safes are defective and inadequate (referring to them as a “house of cards” safety architecture).

- Misbehaviors of Toyota’s ETCS are a cause of UA.

maybe you were pressing both

From what I understand, the computer will (and should) reject the acceleration attempt in favor of the brake.

But, if you own a reasonably modern vehicle, try it :)

If this were true people wouldn't be able to do burnouts in auto transmissions cars?

It's actually fairly common, try searching for "brake override".

That's not how it works. The brakes are more powerful than the engine, sure, but only at full braking power. Ask anyone who owns a rear-wheel-drive sports car what happens when you push both pedals at the same time and they will literally leave you in a cloud of smoke.

The RWD thing is definitely important here. Given that the majority of cars and a significant number of SUVs are now FWD, this doesn't hold water. Given it is the front brakes that are more powerful than the rears, and if your brakes are maybe a tad worn out, you might not be able to take your brakes from a dynamic friction situation to a static friction situation. In the mean time, your engine would just continue injecting heat into the brakes, which could easily overcome them. It might also be that he was not braking with the full force initially, meaning some degree of heat saturation may have already been present.

I could rev my engine while pressing the brake on a 2014 Honda Civic (helped with not going backwards when I needed to start moving from a standstill on a steep hill). Is that modern enough for you?

(On the other hand my 2017 Subaru Forester does not do the same; thankfully it has less of a tendency to roll backwards when releasing the brake on a steep hill.)

First, my main point remains. Most cases of unintended acceleration are human error, regardless of whether any specific case is. I am sorry that you experienced the contrary. But it doesn't change the overall figures.

Second, Toyota dragged its feet on the full variety of problems that they had. But the one that they first acknowledged was the floor mat. As https://abcnews.go.com/Blotter/toyota-pay-12b-hiding-deadly-... says, At the time of the first ABC News report, Toyota attempted to assure its drivers that the incidents of sudden acceleration without warning were solely caused by floor mats becoming stuck on the gas pedals or driver error.

>I had the brake fully depressed with around 100kg of force, and the engine was revving to attempt to overcome the brake

No, it wasn't. An engine's speed (RPM / "revs") is directly proportional to the drivetrain speed -- the RPMs literally cannot increase unless you are 1) accelerating (actual drivetrain speed going up), 2) shift to a lower gear then slip torque converter to rev match (I assume you're driving automatic), or 3) disengaged from the drivetrain (in neutral gear) and hitting the gas.

There's no way for an engine to be "revving" to overcome braking. Hop in a manual car, get up to a normal speed in any gear, then try hitting both the brake and gas.

I'm not a car guy, so maybe I am getting some technical terminology wrong.

But Toyota's software did cause acceleration even without pushing the accelerator, and even while pressing the brake. They were legally proven to be at fault for that.

You can speculate all you want if I was being an idiot and mis-remembering what I personally did, but these facts still stand regardless of my personal experience:

1) Toyota was legally at fault for software caused unintended acceleration in the USA

2) Toyota Thailand never issued a recall for similar makes/years

Now maybe the software of cars sold in Thailand is totally different from the software of cars sold in the USA, but the cynic in me thinks it's just that Thailand's lack of consumer protections, plus Toyota's indifference, is the real reason no action was taken.

You are mistaken. In a car with an automatic transmission, the engine RPM is related to the drivetrain speed, but not directly proportional. The input and output shafts of the torque converter always spin at different speeds during acceleration.

You can try a similar experiment to the one you proposed in a car with an automatic transmission. Stand firmly on the brake (with the parking brake engaged and nothing in front of the car, for safety), put the car in drive, and give it some gas with your other foot. You can easily get the engine to more than twice the idle speed without the wheels turning at all! (Don't do this for extended periods. It makes the torque converter get hot.)

From a standstill yes -- the car is slipping the torque converter between disengaged (neutral, idle in a standstill) and first (to actually get going). Exactly the same as a manual, to get moving from a standstill you have to slip the clutch somewhat, and yes your RPMs will go up.

