Toyotas Suddenly Accelerate; Owners Up In Arms 1146
cyclocommuter writes "Some Toyota owners are up in arms as they suspect that accidents have been caused by some kind of glitch in the electronic computer system used in Toyotas that controls the throttle. Refusing to accept the explanation of Toyota and the federal government (it involves the driver's-side floor mat), hundreds of Toyota owners are in rebellion after a series of accidents caused by what they call 'runaway cars.' Four people have died." The article notes: "The National Highway Traffic Safety Administration has done six separate investigations of such acceleration surges in Toyotas since 2003 and found no defect in Toyota's electronics."
Re:Carmakers lie (Score:5, Interesting)
Floor mat, really? (Score:4, Interesting)
So Toyota says it's floor mat. But here's something I don't understand after reading TFA... all people who had that problem (and lived to tell the tale) insist that they were braking hard as the car was accelerating. If it were really just gas pedal stuck in a floor mat, then surely applying brake would force the car to decelerate regardless?
problems with complexity (Score:5, Interesting)
F-35 joint strike fighter - 5.7 million
Boeing 787 - 6.5 million
Premium class automobile - ~ 100 million
IEEE Spectrum: "How hard should it be to stop a runaway luxury car?" http://spectrum.ieee.org/blog/computing/it/riskfactor/how-hard-should-it-be-to-stop-a-runaway-car [ieee.org]
IEEE Spectrum: "This car runs on code" http://www.spectrum.ieee.org/green-tech/advanced-cars/this-car-runs-on-code [ieee.org]
Re:Put the damn thing in neutral! (Score:3, Interesting)
With Toyota Hybrids, the gearshift lever is just a switch for all positions other than park. Flip it all you want between R, N, D, B and all you're doing is asking the ECU to alter what it does with the 'synergy drive', it doesn't change any gears.
I've played around a bit with my Highlander Hybrid, it does some odd stuff... Put it in Park or Neutral, give it gas, and it'll fire up the gas engine and rev it a bit? Floor the brake pedal, give it some gas, and again, it'll rev the gas engine but not transmit any power to the wheels?
Again? (Score:1, Interesting)
Ok, repeat after me: there is no production car on the planet with an engine capable of suddenly overpowering simple hydraulic brakes.
Know what Audi's engineers found back in the 80s? They found gas pedals bent out of shape by people standing on their "brakes".
This is not "news for nerds". This is the same bullshit driver error as before, just the computers playing boogeyman are a bit more advanced this time.
P.S. This opinion is based on the statements quoted in the article. The laws of physics may not be widely known, but your car can't nullify them.
Re:Floor mat, really? (Score:5, Interesting)
I have owned many Prius's. I currently drive a 2010 one. Let's say that I'm in some place where the speed 85 mph is legal. I can nudge my cruise control speed lever and my speed barely goes up, say from 80 to 81.I nudge at again and again, up to 83. Then I nudge it again and the car takes off, no speed limit. Nudging the cruise speed control lever down has no effect until I've done it about 10 times or more. By then my Prius is doing 97. It's scary because it's so wrong and so out of your normal control. I tested this over and over the night I observed it.
It's scary because you don't think of things like putting the car in neutral when this happens. I am sure you can't turn the car off with the keyless power button, the only option on this model.
Braking does disable this scary cruise control effect. It is a natural response, so the problem is mitigated a great deal.
I have not seen this happen before so I think it's new to the 2010. I have the package which includes parallel parking assist and cruise control distance limiter.
First Hand Experience (Score:5, Interesting)
I was driving a '98 Toyota Camry. Foot on the brake. B-R-A-K-E. Yes, I know the difference. Car in drive. Waiting for a right turn. The car revved up high. I did manage to throw it into neutral, and the engine continued to surge. Luckily I didn't hit anything. And it was pure luck.
The car did not have All-weather floor mats.
I have racing experience, and a background in Mechanical Engineering.
The reason the problem hasn't been found is that it's probably a subtle fault (like the AT&T crash back in the early '90s, or the stress concentrations in the DeHavilland Comet) and they're (by they I mean the NHTSA) probably not looking very thoroughly, due to lack of manpower. They don't do investigations of car crashes the way they do for other serious engineering failures, like plane crashes or bridge collapses.
