Unintended Acceleration - Find the Cause

Mr_Shiftright

So, to summarize, then, we don't know what components a "livelock" will and will not affect, and we don't know how it will affect them----because no one has ever been able to replicate such an event in those types of Toyota's involved in UA.

Is that a fair enough statement? Seems that way to me at any rate.

Affects throttle PLUS shifting but not steering or brakes or ignition.

HMMMMMM.....

How convenient for this "livelock" theory! Why, that's just the way it HAPPENED!

Why is this sounding like a South Sea Islander explaining the earth as riding on a turtle's back? I mean, it DOES offer an explanation but hey, c'mon....

A far better explanation of events is this, IMO:

"somehow, the throttle stuck open (floor mats, defective pedal module, driver error--we don't know) and the driver failed to put the car in neutral and failed to shut off the ignition, possibly because he was in a car strange to him and didn't know how, and was under great stress trying to stop the car and shut it down"

busiris

It would appear that a few posters here just can't seem to move beyond the idea that a CHP officer could NEVER have a lapse in judgement and make an incorrect decision...

Well, no one has more respect for our police officers, but I refuse to make gods out of them. Everyone makes mistakes.

They are human, and subject to the same foibles and limitations as the rest of us.

Few have more training in "life-or-death" issues as surgeons, but there are cases and cases in which the incorrect organ or limb was removed. How could that happen? (In most hospitals nowadays, they patient is advised to use a marker and clearly label the area/limb that is to have the procedure... a simple but effective way to reduce errors).

I doubt a single police officer has as much training as a fighter pilot, yet these same pilots have been know to fire on friendly troops, or fly their plane into the ground. How could that happen?

Approximately 50% of commercial airline crashes are due to pilot error. Yet, they go through rigorous training exercises and simulations to learn what to do in cases of emergencies. How could that happen?

Law enforcement officers have been know to accidently shoot themselves, as well as others, yet they should have the best training in firearms control and handling. How could that happen?

The answer to all of those questions is simple.

Even the most highly trained, best educated individual can make a mistake. If one wants to do a little "google-ing", you can find example upon example of every single instance listed above.

In fact, most "good samaritan" laws were passed so that health care professionals would be protected if they make an error in emergency treatment of an individual. No need for the laws if no errors in judgment are never made...

http://en.wikipedia.org/wiki/Good_Samaritan_law

If anyone wants to simply believe an officer couldn't possibly make an error in judgment while under pressure, be my guest.

Those of us living in the real world know better.

kernick

You're trying to make the case that an error that a fighter pilot might make , or a state trooper may fire at an innocent person - an event that happens in 1-2 seconds, is the same as an emergency where a person like Saylor has what 30 sec or 1 minute, as the same thing?

I agree that if a person has 1-2 sec, they can make a wrong choice, and that's the only choice they get. However if you give someone even myself 30 sec to try and control a vehicle that's WOT, you can bet I'll have tried 10 different things. The mat would be in the backseat, I'd have bent the shifter if need be into N, the key would have been tried, and the emergency brake would have been pulled.

Yes a person can make an error in judgment, but the odds of someone not trying multiple things to shutdown a runaway vehicle, are astronomical.

I doubt a single police officer has as much training as a fighter pilot, yet these same pilots have been know to fire on friendly troops,

Well since the pilot is moving at 500 - 600 mph, and it can be dark or cloudy, the pilot is not making visual confirmation. What's sending the signal from the troops, and on to the plane, but their GPS system. I would have to say they're not flawless, as most military hardware.

busiris

You're trying to make the case that an error that a fighter pilot might make , or a state trooper may fire at an innocent person - an event that happens in 1-2 seconds, is the same as an emergency where a person like Saylor has what 30 sec or 1 minute, as the same thing?

No, I'm not.

I'm simply demonstrating how even the best trained and educated human can make a mistake in a high-pressure situation.

Doctors and surgeons do it... there are tons of examples. CHP officers are not gods, nor are they perfect. They are human, and they sometimes make human mistakes.

And, contrary to your assertion, most pilot-induced commercial plane crashes don't happen instantly, but over an extended length of time.... certainly in the count of minutes, not seconds. Something goes wrong, and then the cascade effect begins, in which the pilots make poor and/or incorrect choices under the circumstances.

The real issue here is this...

Your claim (along with a couple of others here) rests solely on the contention that Saylors must have done everything perfectly, yet he was still unable to remedy the situation. Therefore, something approaching "sinister" must have been happening.

Of the evidence we DO have, we know there were floor mat issues. Another previous driver verified that point.

The evidence we DON'T have is what actions the driver did/did not take in the emergency situation.

Knowing that humans quite often make the incorrect choice in a high-pressure situation, your "theory" continues to ignore that vital fact.

Why?

Because your "theory" falls flat on its face once the assumption that Syalors couldn't possible have made a mistake is proven to be false.... And, it is a false assumption.

None of us know what Saylors ACTUALLY did, but I do feel he did everything he could think of doing to remedy the situation. Unfortunately, it wasn't enough. Nor do we know if the things he did were the correct things.

In fact, he may have frozen. He may have done nothing. Unless you have ESP or can read minds, you don't know any more about what was happening in that vehicle than any other person on the planet.

Now, you can make any claim you wish. And, the theory you put forth is a possibility. Its just not the most likely possibility, with the limited knowledge we have.

When you have to create a "layered" or "concentric circular" pattern to make square pegs fit round holes, eventually you hit a point where you enter the "absurdity" zone.

kernick

Doctors and surgeons do it... there are tons of examples. CHP officers are not gods, nor are they perfect. They are human, and they sometimes make human mistakes.

