Naked Science Forum
Non Life Sciences => Technology => Topic started by: Geezer on 28/01/2010 06:10:56
-
In the US, Toyota is having some very serious problems due to many reports of vehicles suddenly accelerating for no apparent reason.
At first they said this was because the accelerator pedals were getting stuck in the carpet floor mats. So they told everyone to remove the mats until the pedal was modified. That was a few months ago.
Today, they stopped further sales of many of their most popular models because of defective parts in the accelerator assembly that they say can cause the pedal to stay down without any foot pressure.
However, I've seen several reports where people said their foot was nowhere near the accelerator when the car suddenly accelerated for no apparent reason.
There have been several fatal crashes that were attributed to mysterious acceleration in Toyotas and Lexus vehicles.
To me, it just does not quite add up. Would it be cynical of me to imagine there could be a flaw in the engine management firmware, or that there is a defective transducer in the system? Would it be even more cynical of me to imagine that Toyota might already be aware of that, but they plan to quietly uprev the firmware to address this issue?
I once owned a VW Vanagon that had the complete opposite problem. A couple of miles after getting going, the engine would stutter and develop almost no power. Funny thing was, this would usually happen at the bottom of the same hill not far from our house. After some experimentation, I discovered a simple solution. Flick the ignition switch off then on again. Reboot the fuel injection computer! Problem gone.
Eventually I realized that the mechanical thermostat in the cooling system was not closing properly. I replaced it, and the problem never reappeared. The firmware was not able to handle a situation where the engine got colder after it had warmed up, but that's what was happening because of the defective thermostat.
-
There was a problem with Citroens a while ago, where parked cars (with park brake applied) would roll off.
The problem was caused by the park brake being on the front wheels. With the brakes hot from braking, the park brake was applied. When the brakes cooled, there was insufficient pressure on the park brake to hold the car if parked on an incline.
But this Toyota problem sounds odd. Since pressure has to be applied to the throttle to accelerate, I would guess you could be right in assuming a firmware problem. Or there may be an issue with the throttle return system not being strong enough, due to designing it to be too light.
A few years ago I had occasion to hire a 7.5 tonner and ended up with a DAF. When I put my foot on the clutch, expecting some resistance, I nearly put my foot through the floor. There was virtually no resistance at all. If this Toyota throttle has been similarly designed, it could result in problems.
-
I think this is to some extent an education problem, driving schools should drum into new drivers that the moment this happens they hit the brake pedal and the ignition switch,
I have only had this happen to me once when the accelerator pedal on my BMW broke, sad to say I did the wrong thing and put out the clutch fortunately the engine had a rev limiter and no damage was done and one second later I hit the ignition switch.
-
"There have been several fatal crashes that were attributed to mysterious acceleration in Toyotas and Lexus vehicles.
"
By whom were they so attributed? Presumably not by the victim. Who else was there?
-
"There have been several fatal crashes that were attributed to mysterious acceleration in Toyotas and Lexus vehicles.
"
By whom were they so attributed? Presumably not by the victim. Who else was there?
In at least one case that I'm aware of, actually by a victim. The driver called 911 to ask for help because his Lexus was doing 110 MPH and he could not do anything to slow it down. He, and the other occupants of the vehicle died a few minutes later.
-
I just heard a news item on the BBC World Service about the recall of ~6 million Toyotas in the U.S. Nothing about recalling them anywhere else though.
-
Ah, the unstoppable world of fly-by-wire.
Give me a direct cable link any day... saying that they're not immune to sticking.
-
Lee: That's interesting. I heard on the news here this morning that they were applying it in Europe and China too.
The latest story re. the US was that the problem is caused by a defective component that came from a US supplier. Initially there was comment (it may not have come from Toyota) that this only affected vehicles built in the US. Now, unless the news here is wrong, the problem seems to affect other regions too.
This looks like a slow reveal to me. How many more shoes are going to drop before they get to the bottom of this? IMHO, this bears the hallmarks of a gigantic cock up.
BTW, DD mentioned the Audi problems in the US. As I remember, that situation only happened when the vehicle was being parked or reversed. I always put it down to the offset in the pedal arrangement causing people to hit the gas when they thought they were hitting the brake. (That would have been unlikely in the UK because the offset would have the opposite effect.)