Re:Same thing happend to Audi a few years ago (Score:2, Interesting)
My 2010 Prius has a package that includes parallel parking assist and cruise control distance limiter. In some fictional state (let's say 'private property') I tap my cruise control speed lever up and the car speed increases from 80 to 81.I tap this leve again and again, up to 83 mph. Then I tap it again and the car takes off without speed limiting. Tapping this lever down has no effect. The car is shortly up to maybe 97 mph. I repeated this many times.
One doesn't think of things like putting the car in neutral instantly.
The natural braking action does disable this effect.
Re:God damn it this again (Score:5, Interesting)
What's happened, in each case, is that the dealer or driver put winter floor mats, either OEM or aftermarket, on top of the regular carpeted mats. What this means is that, unlike the normal mats, they're not pinned down in any way and will slide forward. In the case of the CHP officer in the rental Lexus, the dealer slapped truck mats down on top of the "normal" Lexus mats
What happens next is easy: the mat jams the accelerator pedal. What happens after that is that people panic, do the wrong things, and plow into people in front of them.
And what happens after that is lawyers.
There's no car you can buy today where you cannot overpower the engine with full braking force. Try it: stand on the accelerator with your left foot for a while, then stand on the brake. Push both down as hard as you can; your car will slow down and stop. It won't be happy about it, but it will. The drivers in this case didn't do that: they panicked and didn't press the brakes hard enough.
Nor did the slap the car into neutral or stop the car. And yes, the car could have a gated shifter or a Prius-style stick. You can also turn the car off: even with an engine-stop button, all you need to do is holdit down. Again, in both cases it requires the driver to not panic.
There's no real way around the human factor in this. I've seen drivers who two-foot drive. I've seen drivers who, when they're presented with a scary situation, take their hands off the wheel and cover their eyes. I've been in the car when a driver's panic reaction was to flail madly at the pedals with her feet and see-saw the wheel---in that case, the car rolled. While the floor mats can create a problem, and while Toyota could fix it by mounting them a little bit higher, you'll never truly idiot-proof a car until the car drives itself.
The solution to the likes of this are systems like stability control, ABS, Volvo or Nissan's Lane Departure Control and Mercedes' and Lexus' Pre-Safe crash mitigation systems: keep the car on-course and stable, allow the driver to maintain control and, if a crash is imminent, apply full braking force, tighten the seatbelts and pre-charge the airbags. Oh, and call 911.
Re:Floor mat, really? (Score:2, Interesting)
They've only existed for about a decade. Do they really wear out that fast, or do you just like the new car smell?
Re:Carmakers lie (Score:5, Interesting)
This is a "feature" of German cars due to a law in Germany mandating that the speedometer must NEVER read lower than real speed... even if the car has non-standard wheels and tires fitted.
Porsche and BMW exaggerate speed the most, and the theory is because owners of these cars are quite likely to upsize their rolling stock (and thus make the speedo read lower). It's annoying, but it's simply in response to a legal requirement.
Car and Driver did a test on several vehicles a bit over a year ago. GM vehicles were the most accurate... around +1 mph on average.
Re:Floor mat, really? (Score:5, Interesting)
I used to have one of those "sudden acceleration" Audi 5000's (1979). It happened to me once and I figured out exactly what happened within five minutes. It wasn't the computer or the floor mat or anything. The accelerator pedal linkage was a solid rod which ran up a few inches from the tip of the pedal, then turned left to pass behind and above the brake pedal. If you put the arch of your foot on the brake pedal, your toes could contact the accelerator rod and depress it. Even light braking action was enough to impart enormous acceleration. The harder you stomped on the brake, the more the engine overcame the braking action. The fix was to put a metal guard plate over the rod behind the brake pedal.
Some automakers (e.g., Audi) are framed. (Score:2, Interesting)
About 20 years ago, the Audi 5000S had the same supposed problem. You can read about the problem at the "New York Times [nytimes.com]", the "Los Angeles Times [latimes.com]", and the "Business & Media Institute [businessandmedia.org]".