This is not the same sort of situation. It is usually due to wrong info. or reading the info. incorrectly in a calm situation prior to getting scrubbed. Maybe the doctor is having a coffee, talking to the nurse, and trying to read a chart.

And, contrary to your assertion, most pilot-induced commercial plane crashes don't happen instantly, but over an extended length of time.... certainly in the count of minutes, not seconds. Something goes wrong, and then the cascade effect begins, in which the pilots make poor and/or incorrect choices under the circumstances.

Yes usually there is a System problem to start with! Congratulations - you now acknowledge that technological systems like an ECM can fail! But the pilots don't just sit there frozen in panic. They will try many things to control the plane. Even if they make a critical error, I'm sure they're trying to conrol the plane in various ways before the crash.

The evidence we DON'T have is what actions the driver did/did not take in the emergency situation.

Well let's just guess the CHP officer didn't sit there "frozen" or that he tried to stop the vehicle by playing with the stereo or the power seats. :P

Your claim (along with a couple of others here) rests solely on the contention that Saylors must have done everything perfectly, yet he was still unable to remedy the situation.

No one claimed that he did everything perfectly. He may have "froze" the 1st 3 seconds. He may then have tried 1 or 2 things which didn't make sense. But during that 30 sec or 1 min. I'm sure he did several things that should have stopped that vehicle. If you give me 30 sec. I'm sure to include braking, pulling up on the accelerator pedal, at which point I would have found the mat wrapped around it, i would have tried the shifter even trying to throw it in reverse. So let's say the CHP officer did 50% correct. Any one of those "correct" things though should have brought the vehicle to a halt.

He didn't need to do 100% correct, he just needed to do 1 thing correctly. And since that vehicle didn't stop, we come to the conclusion that the correct things he did, did not function.

Knowing that humans quite often make the incorrect choice in a high-pressure situation, your "theory" continues to ignore that vital fact.

To support your theory, you can insisting that there was only 1 choice, and then a driver tries nothing else. That's foolish or bull-headed.

Because your "theory" falls flat on its face once the assumption that Syalors couldn't possible have made a mistake is proven to be false.... And, it is a false assumption.

Wrong again for the same reason. He could have made several mistakes, but he also probably tried several things that should have stopped the vehicle. Unless you want to argue we don't if he played with the radio, the rear defroster, and the p.mirrors during that time?

In fact, he may have frozen. He may have done nothing. Unless you have ESP or can read minds, you don't know any more about what was happening in that vehicle than any other person on the planet.

We know he didn't freeze. Wasn't there a phone call made? Your argument is now based on - because there is no data, and dead people don't talk, and those that survived can't be believed because studies show people to miss things, therefore mechanical/electrical UA can't happen. Huh? You're not a police detective, I hope?

"Mr. Detective I saw someone get murdered right here, last night."

Can you prove that?"

"No, I didn't have a camera. But there's blood on the sidewalk."

"That could be from anything. Most likely human error, some drunk cut his hand on a broken bottle. Happens all the time around here."

"But I saw the body get put in the backseat of a car."

"Do you have any data or photo?"

"No."

"You do know though that people take cabs around here all the time. I think you were distracted by some ladies-of-the-evening, and that what you really saw was a cab-driver helping a little old lady into a cab. I mean what's the odds of you seeing a murder right here, vs. someone getting in a cab. Unless you have something else, I'm off for a java."

kernick

Computer glitches, hardware failures and unexplained communication outages happen all the time. But when the affected systems control nuclear-armed missiles, it gets a little scary.

That's why Saturday's brief malfunction at F.E. Warren Air Force Base in Wyoming, disconnecting 50 intercontinental ballistic missiles of the 450-ICBM-strong U.S. arsenal

a squadron of Minuteman III ICBMs suddenly dropped into "LF Down" status, which signifies a complete cut in communication between the missiles and the Air Force personnel in the bunker. Even the intrusion alarms and warhead separation alarms were off-line.

An official tells The Atlantic a launch control center computer, or LCC, that maintains communications with the missiles began to record errors. Those error messages cascaded through the system and prompted missileers to take five LCCs off-line, leaving the missiles "in the dark." Four of the computers were eventually restarted, but that first suspect LCC remained off-line.

The official said Air Force engineers found a similar failure 12 years ago at Minot Air Force Base in North Dakota, which set off a similar cascading sequence. But as of late Tuesday afternoon, the draft of the engineers' initial findings hadn't concluded what happened, according to The Atlantic. Earlier reports of a power failure were declared wrong, while a breach of underground cables beneath the base was suspected.

http://www.aolnews.com/2010/10/27/missile-mishap-revives-alarm-over-nuclear-[non-permissible content removed]- nal/

So I would say here's MORE proof that unexpected, highly improbable errors can happen with any system. They may occur once and then once again 12 years later! I'd guess no one found the problem, or corrected it properly the 1st time?!

And since there are millions of cars on the road each being driven daily, the "highly improbable" possibility becomes "possible". Just like buying 5 million Powerball tickets each day, means you have a pretty good chance of winning in a few months.

Is UA prevalent? No. Is UA mainly human error - Yes. Is UA happening not due to human error or entrapped floor mats - yes. Every device and system in our society has been shown to have flaws, errors, and proven problems. Vehicles are not the exception!

We can argue each individual case as to whether it was human error or not. Regardless of what you decide, there is mountains of similar stories from all areas of technology.

kernick

http://www-ee.stanford.edu/~hellman/Breakthrough/book/chapters/borning.html

Computer systems can fail because of incorrect or incomplete system specifications, hardware failure, hardware design errors, software coding errors, software design errors, and human error such as incorrect equipment operation or maintenance. Particularly with complex, normally highly reliable systems, a failure may be caused by some unusual combination of problems from several of these categories.

