The Toyota Witch Hunt
businessweek.com
businessweek.com
The rest is in the C&D article referenced, I believe, by the author:
http://www.caranddriver.com/news/car/10q1/toyota_recall_scan...
The C&D article does leave the door open to the possibility of the brakes not stopping the car. If you pump the brakes, rather than firmly putting them down once, you may lose vacuum, or overheat the brakes in a series of "slow down a little" steps. The vacuum is apparently hard to replenish when the throttle is open.
The advice about turning the car off works sometimes, but not always. Suppose you're in a rental Lexus like the one the CA state trooper died in. Would you know that to kill the engine, you have to depress the off button for 3 seconds continuously? Three seconds is a long time with the throttle wide open.
It seems the key technical mistake on Toyota's part, as pointed out in the C&D article, is to omit an interlock that kills the throttle when the brake is depressed. This is apparently standard on many other cars.
Toyota's damage control on this has been disastrous.
It's tempting to blame drive by wire, since electronics is black magic to most people. Keep in mind that most passenger jets in the sky are fly by wire, these days.
For you to point blank claim this to be true, when it's been flatly denied several times, seems a bit over the top, don't you think?
http://articles.latimes.com/2009/nov/29/business/la-fi-toyot...
Toyota is claiming it's not the DbW system, but they also thought they had this issue fixed a long time ago. It's a hard thing to prove unless you see it yourself (unlike a broken cable like we used to have before DbW).
Even Woz thinks is the DbW software:
http://news.cnet.com/8301-13924_3-10445564-64.html?tag=mncol...
Complaints of out of control acceleration increased after 2000. I guess it's the millennium bug. In other words: correlation is not causation. Moreover, no numbers are presented to actually support this correlation. What about 1996-1999? 100 cases would be totally plausible.
Even Woz thinks is the DbW software With the rather important difference that it doesn't actually cause any problems, because he can just brake and make the acceleration stop.
You postulate that both sensors short out at the exact same instant, since any other scenario would power down the vehicle. They did it within milliseconds of each other, and they did it at some intermediate, yet believable level. (They didn't "peg at maximum", as that would also alert the software there was a problem. Instead they pegged, identically, at some believable value.
I'll take the other side of that bet any day.
Does Congress have a legal mandate to investigate anything it deems to be interesting/fishy/a potential target for legislation? Has Congress always conducted these types of investigations or is it a modern phenomenon?
Wikipedia: Investigative hearings share some of the characteristics of legislative and oversight hearings. The difference lies in Congress’s stated determination to investigate, usually when there is a suspicion of wrongdoing on the part of public officials acting in their official capacity, or private citizens whose activities suggest the need for a legislative remedy. Congress’s authority to investigate is broad and it has exercised this authority since the earliest days of the republic. Its most famous inquiries are benchmarks in American history: Credit Mobilier, Teapot Dome, Army-McCarthy, Watergate, and Iran-Contra. Investigative hearings often lead to legislation to address the problems uncovered. Judicial activities in the same area of Congress’s investigation may precede, run simultaneously with, or follow such inquiries.
It's not a recent phenomenon at all. For example, the Crédit Mobilier hearings were in 1872.
Credit Mobilier: "The distribution of Crédit Mobilier stocks by Congressman Oakes Ames along with cash bribes to congressmen"
Teapot Dome: "control of U.S. Navy petroleum reserves at Teapot Dome in Wyoming and at Elk Hills and Buena Vista in California, were transferred from the U.S. Navy Department to the Department of the Interior"
And Army-McCarthy, Watergate and Iran-Contra are all obviously directly tied to the government.
In those cases it makes sense for Congress to get involved. It's when Congress start to investigate "private citizens whose activities suggest the needs for legislative remedy" that I get concerned. That type of broad power apparently gives them free-reign to investigate anything they don't like.
It seems to me that in the past (15 or 20+ years ago) Congress didn't get involved in non-government related "scandals" like Toyota and college football, I'd be interested if anyone has examples to the contrary.
http://news.google.com/newspapers?id=DDAbAAAAIBAJ&sjid=d...
There are lots of other examples if you search newspaper archives using Google News. I encourage you to go look for yourself.
It also allow them to organize two minute hate sessions for the media's consumption.
Article I, Section 8, Clause 3
For model year 2009, Volkswagen had 11.5 complaints per 100k vehicles, whereas Toyota had 7.5 and BMW 5.8
People seem to forget that this only happened until after Toyota had repeated recalls, each promising to fix the problem.