The trouble began when "60 Minutes" (of CBS News) broadcast a story about a woman who killed her son when she accidentally pressed the accelerator pedal instead of the brake pedal. Her son was standing in front of the car. The woman, refusing to admit guilt, accused Audi of producing a defective car which accelerates automatically without driver intervention. She even filed a lawsuit against Audi. (Later, the court determined that she was at fault, but that fact was never broadcast in the original "60 Minutes" program.)
The sales of Audi vehicles fell dramatically after that "60 Minutes" program.
The Audi 5000S was never defective, but it did have 1 minor inconvenience. The accelerator pedal and the brake pedal were much closer to each other than they were in a traditional American car. This closeness was something to which a small subset of American drivers could not become accustomed. They sometimes did press the accelerator pedal when they intended to press the brake pedal.
As for the Toyota Camry, is it defective? The probability of it being defective is higher than the probability of the Audi 5000S being defective. Consumer-safety standards in Japan are lower than the standards in the European Union.
Even from an engineering perspective, the Toyota Camry is a dangerous design. For example, the transmission is mechanically separated from the automatic-transmission lever (that the driver uses to change gears). The lever is connected to an electronic box that sends some electrical signals -- along copper wires -- to the tranmission to control it: the process is drive-by-wire. Supposedly, Toyota used 2 identical sets of wires (for reasons of fault tolerance) from the electronic box to the transmission.
Another participant in this discussion claims that Toyota also mechanically separated the accelerator pedal from the fuel line. Toyota appears to have used drive-by-wire throughout the design to eliminate some metal -- thus saving money.
Do not trust the fault tolerance in mass-merchandise products. Fault tolerance is expensive and is meant to be expensive. Toyota likely tried to save some money on the fault tolerance, and it was not able to protect the vehicle from the 1-in-1,000,000 chance of a transient fault in the electronic circuits. The chance of a glitch is low, but the probability that it occurs exactly once among 200,000 vehicles is high.
The fact that only a handful of people have been affected by the freak accelerations matches a distribution of a low-probability electrical glitch. If you own a Toyota Camry, I suggest that you sell it as quickly as possible and get an old-fashioned-technology vehicle without the drive-by-wire. The Ford Fusion exceeds the quality of the Toyota Camry, does not use drive-by-wire, and costs much less than the Toyota deathtrap. Think about it.
Re:Carmakers lie (Score:2, Interesting)
GPS is horribly inaccurate to start. By that I mean its accuracy is out by a larger margin than you report the Hyundai showing. On a good day you'll be within 5% of your actual speed depending on the number of satellites in view. More sats, more accuracy.
Those road side radar things, I don't know about their accuracy, but I would imagine angle of approach will have something to do with how accurate they are.
Tire wear will have an impact on your car's actual speed, the more wear the faster your tires have to rotate to maintain the same speed. Depending on your speed this slight difference could account for your higher mph report.
Basicaly I am trying to say is that anything better than about 5% margin of error and you should be happy with the reported speed.
Re:That's why cars had ignition keys. (Score:3, Interesting)
There is a butterfly valve on the air intake designed to keep this from happening (chokes the air from the engine) but I think it might have sucked that through too.
Re:Floor mat, really? (Score:5, Interesting)
So Toyota says it's floor mat. But here's something I don't understand after reading TFA... all people who had that problem (and lived to tell the tale) insist that they were braking hard as the car was accelerating. If it were really just gas pedal stuck in a floor mat, then surely applying brake would force the car to decelerate regardless?
Funny thing about the brake. It's operated in the same way as the accelerator, and located in a place where most people don't normally look while they're operating it.
This was famously the case with Audis in the early '90s. Audi, designing for the heel-to-toe autobahn driver, put the brake and accelerator pedals closer to each other than on most US-market cars. Cue a number of reports in the US from drivers screaming, "I was mashing the pedal as hard as I could, and the car just wouldn't stop! In fact, it kept going faster and faster!"
No defect was ever found (though that didn't stop the media from demonstrating it) and the problem was only reported in the US, although the same cars were sold worldwide.
Re:Carmakers lie (Score:3, Interesting)
GPS was never intended to measure speed. All it does is get position and repeat. Speed is calculated based on change of position over time, which is good for averaging but NOT good for spot measurements.