Hardware failures are perhaps the most familiar cause of system failures, as in the June 1980 NORAD false alerts. Individual components can be made very reliable by strict quality control and testing, but in a large system it is unreasonable to expect that no component will ever fail, and other techniques that allow for individual component failures must be used. However, when one builds very complex systems - and a command and control system in its entirety is certainly an example of a complex system - one becomes less certain that one has anticipated all the possible failure modes, that all the assumptions about independence are correct.

Believe in the infallibility of your technology all you want; you'll sleep better at night. Consider that everyday we live, Russia and the U.S. has systems that have failed and have nearly led to nuclear war on several occassions. We were spared in 1983 by a Soviet named Petrov, who ignored the errors of his computer system, that told him to send the signal to for an all-out launch of Russian ICBM's. If these systems which are so highly tested and have multiple back-ups fail, then everything else is highly suspect.

Another potential cause of failure is a hardware design error. Again, the main source of problems is not the operation of the system under the usual, expected set of events, but its operation when unexpected events occur. For example, timing problems due to an unanticipated set of asynchronous events that seldom occur are particularly hard to find.

busiris

Well, that was quite a rambling reply that simply shows you continue to deny reality.

Here's some reality for you...

1. No one has implied that something didn't happen.... even a fault with the car's electronics. That has been repeated multiple times, yet you STILL don't seem to "get it". No one suggested that Saylors simply stomped on the gas like a madman.

2. The single largest factor in auto accidents is the driver...Yes, driver error.

3. Your refusal to even acknowledge driver error as a possibility, even thought its the number ONE cause of auto accidents clearly shows your bias against the most plausible cause (but possibly not THE cause) of the accident).

4. Again, we know what the front PASSENGER was doing at the time. He was making a phone call. We DO NOT KNOW exactly what Saylors was or was not doing.

5. Like any good conspiracy theorist, you continue to grab things that support your position, yet at the same time, discount anything that disagrees or tends to disprove it.

I'll spend no more time with you on the subject. Its simply a lost cause. To put it bluntly, you're like a broken record... continuously repeating the same track, over and over.

No accident investigator ever succeeded in solving a case by opening his investigation in which he automatically ruled out the most likely scenario, especially when there was clear evidence supporting a set of circumstances (stuck floor mat).

Just like those who wish to believe the 9/11 attacks were something far more sinister than actions of crazed religious fanatics, there will always be those who believe Saylors was much more likely driving Stephen King's Christine than he simply didn't do the right things to stop a vehicle with a stuck gas pedal... one that indeed could have been stopped.

BTW... Have you even read the official police report on the accident?

stever

Something went wrong with the car that resulted in the crash. I don't think the jury will buy 100% "driver error". My guess is that's one reason why Toyota settled.

I think the dealer will get hammered by the jury even if there's no consensus that the UA was caused by floor mats or something else.

That's some reality for you (and it helps the conversation move along if you'll avoid making such personal characterizations).

kernick

3. Your refusal to even acknowledge driver error as a possibility, even thought its the number ONE cause of auto accidents clearly shows your bias against the most plausible cause (but possibly not THE cause) of the accident).

Egregiously incorrect again. In my post #1920 I said: "Is UA prevalent? No. Is UA mainly human error - Yes. Is UA happening not due to human error or entrapped floor mats - yes."

Both human error and systemic errors are involved. That means - I agree most UA is driver error; but some can't be explained by human error.

I have no bias. I'm not going to make any $ from this, I haven't been involved in UA. I'm an engineer, and my training and experience tell me that all options are possibilities and can and do happen. Everything has a probability. With millions of vehicles on the road, making hundreds of trips per year, that is a lot of chances for failures to materialize. Billions and billions of spins-of-the-wheel, testing the complex software and hardware of hastily developed and lowest-cost procured systems.

wwest

Insofar as dealer liability is concerned it really doesn't matter if the floor mat was the causative factor or not.

The jury will find that the dealer was negligence due to the fact that the PROPER reaction to Barnard's incident was not done. Low level employee or not, the "dealer" had knowledge of an extremely dangerous situation and did not take corrective action.

That's also probably why Toyota didn't include a dealer "no sue" in their settlement.

IMMHO had Toyota thought the floor mat was most likely the problem then they wouldn't have settled until the suit against the dealer was resolved.

Jury finds that the floor mat was the most likely causative factor that gets Toyota off the hook completely.

There is one piece of evidence that seems to be being overlooked. Post crash one of the retraining hooks was found broken off and under the floor mat. That "might" indicate that the "dealer" had properly secured/restrained the floor mat as a result of Barnard's input.

stever

Jury finds that the floor mat was the most likely causative factor that gets Toyota off the hook completely.

Ok, but that could also whack a judgment by whatever the "contributory" negligence percentage is by Saylor (assuming that California uses that theory). Dealer fault on the mat, but driver should have checked it.

wwest

But IMO the jury would find that Saylor had no knowledge or experience of the problem at the time so it wasn't reasonable to expect him to check. Which is probably why Toyota went the (RED HERRING) RECALL route rather than any of the simpler choices. Ouickest way to get the word out and transfer at least some of the liability to owner/drivers, just in case.

My first experience with floor mat UA was over ten years ago and I was not checking for a floor mat problem even on rental cars until after the recent second incident.

kernick

If I'm on the jury I don't consider it to be reasonable that someone, especially in a rental situation, should have checked the mats, or any other condition. If your brakes fail on your car, and I'm on the jury, I'm certainly not going to take the position that you should check your brake-fluid-level before starting the vehicle everytime, and that you should do a series of checks in your driveway.