Is there a ghost in the machine? I don't know, Woz certainly seems to think so. Replacing floor mats and then gas pedals seems awfully like what happened with the Therac 25 radiation overdoses in the 80s.
(The Therac 25 manufacturer first added an extra hardware switch to the radiation therapy machine and claimed it was "an order of magnitude safer". When more people died, the quick emergency fix was to remove the "up" button from the VT100 console.)
Actual American political debate hardly happens anymore. What we have now is theater. The mentality of the LCD American voter is simply abysmal. You can see it from the low or sloppy intellectual content of the "product" which is presented to them by politicians, journalists, and other media workers.
Heh, that works on two levels--not just "lowest common denominator" but also the American voter who watches too much TV news on their LCD HDTV.
It has happened to me once when I went over a pothole and my wife a number of times (she drives more).
I'm in Canada and my local dealer can't reproduce the problem yet, but put yourself in the position of a Toyota owner who is no longer confident in being safe and who would get less than half the purchase price for a car with < 70,000 KM on it.
The Honda dealer doesn't want Toyotas as tradeins; they can't sell them.
Can you feel the anger?
I don't want to have to buy another car; I want to work on my startup. But I don't want to worry about my family's safety or buying some used car.
Toyota should be forced to buy back their crap products at some reasonable percentage, since these problems remind me of "heisenbugs" in code...
I would bet any sample of 3000 complaints about the Ford Pinto or any product had at least some stupid and some dishonest ones.
http://www.leftlanenews.com/report-toyota-electronics-design...
This is what the article says about the prof:
Instead, we were treated to Dr. David Gilbert of Southern Illinois University, also a guest of Mr. Kane's, who claimed to have found how Toyota's electronic system could totally malfunction, creating a runaway car—and claimed he'd found the error in less than four hours. Spoiler alert: Dr. Gilbert was assigned this work by Kane's safety advocacy firm, with at least partial funding by trial lawyers.
Here, too, is a problem: Dr. Gilbert said he relayed the results of that test and his concerns directly to Toyota. In short order Toyota looked into Dr. Gilbert's claims and found them not to be valid in terms of creating unintended acceleration. Then, to the company's surprise, it watched his appearance with Brian Ross on ABC News this past Monday night, Feb. 22.
According to Toyota, it now appears that Dr. Gilbert had done something completely different in order to get a Toyota Avalon to accelerate under its own power. Toyota offered to evaluate Dr. Gilbert's Avalon, with ABC in attendance, and see what he did electronically to cause it to accelerate.
Additionally, Toyota is fairly adamant that Gilbert's "test evaluation" on ABC News was not the original "discovery" he relayed to them on Feb. 16.1. to a certain degree, the US is punishing Japan due to their recent actions in the bond markets. This is a very important problem for a very small amount of people, but it makes great headlines. Do you think the US government is genuinely interested in resolving this properly from an engineering perspective?
2. that 100% "drive by wire" systems should not be allowed...there should be a big red "holy shit" kill switch on the dash of modern cars that kills power to everything except steering and brakes
So, imho, its a very serious problem, that affects very few people (in the big scheme of things), and there is a very simple solution, for going forward technology at least, if not existing models. But do you guys get the feeling that this is what the conversation is actually about? Something seems a bit odd to me.
But then I'm one of those conspiracy theorists that think when you're running 15% budget deficits and have real (as measured using historical methods) unemployment of 20%, and many bond issues are taken down 50% by the federal reserve (literally printing money to buy your own debt), that you will start to observe unusual things happening in the marketplace. The adherents of Occam's Razor would imply that that I am simply crazy, becasue the simplest answer is always the correct answer. So I guess thats it.
If people are upset with how their Toyota cars are operating, then no one stands to lose more than the Toyota company itself in lost revenue. So what additional incentive or pressure could congress apply here except increased publicity? Let's see...we lose a billion dollars, or get chewed out by some old guys for a few hours, which sounds worse?
It seems really silly to watch politicians chastise a CEO and talk down to him. Good engineering (especially on a massive scale like that) is really hard. If they think they can do better then no one is stopping them (members of congress) from starting their own car company to try and compete. Otherwise, I'm not sure why they feel qualified to criticize.
Something else that bugged me is the quote from one of the witnesses which said 'Shame on you Toyota for being so greedy'. So...you're saying they intentionally pissed off all their customers as some sort of plan to make more money? I'm so confused by this mentality people have that greed in corporations is a bad thing. Yes, you could argue they sacrificed long term quality for a short term savings, but this is hardly greedy, it's just bad business practice that leaves you with LESS money.