-
Ah, the unstoppable world of fly-by-wire.
Give me a direct cable link any day... saying that they're not immune to sticking.
Is the throttle positioned by a servo? I thought there was still a mechanical connection to the throttle flap. Although, the latest cruise control systems don't alter the accelerator pedal position, so there must be some sort of electronic override.
-
Here's a recent news report. Toyota maintains it's a sticky pedal problem. I hope they are right and there are no other problems.
http://news.yahoo.com/s/ap/20100129/ap_on_bi_ge/toyota_recall
-
Lee: That's interesting. I heard on the news here this morning that they were applying it in Europe and China too.
Yeah, just heard that it's up to ~8 million world-wide now.
-
After accounting for earlier recalls for floor mats and some vehicles being recalled more than once, the global recall amounts to more than 7 million vehicles, a staggering number.
Hidekai Homma, a Toyota official in Tokyo, said media reports giving that number as the equivalent of Toyota's annual vehicle sales are providing an exaggerated picture of the problem.
"We don't welcome this kind of thinking at all," he told The Associated Press. "No matter what, we believe we have responded speedily to a problem that has come up."
(Above from this link http://news.yahoo.com/s/ap/20100130/ap_on_bi_ge/toyota_recall)
Er, bollocks, Hidekai San. I'm sure Toyota does not welcome this kind of thinking at all, but some people think Toyota has been less than proactive in getting to the root cause of this problem. Reacting "speedily to a problem that has come up" (assuming this is a fair translation) suggests that Toyota customers are testing the product for Toyota. If that is the case, the customers should understand that they are conducting a Beta test, and get a substantial discount. Toyota should also bear the cost of any liability insurance.
-
Toyota are recalling around 220,000 cars in the UK for this problem.
http://news.bbc.co.uk/1/hi/business/8487984.stm (http://news.bbc.co.uk/1/hi/business/8487984.stm)
-
Ah, the unstoppable world of fly-by-wire.
Give me a direct cable link any day... saying that they're not immune to sticking.
Is the throttle positioned by a servo? I thought there was still a mechanical connection to the throttle flap. Although, the latest cruise control systems don't alter the accelerator pedal position, so there must be some sort of electronic override.
"The electronic throttle control—the digital fly-by-wire unit that links the gas pedal to the fuel delivery unit. Unlike some manufacturers such as Volkswagen and the other German carmakers, Toyota has yet to implement what's called a smart throttle. It's a software algorithm that allows the brakes to override the accelerator."
http://www.scientificamerican.com/article.cfm?id=what-might-cause-a-gas-pe
This article implies some electronic 'interference' between the pedal & the throttle. It is also weird that something as simple as overriding the 'gas' pedal with the brake has not been implemented.
Αnd in fairness, I don't suppose there's a lot of drivers who will be toe-and-healing anyway [:-\]
-
The plot thickens. Maybe it was a cost saving measure, or they didn't want to pay a royalty to someone for a patent.
Toyota has been trying to "spin" their way around this for so long. I'll bet the level of inter-departmental finger pointing within Toyota has reached epic proportions.
I know one thing, if my wife had one of these, she would refuse to drive it ever again, fix or no fix.
-
Steve Wozniak (co-founder of Apple) just went on the record and said his Prius does it, and it's nothing to do with the accelerator. He is convinced his car has a software problem.
He could not get anyone at Toyota or the NTSB to talk to him, so he went public.
-
He could not get anyone at Toyota or the NTSB to talk to him, so he went public.
Blimey!
I reckon it must be a conspiracy by US car manufacturers to discredit the Japanese & boost their sales - Only explanation!
-
I think you credit the US Auto industry with more intelligence than they have previously demonstrated.
The interview with Wosniak that I saw was interesting. He said he only needed to lightly touch the cruise control buttons on the steering wheel on his Prius and it would accelerate to top speed. Of course, this only happened intermittently. In his case, he said he could fix it by hitting the brake pedal. I have no reason to suspect Wosniak is making any of this up.