But that doesn't apply when the car is already in gear and in motion. In both an auto and a manual, if you're applying enough brake to prevent the car from accelerating, the engine will never be "revving" to fight your braking. It will only increase speed if your drivetrain speed increases.

I didn't explicitly state it in the above post because I don't want to crap on the parent too much, but the if he was hitting a pedal and the engine was "revving"... he was probably hitting the gas and accelerating.

> he was probably hitting the gas and accelerating.

...Or the car was actuating the throttle as if he was hitting the gas, a.k.a. unintended acceleration.

I see a lot of people like you in this thread and elsewhere dismissing unintended acceleration experiences.

Before you continue to do so, I would urge you to read the NASA report on the Toyota unintended acceleration case and related documents (https://www.nhtsa.gov/staticfiles/nvs/pdf/NASA-UA_report.pdf, https://users.ece.cmu.edu/~koopman/pubs/koopman14_toyota_ua_..., and the link in the top level comment, https://www.edn.com/design/automotive/4423428/Toyota-s-kille...). It lays out a damning indictment of Toyota's ECU hardware and firmware development culture, that flies directly counter to the happy picture promoted by the "Toyota Way" PR. Toyota never acknowledged the mistakes they made - instead they quietly sacked the head of that division and then most of the staff, and rebuilt it from the ground up. Today's Toyota ECUs appear to be designed quite differently.

There is a long history of American automakers and their cohorts in the media and elsewhere who use these scandals to scare buyers away from Japanese makes. See what they did to Suzuki with the Jimny in the eighties for the most egregious example. I watched this “scandal” closely at the time, and it smelled like a rat all along. Fortunately, these tactics are ineffective, but there will always be anecdotal persistence on discussion threads like this. Nothing in any credible source materials proved this issue. Settling lawsuits is not an admission of guilt; it’s Japan Inc rolling their eyes, pulling out their pocket books, and making a pragmatic cost/benefit analysis. The lengths to which the case for the unintended acceleration flaw went is proof enough of what their aims were. If you want a counterpoint to the study you cited, here’s one from the National Highway Traffic Safety Administration when American media and automakers pulled the same stunt to slander German makers in the 1980s: https://www.autosafety.org/sites/default/files/1989%20NHTSA%...

Putting aside other aspects of your bias, it's ironic that you insinuate that a "slander" happened, because the opposite effect occurred (as described by slide 9 in my second link). The US Transportation Secretary claimed that Toyota was exonerated, even though the report given by NASA did not support that claim.

The “slander” (not a legally accurate use of the term) referred to the media campaign, which was effective around 2010-12 and has lingering effects, as seen here.

You're trying to dismiss a series of technical reports on Toyota's software engineering defects - reports that directly acknowledge that it's impossible to prove causality beyond reasonable doubt because of the software engineering defects - on the basis of their similarity to an unrelated report from 1989, and because you "smell a rat".

Just to flag - the US Government (which operates NASA) was a major stakeholder in GM and Chrysler, major competitors of Toyota, when they investigated Toyota.

Again, no. Torque converters do not have gears or neutral. They are hydraulic couplings. The output shaft of the torque converter turns slower than the input shaft when the car is accelerating, and there is no fixed relation of the input shaft speed to the output shaft speed. The gears are all in the transmission after the output shaft of the torque converter.

This is just plain wrong. The poster didn't say they were at a standstill but that they had the brake pedal pressed. That means the engine could have downshifted into 1st gear and started to rev noticeably. The engine would also produce more power (and sound very different) when taking up load, even under the same RPM. Finally, as the other reply pointed out, the engine can spin a clutch or CVT.

The poster said that the brake was fully depressed with around 100kg of force. It won't take long for that to be in a standstill. And there's no way acceleration can overpower that kind of brake power. The sound might be scary but it won't change much about brake performance.

Bad clutch plus bad brakes (but not so bad as to be unserviceable in light, normal use). Story is not implausible.

Cars of that vintage still had cable operated throttle bodies which could stick for a number of reasons.

Often badly-placed floormats could also cause this issue. I remember a few automaker recalls in the prior decade that comprised of adding hooks to the floor boards to avoid floormat movement.

Reply on news.ycombinator.com