And there are several ways to calculate change of position, not all GPS's do it the same. Haversine, Spherical Law of Cosines, etc. Some may even use geometry on UTM (which can work cause rhumb lines are great circle lines in mercator projections). But again, what happens between refreshes and recalculations is lost in the averaging.
Those road-side radar speed signs...real police radar (the kind admissible in court) have a calibration regimen in order to stay accurate. The roadside radar signs....not so much.
I'm not arguing your speedometer is off...could be tire size, bad design, conspiracy...whatever. But GPS and radar signs aren't the standard by which to judge.
Re:PEBAAC (Score:5, Interesting)
It is exceedingly easy to test the Throttle Positioning Sensor in modern vehicles. In fact, your ECU probably tests idle throttle position every time you turn the key on for a while without staring the engine. The ECU will also log 'implausible signal' for TPS that get an out of range reading, or inconsistent reading throughout the range.
You are basically correct. I have first hand hacking experience with the drive by wire throttle because my Grand Challenge team automated a Toyota Prius for the last Grand Challenge. There are 2 completely independent signals that go from 2 independent sensors on the pedal to the computer throttle component. The signals have to move in lock step with each other or the computer will detect a fault. If a fault is detected the throttle goes completely off and the car has to be turned off and turned back on to recover.
So for the throttle to stick down both pedal sensors have to fail in the same way at the same time, which seems highly unlikely to me. Or there could be a bug in the computer control section, bus as a software engineer I can assure you that that would be impossible. ;-)
Re:Carmakers lie (Score:2, Interesting)
It's akin to setting your clock 5 minutes ahead so you won't be late.
It's fucking stupid and it defeats the point of the instrument.
Totota problems..... (Score:3, Interesting)
A friend of mine just had her toyotas engine die because a cheap $20 hose was prone to failure, When it failed, the oil all leaked out WITHOUT the oil light coming on. This issue has occured for quite a few people with the Toyota Avalons (though almost all the ones with this issue were made in the same year)
Did I miss the joke? (Score:3, Interesting)
Hey, when you come up with that Ford Fusion that has a non-electronically controlled automatic transmission and that doesn't have the electronic throttle control that they're wallpapering the world with press releases about, I've got some nice land to sell you to park it on...
brakes, neutral, brakes, ignition off (Score:4, Interesting)
You put the car in neutral and the engine goes to 8000RPM. That will freak you out, I guarantee.
Not really. I had a 1980 Ford Granada (US/Canadian model with 4.1L 6 cylinder automatic http://en.wikipedia.org/wiki/Ford_Granada_(North_America) [wikipedia.org], a totally different car to the European model) in the early 80s. Its throttle was entirely mechanical, and the linkage to the butterfly valve involved a rod which twisted when the gas pedal was pushed down. One of the cheap clips holding this rod at one end got broken, and as a result, when I floored the throttle (required for manoevering in fast traffic on urban expressways), one of the engine hoses got jammed under it, effectively locking the trottle wide open.
I think the other highway users were more freaked out than I was, since I was driving with the brake lights on while holding the car at the speed of the traffic (65-75mph). There was no shoulder to stop on, but it was only a few km to the next exit ramp, where I dropped to neutral and then switched off the motor (as it raced towards bursting speed) and coasted to a safe halt with manual brakes and manual steering. A quick look under the hood revealed the problem, but I waited a few minutes extra to allow the motor and brakes to cool properly, before continuing home.
Despite driving more cautiously, I had the same thing happen two more times, before the clip was replaced a couple of days later.
Re:Totota problems..... (Score:4, Interesting)
Ever drive a modern BMW? No dipstick. That's right, no checking the oil at all.
Besides, checking oil level only determines if there is enough oil in the sump, not if there is a major failure of the oil system. For that you need a gauge or an idiot light.
Drive By Wire (Score:4, Interesting)
The Audi 5000S was never defective...
That's right it wasn't. It was people blaming the equipment for their own failures.
As for the Toyota Camry, is it defective? The probability of it being defective is higher than the probability of the Audi 5000S being defective.
Umm, please show your calculations. You already admitted that the Audi 5000S was not defective so this should be interesting.
Consumer-safety standards in Japan are lower than the standards in the European Union.