The symptoms he described suggest a lack of ESD immunity.
-
I agree DD. Liu is suggesting this is some kind of "black art". It's not, and there are many reasonable hardware and software defenses that must be incorporated into systems. Sometimes they add cost, but not much. Not only that, as I'm sure you are aware, there are many well known test methods that can be used to prove the effectiveness of these defenses.
It seems that, somewhere along the line, Toyota became extremely sloppy, and the news reports also suggest they may have become a little arrogant. They held a very enviable position based on quality and reliability but their reputation is now shot.
It takes years of hard work to build customer confidence, but it can be destroyed in a matter of days.
BTW - did you see the latest. Toyota owned up that they had fixed a brake problem in the latest version of the Prius, but they don't seem to think a retrofit is required on the previous models.
BS! It sounds a lot more like they planned to sneak the fix into the new model and let previous owners "roll the dice" with their lives, but they got caught. Now that, as Neil would say, is kwality.
-
It's really quite incredible to watch. The company seems to unravelling before our eyes, and nobody seems to be in charge of anything. It's like a crowd scene from a Shakespearean play. If somebody does not get a firm grip on the controls very soon, it's going to auger in, certainly in terms of the US market. You have to believe all the top US guys are updating their resumes. I suspect the guys in Japan think they can ride this one out!
It's a shame for all the workers in the US plants. They'll probably get hit hardest.
Talk about a casebook study of how not to run a business.
BTW - Perhaps you saw it, but there was an "expert" on GMA this morning who said the accelerator problem was BS. He was adamant they had RFI problems.
I suspect that's part of it, but I would think it's probably a combination of RFI and inadequate firmware defences. Even if RFI gets into a system, it should at least failsafe. There are all kinds of sanity checks - both hardware and software - that can be implemented to keep a system from doing stupid things. Of course, if the manager is telling the engineers there is no more time and it's "good enough", we know what tends to happen.
-
Do you mean about not recalling the Prius? If so, I completely agree.
The expert didn't work for Toyota. I hope I made that clear. There is so much stuff flying around on this subject already - I don't want to spread any misinformation.
I have no idea about the configuration. If there is a EMI/ESD problem (they are really variations on the same theme) it may have nothing directly to do with the accelerator input. There are lots of sneaky ways for EMI to screw up a system. Unfortunately, shielding does not help much because most of the problems are electromagnetic rather than electrostatic. Electrostatic discharges are a problem, not because they actually zap things, but because there are enormous rates of change of current flowing through ground wires etc. when the energy discharges. Those currents produce significant electromagnetic pulses that couple into any available conductors (including metal casework!) and they in turn re-radiate the energy. If you accidentally design a bunch of receiving antennas with large loop areas into your hardware, you are pretty much screwed!
However, assuming the they don't have EMI problems that are managing to mess up memory contents and CPUs, they can probably harden their systems with some firmware changes. But first you have to really understand the problem and then it takes time to come up with the right fix. If you are not really careful, it's quite easy to fix one problem and create a load of new ones.
Just thinking about this is making me nervous. I used to try to manage stuff like this for a living. The money was good, but the stress was horrible. Weekends were no fun because there was usually at least one crisis hanging over the engineering organization. Fortunately, I was able to retire before it did me in!
-
Just a thought, does anyone use optics for this type of transmission?
That would rule out all noise issues would it not?
The other query is, as all modern autos use CAN-bus (or similar) to save on copper wiring-looms, does the pedal position signal work though the same interface?
-
Yes, fiber optics or optical isolators is a good way to block EMI. Don't know how much the car manufacturers use them. Also don't know if they take the accelerator over the bus or route it straight to the engine management computer(s).
Anybody?
-
Surprise!
"WASHINGTON (AFP) – US investigators doubt Toyota fully understands its accelerator problems and are even less sure that new floor mats are an adequate solution, said a memo to lawmakers made available Tuesday."
"There appears to be a growing body of evidence that neither Toyota nor NHTSA have identified all the causes of sudden unintended acceleration in Toyota vehicles," said the memo"
http://news.yahoo.com/s/afp/20100209/ts_alt_afp/japanusautorecallcompanytoyotaprobe