The emphasis is on: for the driver - what is reasonable and standard practice; for the designer/manufacturer - if the system fails, does it fails safely, and have an absolute "kill" switch.

wwest

"..and have an absolute "kill" switch..."

And one without a special/unique 3-second "password".

plekto

We've seen the term "live lock" and "hard crash" used here before to describe what's happening with these computer systems, but let's go into it a bit further, because some people perhaps don't understand the terms.

I like to use the term "hard crash".

In the early days of computing, there were "soft resets" which cleared the memory and restarted the OS and "hard resets" which cycled the physical power. This is where the term "hard crash" comes from. On your PC, you see the soft reset as the famous ctrl-alt-delete combination and in modern OSes as "logging out". The motherboard, drives, and so on all are still working. And this is fine most of the time. But sometimes that does't work. The machines get into a scenario where you have to shut it down.

Now consider where you typically see "Press three seconds to turn off".
A: Every modern PC power supply does this. The override when the machine hangs is to press the power for three seconds and it will turn off the power supply. This is actually a soft reset that's controlled by the motherboard and power supply.(usually routed through the motherboard but the IC is in the power supply itself)

I am 100% positive that the engineers at Toyota are using a modified ATX power supply setup for their computer system. That is, the "start" button on your vehicle is exactly like the "power" button on the front panel of your PC!. This simply is a disaster waiting to happen, as it means that the car is always on and that turning it "off" under normal circumstances actually just puts it into the equivalent of your PC's "sleep mode".

Cycling the power by pressing in for three seconds is the actual "power off".

But... and this is huge...

Modern PCs, when they do crash in such a way that a hard reset is required, have a known problem in that in about 1/20th of typical scenarios like that require you to physically cycle the power on the supply itself, the front panel button won't respond, either. Even if you hold it in for twenty seconds.

This is a "hard crash". Total lockup usually caused by a motherboard error and not actual "live lock" trying to execute code. This is actually WORSE than a "live lock" scenario.

If this type of scenario were to ever happen to a vehicle, it would cause the scenario we are seeing. Everything is still on, but it's frozen. The power button won't respond. But what if there's no power button at the back of the PC? You see this with some Dells, Apples, and similar small PCs. You have to yank the power plug at the wall as there is no 15 cent on/off button on the power supply itself.

You cant yank the power at the "wall" on these cars while they are moving, as that's the main positive connection to the battery that runs to the fuse panel - and that's now under the hood and not inside the car(!).
****

Believe it or not, the fatal flaw in the Toyota UA cases appears to be the start button and power supply to the computer, which they lifted (parts of) from a typical PC without understanding why that's a horrendously stupid idea.

And, yes, that's my official "entry" into this sham of a contest - as if Edmunds ever intended to pay a dime given the sheer idiocy and hoops required to jump through to satisfy them. But at least it's some sort of moral victory to know that I know why the UA incidents happened.

busiris

Something went wrong with the car that resulted in the crash. I don't think the jury will buy 100% "driver error". My guess is that's one reason why Toyota settled.

Again, I don't know of a single claim that even remotely suggested that something didn't happen to initiate the UA issue, so if possible, lets see if we can all move beyond that mischaracterization of the events.

I think the dealer will get hammered by the jury even if there's no consensus that the UA was caused by floor mats or something else.


I agree with you there. Its simply unreasonable to believe a normal driver would speed down the highway with his family on-board and crash his car. And, based on the information in the police report, there was clear evidence that a problem existed. IMO, case closed.... guilty as charged.

That's some reality for you (and it helps the conversation move along if you'll avoid making such personal characterizations).

If I came across as making a personal attack, it was not my intention.

However, I do find it a bit humorous that several of the comments made here are from those that have never even bothered to read the official police report on the Saylor's incident.

I call that a "ready....FIRE!.... aim" scenario...

Mr_Shiftright

Well you can't rightly say you "know", you can only offer a good guess.

Which....might be a first step toward problem-solving, but doesn't answer the question.

I'm not so sure a jury would be sympathetic, at least not financially...when they ask for the "proof" all they're going to get is a big....UH......

I'm sure Toyota would love for any lawsuits to be out of the way as soon as possible, especially ones getting lots of media attention.

As we saw with the Audi debacle, it doesn't take much to damage a company's reputation--evidence not required.

kernick

And how would a consumer have proof? Given what's on a vehicle do you think there could be proof, if the vehicle does not do it the next few times when a dealer's service department checks it out.

Maybe Toyota could install reinforced video cameras in all their vehicles that would retain the last hour of audio/video focused on the pedals, steering and gear selector?

Not many vehicle manufacturers spend the $$ to put in systems so when things do go wrong, that lawyers have an easier time proving the faults. Would you pay to build your own noose? Or since we're talking Japan - you're own sword to fall on.

Mr_Shiftright

that's the consumer's problem, if he's insisting that a defect caused the UA and the factory says "prove it". That's how automotive forensic experts make their money.

I mean, would you really just pay me if I insisted, for example, that the software your company sent me infected my computer and crashed it, or would you ask me for proof?

busiris

that's the consumer's problem, if he's insisting that a defect caused the UA and the factory says "prove it". That's how automotive forensic experts make their money.

Interesting comment...

Here in the upstate of SC, there was a case in which a 16 year old driver was at the wheel of a Ford Explorer (several folks on-board with her). She lost control of the vehicle, it flipped, and unfortunately, the 16 year old was paralyzed and another passenger killed.

The family sued Ford for millions, claiming the cruise control was defective and would not dis-engage (along with the usual, no brakes, can't shift to neutral,....). The plaintiffs had an "expert automotive witness" who testified about cruise control operations and potential failures, and the jury awarded millions to the family.