Even if true (and you've provided no evidence that it is true) that has precisely nothing to do with a car sold in the United States where US consumer safety standards apply. Never mind that the Toytoa Camry is produced right here in the US (also in Japan, Russia, China and Australia).
Even from an engineering perspective, the Toyota Camry is a dangerous design. For example, the transmission is mechanically separated from the automatic-transmission lever (that the driver uses to change gears). The lever is connected to an electronic box that sends some electrical signals -- along copper wires -- to the tranmission to control it: the process is drive-by-wire.
Drive by wire does not make it a more dangerous design. It has DIFFERENT failure modes but different is not the same as dangerous. Fly by wire has become state of the art in airplanes where they have much stricter reliability standards so the technology clearly CAN be safe. While it is certainly possible Toyota has a defective system, I want to see some actual evidence of a fault beyond a few anecdotes of customers.
Do not trust the fault tolerance in mass-merchandise products
You do that every day whether you are aware of it or not. There is a reason we have product safety and liability laws. You trust your life to mass merchandise products every single day of your life.
If you own a Toyota Camry, I suggest that you sell it as quickly as possible and get an old-fashioned-technology vehicle without the drive-by-wire.
Good luck with that. Lots of cars are already drive by wire and within a few years nearly all will be. Enjoy driving unsafe older cars.
Fault tolerance is expensive and is meant to be expensive.
Actually it doesn't have to be expensive at all. A pipe wrench is a great example highly fault tolerant engineering but it isn't expensive. Fault tolerance CAN be expensive but it doesn't have to be. With an appropriate design it can even be cheaper.
The fact that only a handful of people have been affected by the freak accelerations matches a distribution of a low-probability electrical glitch.
It also matches the distribution of a handful of people standing on their accelerator pedal and being too embarrassed to admit they weren't using the brake. Remember the Audi? It's entirely reasonable to believe this is people trying to get money via our legal system instead of an actual engineering fault.
The Ford Fusion exceeds the quality of the Toyota Camry, does not use drive-by-wire, and costs much less than the Toyota deathtrap. Think about it.
The Ford Fusion DOES use drive by wire [fordvehicles.com]. Every hybrid car is drive by wire and soon enough so will (nearly) every non hybrid. Drive by Wire has FAR too many advantages in both cost and features to not be used.
Regarding quality, JD Power thinks you are full of crap and I tend to believe them more than you. 2010 Ford Fusion [jdpower.com] vs 2010 Toyota Camry [jdpower.com]
Re:Same thing happend to Audi a few years ago (Score:2, Interesting)
Re:Put the damn thing in neutral! (Score:5, Interesting)
The guy apparently did try pulling the car out of gear and into neutral, but it didn't do anything (many auto transmissions are electronically controlled and have failsafe mechanisms to prevent desctruction). He also tried shutting off the ignition - but this vehicle like many other fancy new vehicles doesn't have a key ignition. It's a button, and when pressed while driving it won't turn the engine off. Buried in the owners manual is a single sentence that essentially says that in order to turn the engine off while driving, you have to hold the button down for three seconds.
Now I've never gone 120MPH on a congested roadway, but I could only guess that when you're trying to avoid hitting anything, it's not a simple task to 'reach down and pull the floor mat off the pedal'
Although recently I had the accelerator pedal stick to the floor on a 2009 F-350 diesel. It happened as I was accelerating from a complete stop. I immediately pressed the brake to the floor which prevented me from shortening the wheelbase of the Civic in front of me, but with 650 lb-ft of torque it was still accelerating. I was able to lift the accelerator with my foot and bring it back up.
Re:My sudden-acceleration Audi experience (Score:5, Interesting)
Virtually all (all that I have ever worked on) gasoline cars use the vacuum created in the intake plenum to operate the brake booster. Some cars use an electric vacuum motor to maintain the power breaks in designs where there is not a consistent vacuum or not accessibility to the plenum.
If you have a leak in your plenum or vacuum booster line, your engine should run rough and your breaks will be much harder too push.
The system will never prevent the application of the breaks, but it does mean instead of having power breaks, you are relying on the mechanical advantage of the pedal and your own leg power to stop the car. If you go back to the 60's you'll see "Power Brakes!" as an option you could add to your car.