This may be unrealistic, but I really wanted to see the Toyota CEO reply to congress with some sort of statement like, "We won't be appearing at any hearings. We are a private company and don't answer to government. We only answer to our customers. So if any of our customers have questions here is a toll free hotline. Members of congress, if you are a Toyota customer, feel free to call as well, but you're role in government won't be taken into account in how we respond."
I pray people realize this is folly... I don't know what we're gonna do if stupidity brings down the pillar of the auto industry, through no fault of it's own.
http://en.wikipedia.org/wiki/Audi#Reported_sudden_unintended...
Of course, they should just shift it to neutral and let it redline.
as for the first part, doesn't have to be your hand. Depending on the vehicle, you might be able to get your toe under the pedal.
This is not necessarily true. It's entirely dependent on the brakes involved and the engine involved; any given vehicle could go either way. I personally have witnessed steel brake rotors melted and dripping off the vehicle; the brakes had locked up at highway speed and the driver had continued driving, reporting afterwards that the vehicle seemed a little sluggish but not excessively so, so he hadn't stopped. For that vehicle: the brakes lost. For other vehicles, perhaps ones with dinky engines and big fat brakes: the brakes may win. It's certainly not the sort of thing where one can say "brakes always win".
Car and Driver's article ( http://www.caranddriver.com/features/09q4/how_to_deal_with_u... ) seems to assume that Toyotas don't cut the engine power when the brakes are applied, but their own tests belie that. I would guess that Toyota's software cuts the power LESS than the competition's software does. But if you're stopping from 70mph with and without the throttle at full, and you notice a minimal difference - as C&D did - I can guarantee you that the throttle is being reduced when the brakes are applied. Maybe not to zero, but reduced.
Once I accidentally slammed on the gas thinking it was the brake. (I got confused switching from a manual transmission to an automatic.) It was scary...the car jumped a curb and my driver's ed teacher was furious. He slammed his brake and it stopped the car.
Was reversing into a spot, and hit gas instead of brakes.
If you are in a car moving along and you have to kill the engine, or the engine dies, get on the brakes as fast as you can and don't let go. You can ease up on the brakes almost to the point of letting go of the pedal, but don't come off all the way.
The reason is the brakes will maintain the power-assist vacuum from the engine, even when the engine is later switched off, until they are released completely. The 'power-assist' strength will remain for tens of seconds, at least. I've tried this many times before. I can't guarantee it works on modern cars, but it sure did work on my '86 Toyota with standard hydraulic vacuum-assist brakes.
The vacuum assist is just that - an assist. The brakes still work if you don't have it, you just have to put some extra force on the pedal. It isn't even that much force; I've driven older vehicles that don't have any assist at all and it's not that much harder to push the pedal.
The major failure of all automobile manufacturers is that they haven't implemented a fail safe mechanism in the event of engine failure at velocity. If they had it would literally be as simple as turning the engine off and the brakes deploying at a reasonable (IE not maximum) pressure.
The fact of the matter is that every vehicle except consumer automobiles fail safe from aeroplanes to tractor trailers (power failure leads to de-pressurizing of the air compressor and the brakes apply harder as the air level decreases). Toyota shouldn't be being hauled up, they all should be for risking civilian lives. IMO you accept certain risks when you pilot an aircraft, train, or transportation vehicle, however these are all typically far safer than the vehicles readily handed to consumers. That just is not right.
Though I guess they already have that in the form of "emergency brake". Do you suggest they implement electronics to perform these functions when the e-brake is depressed?
The increased use of electronic control in automobiles means that cars now have the problem that software has had for a while; real but difficult to reproduce bugs. When the control of the car has gone away from physical devices like rods and wires, not only is it harder to find the control but any problem that exists just seems creepier. This is another that I suspect is driving the current reaction.
What bugs me is that people don't keep things in perspective. I see people all the time claim they want to go back to the way cars were (more mechanically simple, or whatever). What they seem to forget is that cars broke down a hell of a lot more often back then.
In fact, they broke down so often that it wasn't such a big deal. Now we expect them to work flawlessly and when something goes wrong it somehow seems worse.
It's almost like Microsoft designed these cars: "What? I push Start to shut down the car/computer?"
It's also very convenient for opening the trunk without a key as well. It has proximity sensors by the driver and passenger doors, as well as the trunk.
Yes, the computers do.
However, if you're going to have an accident, engine damage is pretty low on the list of concerns, even if you ignore the fact that accidents cause engine damage.
Switching off the ignition can lock the steering....