Of course, Ford appealed. In the appeals hearing, it was brought to light that the "expert witness" was anything BUT an expert on cruise control.

The appeals court threw out the award.

http://www.antony-anderson.com/cruise/9.5 SA Links/FORD.html


http://www.judicial.state.sc.us/opinions/HTMLFiles/SC/26786.htm
From the link...

On December 11, 1999, Watson was driving a 1995 Ford Explorer along with three other passengers including Patricia Carter. Shortly after entering Interstate 385, Watson lost control of the vehicle, which then veered off the left side of the interstate and rolled four times. Watson and Carter were ejected from the vehicle. Watson suffered severe injuries that rendered her quadriplegic; Carter died in the accident. Respondents filed a products liability suit against Ford, D&D Motors, Inc., and TRW Vehicle Safety Systems, Inc. alleging that the cruise control system and the seatbelts were defective and seeking actual and punitive damages.

At trial, Watson testified that when she entered the interstate, she promptly set the cruise control, but shortly thereafter, the Explorer began to suddenly accelerate. Watson testified that she reached down in an attempt to grasp the gas pedal, but was stopped by her seat belt and that she then pumped her brakes to no avail before crashing. Watson’s father testified that on two occasions prior to the accident, the Explorer suddenly accelerated while he was driving. As a result, he took the vehicle into D&D Motors, and the technicians determined that the new floor mats were upside-down and needed to be turned over.[1]

Respondents’ theory of the case was that the Explorer’s cruise control system was defective because it allowed electromagnetic interference (EMI) to affect the system. EMI is an unwanted disturbance caused by electromagnetic radiation that interferes with an electric circuit. To support this theory, Respondents presented Dr. Antony Anderson, an electrical engineer from Britain. Dr. Anderson testified as to his theory that EMI can interfere with the speed control component of a cruise control system and cause a vehicle to suddenly and uncontrollably accelerate. He concluded that on the day of the accident, EMI interfered with the Explorer’s cruise control system, which caused it to suddenly accelerate and resulted in the accident. Dr. Anderson further opined that Ford could have employed a feasible alternative design to prevent EMI. Specifically, he testified that Ford could have used “twisted pair wiring” in order to prevent EMI from passing between the wires and had Ford used the twisted pair wiring, the accident would not have occurred.

In addition to Dr. Anderson’s testimony, Respondents presented testimony from Bill Williams who was qualified as an expert on “cruise control diagnosis” as well as evidence from four witnesses who testified as to other similar incidents in which their Explorers suddenly accelerated without the driver’s input.

Ford argued that Dr. Anderson’s EMI theory was unreliable and lacked any scientific foundation, and to counter the theory, Ford presented their cruise control expert, Karl Passeger. Passeger testified that EMI signals have no effect on a cruise control system and that the system contains a watchdog feature that automatically checks for improper signals and resets the cruise control computer if it is not operating correctly. Additionally, Ford suggested that the floor mats could have caused the sudden acceleration as they had on previous occasions.

Sound familiar???

houdini1

Well, before I send you the check, please explain how this "hard crash" seems to be repairing itself over and over. We've probably all had computers die on us before and, once really fried, I have never had one come back to life.

Mr_Shiftright

Well there's a sad case of a vehicle already prone to flipping over, a very inexperienced driver, and a very real emergency situation.....this was a toxic recipe for disaster from the get-go.

Here again, the driver didn't know what to do....you don't "pump" ABS brakes, and you DO shut off the ignition pronto. Ducking down to pull up a gas pedal is not my idea of a good maneuver in this type of situation. I wonder if that's what did her in.

I'm rather amazed that Dad would have even let her drive the car given his past experiences with it. AND with a car full of people, no less.

Poor girl---with some basic training and/or a year's driving under her belt, she'd probably be okay today.

plekto

It's not the computer actually dying(Note that I didn't say it "died" or "bricked itself"). It's that whatever caused the computer to crash is also affecting the motherboard itself in such a way that it's also locked up and unable to even respond to the power button.

It happens about 5-10% of the time a computer crashes as near as I can tell, and you've certainly had this happen yourself where the computer is frozen and you have to physically unplug it or turn the power off. When you reboot it, everything is fine again.

But getting back to the cars There is no physical power switch and that means that the above scenario can happen where everything just simply freezes and there's no easy way to override it(like a physical key as an example). It's not the code. It's that the system is not dealing with crashes correctly and there's little or no way for the driver to easily solve it when it does happen.

Mr_Shiftright

But the ECM on a car is prisoner of the ignition switch, which is circuited between the battery and the engine management system. I don't think a PC is wired this way with its power button.

To compare to a PC, picture the ignition switch on the power cord of your "locked up" PC.

wwest

"..3. Your refusal to even acknowledge driver error as a possibility.."

In the Saylor case driver error WAS NOT a possibility, nor in the Barnard case, or Mrs. Smith, and certainly not even in the case of the person who drove to the dealer with the engine roaring madly. And not even Sikes cannot be considered driver error since the only defense against vehicle failure was driver intent.

Just because you lack knowledge and experience regarding the complexities, vagaries of real time computer programming, and the possibility of computer hardware faults does not give you license to make such broad, false, statements.

"...one that indeed could have been stopped..."

But then that statementof FACT, given that YOU admit there is no evidence one way or another that he didn't try everything in his power, is BS.

wwest

"..prisoner of teh ignition switch.."

Sorry, totally wrong, false.

PB start/stop first: Absent the computer being "live" 100% of the time the momentary PB switch closure would not be detected and the engine would not start. You do know that the only way to erase the learned parameters is to disconnect the battery...? The engine/transaxle computer is powered 100% of the time, 24/7.