Another option to shut down the cruise would be to put the car in neutral. If the engine continues to rev uncontrolled, it likely isn't the cruise control that is at fault.
I have experienced 3 sudden acceleration incidents. 1 was in my Fiero when the 15 year old Cruise Control vacuum got stuck (turning off the CC restore normal driving) and the 2 others, in a '87 Dodge Raider and an '06 Golf TDI we both due to floor mats not being properly installed. The velcro backing on the Dodge's mats had worn out, and the dealer threw in rubber mats on top of the stock mats in the Golf. In both cases the mats had crept forward enough to interfere with the gas pedal.
I'm not saying that there isn't a problem with any specific design, but in my personal experiences the faults have tended to center around pedal interference and/or aging mechanical devices.
-Rick
Re:Carmakers lie (Score:2, Interesting)
Re:Same thing happend to Audi a few years ago (Score:1, Interesting)
Many years ago I had a Mercedes 280SE 4.5 -- a wonderful sedan with their first injected V8 crammed into the engine compartment. The injector controller lived at the front right by the radiator. Normally the car ran perfectly, but sometimes if I was stopped at a light next to someone visibly talking on a CB, the engine would start missing badly. When the other vehical moved away the idle stabilized and I could pull away. Of course, the dealer could never find a problem -- but the circumstances seemed quite clear. My experience makes me wonder how many of these issues have a root cause in RFI?
Re:Carmakers lie (Score:3, Interesting)
You did know that consumer GPS doesn't have the real height in it, right?
Umm,you're not making sense. GPS gives you your position in 3 dimensions - height is an integral part of the calculation.
it uses an altimiter to guess at your height.
No... no it doesn't. None of my GPSes have a barometric altimeter in them. With a good constellation of at least 4 satellites you can get your 3D position with no external data (your GPS may use external data so as preexisting knowledge of your rough location to speed up the satellite acquisition, but that data is not actually required for the GPS to work). It is true that if you have a poor constellation (e.g. only 3 good satellites) then many GPSes start making assumptions about things like your altitude to fill in the missing data, but if you have a poor constellation then the DOP is going to be insanely bad anyway.
Many GPSes use doppler shift as well as measuring how your position changes over time to calculate your speed, so the accuracy of your position may not be that important either.
Expecting your GPS to be more accurate than your car is folly. your GPS is what? 3 feet? 5 feet? hmm pretty easy to get a 0 speed on that even when walking.
You are making a faulty assumption. You are assuming that if a GPS is "accurate to 2 metres" that you are likely to see a large positional shift between one sample and the next. However, this isn't the case - whilst the position may be 2 metres off, that error is going to be relatively stable over short periods of time. i.e. if you read your position and are told it is 2 metres west of your real location, when you read the position 1 second later it will still be roughly 2 metres west of your real location, it won't suddenly have jumped to 2 metres east. So despite moderate positional errors, the speed can be quite accurate, especially when combined with doppler shift to help filter out sampling errors.
Re:Same thing happend to Audi a few years ago (Score:1, Interesting)
I have a 2009 Highlander Hybrid. It happened to me last week. No floormats to get in the way. Accelerator pedal returned to "home: position. The event went like this: full acceleration by me to merge into traffic and complete release of accelerator. The acceleration continued at 100% for another full second. Over 150KW of power. Now I wonder what kind of computer fault would have happened if I had pressed on the brake to compensate for the uncontrolled acceleration.....
That is the exact same phenomenon I have experienced in my 2004 Matrix on a couple of occasions. Press accelerator to floor to pass traffic; lift foot back, but pedal remains on floor and acceleration continues for ~5 secs, then the accelerator pedal returns to normal position. On both occasions the weather had turned rather cold after a long period of warm weather which is what I talked myself into contributing to the cause.
Re:Put the damn thing in neutral! (Score:2, Interesting)
Those that drive manuals would be more clear headed because driving is an active activity where one is engaged with the machine. I don't mean to be a dick but people who drive manuals inherently understand how the fuck a car works because it is required to get the thing moving and stopped.
Interesting theory. That would mean that Germany, where most drivers have never even driven an automatic, should be full of expert drivers. Not so.