You also lose the power-assist on the brakes.
Given the above, you might not notice that you've also lost the power-assist on the steering.
Turning the engine off with the key need not lock the steering, but it will lock the steering if you turn the key past off to lock.
In a run-away engine scenario, are you likely to stop turning the key in a position that you don't use very often? Or, are you likely to turn it as hard as possible?
There may be an interlock that is supposed to keep the key from locking the steering while in drive or neutral, but do you trust it? (What about manual transmission cars? I know that I can lock my steering while in gear. I forget whether it requires clutch-in though.)
(Note, post happen to have the same title, but totally different content)
I just found Hacker News yesterday, this site is amazing.
Say goodbye to your free time and productivity.
But they do decide on the basis of which drum they think people expect them beat.
If this theory is true, it's saddening because Toyota's image is completely tarnished in many people's eyes, and they won't be able to recover from this for years (if ever).
The real question is, just how in-control is the computer. And how much has to go wrong such that uncontrollable acceleration happens and nothing can override that acceleration such that no matter what you do the computer makes the car go into "drive" and accelerate to the max, without responding to other parts of the system like the breaks and such.
The ignition switch would have to lock or otherwise malfunction
The shift lever or clutch or automatic shift lever would have to malfunction
The brakes would have to overheat (does not really happen much in a production car). It is not 100% true that brakes will always overpower the engine, but the cases in which they don't are the ones with massive 700hp engines with big 'ol turbos and stock drum brakes, and in that situation it's the owner's responsibility.
There would also have to be no runaway truck exits.
If you were creative and knew something about cars, you could pull one of many fuses from the fuse box. On cars I've serviced, the computer fuse is always in the driver kick panel, so just rip off that panel and start yanking fuses while watching the road.
It should be noted that the e-brake on fwd cars is not your best choice (though still a decent choice); while a locked up wheel provides drag, it does not apply as much drag as a rolling wheel with maximum braking (which can easily be obtained cause of ABS). Also at extreme speeds you run the risk of a spin, which puts you possibly at risk for a roll.
In recent Toyota vehicles the pedal isn't connected to the throttle--it's drive by wire. In theory the computer could continue accelerating regardless of the gas pedal's position.
You also toggle bits in your code. One in the main loop, another in any interrupt service routine that you are dependent upon. These bits then go to a GPIO so that the GPIO toggles. If the ISR freezes or the main loop freezes, the GPIO quits toggling. You then have an hardware, RC-Diode type circuit, that stays high if it's input toggles, but goes low, if it's input goes low and stays low, or goes high and stays high. The output of this circuit drives a relay, so that if the CPU quits toggling the GPIO, the relay clicks off. The output of that relay is a dead-man's switch that powers down the vehicle. The end result is a hardware fail-safe if the CPU misbehaves. There are other tests to confirm that the relay isn't frozen, that the RC-Diode circuit isn't frozen, etc.
So, in theory, if a high-school sophomore designed the vehicle, you might have a problem. Otherwise, I think you don't give the engineers that design these systems enough credit.
"Well, I have many models of Prius that got recalled, but I have a new model that didn't get recalled. This new model has an accelerator that goes wild, but only under certain conditions of cruise control. And I can repeat it over and over and over again--safely."
"This is software. It's not a bad accelerator pedal. It's very scary, but luckily for me, I can hit the brakes," he said.
-- Steve Wozniak
More suspect, in my opinion, is that in a cruise control, you have a feedback system, based on car speed. The car speed is almost certainly redundantly sensed, so that's no worries, but the feedback loop itself could potentially go oscillatory if there were other variables introduced that hadn't been designed for. Those variables could be pretty subtle. For example, maybe the gear motor that you use to mechanically control the engine gets sourced from another distributor, and they give you a better one, that has more torque. Perhaps that throws off the stability analysis that you had done. Tons of things could change somewhere between the 500K's car and 1M's car you produced, lots of different vendors and permutations could come into play that could throw off the stability of a cruise control, I would think.
EDIT: They're probably using feedback in the control motors as well (servos), so that's a non-realistic example, but it illustrates the problem.
Brain-dead cruise control behavior is nothing new. My '92 Porsche 968 would cheerfully redline the engine if you disengaged the clutch with cruise active.
I'm not sure if automotive systems are held to any certification standards. Maybe someone working in this field could answer that?
Unfortunately, this is probably considerably more dangerous than driving while on your cellphone.
Most of the other actions suggested can be taken while in standard driving posture. Those will be much safer than reaching for fuses under the dash.
In any case, I was just throwing out an idea.