Other than the possibility of it directly engaging the starter solenoid even the key type ignition switch no longer directly controls the engine "start/run" capability.

busiris

In the Saylor case driver error WAS NOT a possibility, nor in the Barnard case, or Mrs. Smith, and certainly not even in the case of the person who drove to the dealer with the engine roaring madly. And not even Sikes cannot be considered driver error since the only defense against vehicle failure was driver intent.


I'm sorry, but that statement is absolutely false. As long as the car is running and in motion, the driver has options.

Just because you lack knowledge and experience regarding the complexities, vagaries of real time computer programming, and the possibility of computer hardware faults does not give you license to make such broad, false, statements.

But I DO have that knowledge. I ran multiple IT shops during my working career in the telecom industry, so I DO have an understanding of what I am talking about.

wwest

"..No accident investigator ever succeeded in solving a case by opening his investigation in which he automatically ruled out the most likely scenario, especially when there was clear evidence supporting a set of circumstances(stuck floor mat)..."

Okay, let's take a serious look at the stuck floor mat as a possibillity.

I have personally encountered this situation twice in the past 10 years, the first involved a test drive in a new Mazda Miata with Manual transmission. The dealer had placed a rather thick cardboard floor mat in the car so as to protect the carpets for the eventual buyer. That cardboard floor mat had slid forward over the gas pedal but behind the brake pedal.

So every time I applied the brakes heavily the engine would roar to life. I pulled over about the second time this happened, discovered the cause, and tossed the troublesome floor mat out of the way. I did inform the dealer salesman of the happening on my return.

The second instance was about 3-4 years ago. Mazda, again, floor mat was not hooked and I ended up driving out half way into an intersection because the more I applied pressure to the brake pedal the more the engine fought back.

You can be sure the next stop, not 50 feet away, involved my "hooking" the floor mat.

So tell me, what might your action be if you were driving Saylor's Lexus and
encountered that initial UA...?

Would YOU pull back out onto the roadbed after such a harrowing, life-threatening (wife and daughter in the car) without finding out the initial UA cause and at least thinking you have resolved it to your satisfaction..?

I thought not.

wwest

"..I'm sorry, but that statement is absolutely false.."

That statement was made to undermine the FACT, your stated FACT, that in the cases stated driver error was a possibility.

"..driver has options.."

Yes, but which ones were tried but apparently didn't work...?

Was there even a guard rail to "rub" against to help bring the car to a stop...?

Sorry, IT is NOT a real time computer processing environment, not even close. If an IT computer "livelocks" and you just reboot and restart the process, unless it repeats reliably, in which case "you" call in the programmers.

"..I ran.."

You don't say you ever wrote or even had the responsibility to debug code.

Plus which if you truly understood real time computer coding, input polling or even interrupt driven, we wouldn't be having this prolonged adversarial discussion....unless you just like to argue.

kernick

that's the consumer's problem, if he's insisting that a defect caused the UA and the factory says "prove it".

My uncle who was a mechanic himself, had a car that was brand new, and the dealer didn't want to address the issue. He told the dealer that if he didn't "he'd show him what was wrong when he drove the car thru his front window". I would second the motion. I guess you're not a proponent of the Lemon law?

I mean, would you really just pay me if I insisted, for example, that the software your company sent me infected my computer and crashed it, or would you ask me for proof?

Just like tobacco companies denied their product caused cancer, and they said "you have no proof", for 30 years? It's just circumstantial evidence right?

You really seriously believe the junk you're saying? You think that the regular consumer should be able to diagnose computers and vehicle electronics to have to prove a company's products caused a problem? You think the average consumer is able to organize and then hire experts who will be able to examine the function of a company's products? The company is just going to turn over their code and designs to some experts who are being paid to sue them.

Were you a tobacco industry lobbyist?

kernick

We've probably all had computers die on us before and, once really fried, I have never had one come back to life.

I've never had a device physically "fry" either; a "hard crash" is paralysis of the hardware. No mouse, keyboard or power button on the PC works. It's not that rare - my prior PC that ran XP would have it happen every few months. My modem and router will do that occassionally too. Tech service told us unplug it for 30 sec and then plug it back in. It works great. I have a Dell laptop at work that is usually plugged in; I've had that crash before to the point where I've had to pull the battery out to get the system unlocked.

I'm sure if you want to read many people's examples you could google "computer unplug reboot" or some such. Here's 1 example to get you started: http://www.velocityreviews.com/forums/t218050-dell-desktop-freezing-have-to-unpl- ug-to-restart.html

kernick

I ran multiple IT shops during my working career in the telecom industry, so I DO have an understanding of what I am talking about.

IT !! About as worthless a bunch of screwups as Human Resources in companies! Make a career of insisting on unnecessary upgrades, create problems from the upgrade, then solve them, and repeat the process.

frankok1

I can't see how it can be a computer latch-up issue that causes the throttle to open. Both sensors in the pedal should not increase about near equally. Same goes for the input to the throttle sensors and valve control motor and power would be lost to the throttle. Perhaps engine stalling but I can't see UA. See videos on bottom of http://www.lexus.com/recall/

A glitch due to the code itself is another issue NASA could not find. I wonder if they really had complete access to the "crown jewel" - hopefully this group will have a chance:
http://www.wfmz.com/automotive/27000673/detail.html
Source Code Is Automaker's 'Crown Jewel'
Ultra-Secret Code Is The Heart Of Toyota's Throttle Control System
GILLIAN FLACCUS, Associated Press
excerpt
Selna indicated that he favored allowing the plaintiffs to access the source code through Toyota's server and not allowing the plaintiffs' analysts to have a copy on their own servers.

The judge also proposed allowing Toyota to alert the court and request an emergency hearing if it believed plaintiffs were accessing the code in an improper way.

Plaintiff attorneys hope to begin analyzing the source code for Toyota's electronic throttle control system in April and expect the process could take up to 10 months, plaintiffs' attorney Mark Robinson told the court. The plaintiffs will hire 10 engineers who will examine the source code in two shifts in a specially designated and secure room in Maryland, he said.

The room will be guarded and will have a surveillance camera and a screener at the door to monitor who comes and goes, and any material that is not shredded at the end of each day will be placed in a safe. Those wishing to enter the room will submit to scans of their irises and hands.

Mr_Shiftright

From my point of view, the case for driver error is increasingly obvious as each failure to show a fault is posted on the wall of the media.

The argument of last resort is always "cover-up". That says to me that the ammunition belts full of facts are spent and have not really hit their target, although they did cause a few flesh wounds.

So before the position is overrun entirely, if I were the defenders of Phantom Bug Hill, I'd retreat to safer ground. :P

busiris

Sorry, IT is NOT a real time computer processing environment, not even close. If an IT computer "livelocks" and you just reboot and restart the process, unless it repeats reliably, in which case "you" call in the programmers.

"..I ran.."

You don't say you ever wrote or even had the responsibility to debug code.

Plus which if you truly understood real time computer coding, input polling or even interrupt driven, we wouldn't be having this prolonged adversarial discussion....unless you just like to argue.

Now, there's a good example of a personal attack.

I guess I'll leave it up to you to define what other's expertise may or may not be.

Don't like the message?

Attack the messenger.

At least I take some comfort in the sense you didn't refer to me as Charlie Sheen.

LMAO!!!!

busiris

From my point of view, the case for driver error is increasingly obvious as each failure to show a fault is posted on the wall of the media.

The argument of last resort is always "cover-up". That says to me that the ammunition belts full of facts are spent and have not really hit their target, although they did cause a few flesh wounds.

So before the position is overrun entirely, if I were the defenders of Phantom Bug Hill, I'd retreat to safer ground.

I agree fully.... but, alas, I don't see it happening.

Mr_Shiftright

Well you have to give credit for some very imaginative explanations. I mean, I don't buy them, but I did find them very thought-provoking nonetheless.

I even added another epicycle to the multiple failure scenario---after the computer shut off the brakes, went to WOT and immobilized the gear shifter, it also activated the anti-theft system and thus defeated the ignition switch.

The only part I'm still working on is how everything comes back to normal and never happens again for anyone but the initial driver.

busiris

The only part I'm still working on is how everything comes back to normal and never happens again for anyone but the initial driver.

Perhaps we've been asking the wrong folks to examine this.

Maybe we should ask someone like Penn Gillette of Penn & Teller to give us a suitable explanation. Or, David Copperfield might let us in on the "illusion".

After all, at some point we'll get to the point where magic must be involved!

LOL!!!

Mr_Shiftright

No, magic is a cop-out here.

What would make a car do things for one person but not another?

wwest

"..I can't see how it could be a computer latch-up issue that causes the throttle to open.."

All you have to do to "see" is engage the cruise control and then hold it in the "accel" position. The throttle will go wide open until such time as you release the control. Or in the proposed case until the ECU stops executing the "livelock" that has it continuously executing the "accel" mode.

"Look at the source code" ...less than useless, a total waste of time. The object code, the computer's instruction execution definitions, and the computer's errata sheet, is where "it's at".

Assuming the specific computer is in wide enough use that a comprehensive errata sheet is available, extremely unlikely in the case of NipponDenso as the most likely, probable, computer chip supplier.

wwest

If this were a US company then our more than plentiful US attorney's would make "cover-up" very, maybe even extremely, unlikely. But with the japanese at the helm, not so certain, even doubtful.

"..facts are spent.."

Not facts, only suppositions based on circumstantial evidence.

wwest

"..At least I take some comfort in the sense you didn't refer to me as Charlie Sheen.."

You would consider my calling you Charlie Sheen as being to kind in comparison to what I did post about you. Stever deleted it so I had to edit and repost with "kinder" words.

wwest

"..after the computer shut off the brakes.."

Whoever gave you that bit of information is TOTALLY out to lunch, in NONE of the 5 cases is there an indication of brake "shut off".

"...immobilized the gear shift.."

I have proposed two scenarios under which that might be a possibility. CC "accel" livelock or extreme torque on the drum clutch gear teeth.

"...defeated the ignition switch.."

No, not SWITCH, start/stop PB. Barnard admits not knowing the "password" required to use the PB to stop the engine with the shifter not in park or neutral. So it is highly likely Saylor also didn't know the "password".

That's a horridly poor human factors engineering design and I'd bet good money a fix is already in the works. Probably requiring a 3 second PB depression for ANY engine stop command circumstance.

"..everything goes back to normal..."

Who said...??

plekto

The only part I'm still working on is how everything comes back to normal and never happens again for anyone but the initial driver.

It's terribly easy. In a typical "hard lock" scenario with a PC, it usually is a condition where the Southbridge goes nuts due to excess heat or an I/O conflict. Everything freezes and you're done. When you reboot, you're good to go unless that exact scenario happens again. Often just letting the chips cool down for a few seconds is enough to make them stable again.

Many times, it's just simply not repeatable as it's a combination of low voltage/poor power(dying or sub-standard power regulation capacitors are common on PCs) and heat problems combined. It has a brown-out in essence due to some error and while it would usually recover, every so often it just doesn't. On a car, a bad alternator or fuse block can cause wild voltage swings - enough to easily bring most electronics very close to failure or freezing up.

Overheating Southbridge controllers is still a common failure mode for many PCs as all too often they don't put a fan on the heat sink or a large enough one to begin with. If the machine gets into a tight execution loop (essentially live-locked temporarily), the CPU usage and heat can spike quickly and lead to this sort of crash as well. Your video card or CPU fan suddenly turns on full throttle and starts to whine like crazy while nothing appears to be happening on screen(very very slow graphics or mouse movement).

Remember, the Southbridge is the I/O controller for the peripherals connected to the main board. If that goes, all of the sub-systems freeze. That would explain no gear, no throttle, no power - it all just stops making its way to the CPU. This also will likely make the power reset stop working as well. From what I can tell, the system used for many new cars with drive-by-wire systems is identical to a PC in this way. Everything is essentially a fancy expansion card plugged into the main board - or at least from the I/O controller's viewpoint. So just like how your video and network card and mouse all go dead when the machine crashes, your throttle, brakes, and so on go "dead" as well.

This is why drive-by-wire is a bad idea unless there is a fail-safe redundant system in place as an override.

As for the "code" - the code is fine. 100% fine.

Note -this also isn't the signal path or electronics in an engineering sense. Everything is designed correctly. The code works. But who would have guessed that some tiny part that was sub-standard caused the main computer board to freeze up every so often?(this is the most logical conclusion, IMO)

Mr_Shiftright

Yes but all unproven and not even replicatable. These suppositions may not even be possible, much less having occurred.

It could be that these suppositions are no more valid than saying "and then the bullet stopped, changed its course 180 degrees, and struck the victim".

I mean, what's the point of going THAT far out to explain something?

I wouldn't call the ignition button bad engineering, since hundreds of thousands of people, perhaps millions, cope with them every day with no problems. And I've never seen one complaint on Edmunds, in magazines, in private engineering chat rooms, etc., that an ignition button ever failed to shut off a car.

It's not that any single one of your speculations aren't interesting---but it's the nearly mathematically impossible arrangement of them in combination, + their phantom behavior, I simply cannot accept. And cars are not PCs, they are cars. They are engineered to do different things.

the simplest explanations are often the correct ones. What is mysterious to us in problem solving is how hard it is to see the simplest solution sometimes.

Many discoveries in science for instance, are accompanied by multiple head-slaps around the world, as scientists say" "Of course, why didn't I see that!"

busiris

You would consider my calling you Charlie Sheen as being to kind in comparison to what I did post about you. Stever deleted it so I had to edit and repost with "kinder" words.

Personally, I would say you are taking this forum way too seriously.

Perhaps you think you'll find the "smoking gun" of UA by spending time here, or maybe you think you already have. If so, I wish you good luck.

My advice is to ...1- ease up a bit, and 2- adjust your attitude to have a little less "I'm the guy with all the right answers" feeling.

There are some pretty smart folks on these forums, and you ignore that at your own peril. Personal attacks on individuals are great examples of desperation, and I say that with my own personal experience.

One of my old "mentors" gave me some advice many years ago that I always try to remember...You don't have to be a Swiss watch-maker in order to understand how a watch operates...

busiris

You know, at the end of the day, if I were skeptical about the safety of the car I was driving and wished to trade it for some other vehicle, I would risk being killed on the way to the dealership by an intoxicated driver far more than being injured/killed by electronically induced UA.

Even if UA exist, the odds are so far in your favor that any one driver will never see it (or know anyone personally that will see it) in their lifetime.

I have often laughed to myself when seeing someone talking about the dangers of UA, while at the same time holding a lit cigarette in their hand.

We humans are a peculiar bunch, indeed...

Personally, I feel as if I have beat this "dead horse" enough.

wwest

"...I wouldn't call the ignition button bad engineering..."

You would if you understood.

The 3 second PB depression is required ONLY if you wish to stop the engine with the transmssion remaining in gear.

"....millions, cope with them every day with no problems.."

No, only for NORMAL PB start/stop use, stopping the engine AFTER we shift to park thus negating the 3 second rule.

So, let's jump to tens of millions and it would still be true that the need to use the 3 second rule would remain practically non-existant. From a human engineering standpoint it's like having to break out the owners manual and read up on the PB's special engine "off" circumstance if you ever have need to stop the engine in this special circumstance.

Idiocy, just.

And bye-the-bye, the 3 second rule does have "just" cause. You wouldn't want the engine to EVER stop unintentionally, say if you tapped the PB lightly, accidentally, with no intention of actually stopping the engine.

The design flaw is in not requiring that same 3 second rule for EVERY engine stop. That way the driver wouldn't have to go searching through the owners manual if/when they ever encountered the need to stop the engine in this special circumstance.

"...the simplest explanations are often the correct ones.."

First, I am not about to call you ignorant...

But you are quite ignorant, along with maybe >99% of the world's population, insofar as experience and knowledge of the complexities and vagaries of composing real time "process control" computer code, or even moreso DEBUGGING computer code that seems to go awry once every millenia.

I could tell you story after story....

"..Of course, why didn't I see that!.."

Oh my YES, after week upon week of attempting to replicate a customer's rare, horribly intermittent, software failure mode, finally getting there, and then discovering a rather simple coding bug along the lines of...

Ready, FIRE, aim...

The simplest explanation.

wwest

I'm not currently in the market for a new car, and likely will not be ever again. I do own one DBW vehicle but it has a clutch.

But you can bet good money that should I for some reason end up with a DBW vehicle and no clutch it would soon get a FAILSAFE modification to shut down the engine in a UA situation.

Open the SFI circuit(***) if the brakes are applied and the TPS doesn't indicate idle within mere seconds.

*** That's the factory method my '01 F/awd RX300 uses to dethrottle the engine for VSC or TDC activation.