Given the media hype, people who had just a throttle problem with a manual transmission (even if they simply dealt with it a and went on with their day) would have still claimed the incident as happening. A handful would have, even based upon human nature. Not zero. Zeros are really bad in math (and doubly so in statistics) as they show you that there is a basic flaw in your hypothesis.
20% of the fleet is producing zero statistical data. That's beyond unlikely if it's just human error alone.
I also strive for perfection in my engineering job knowing you may not get there. That is especially the case when a rare event such as a coupling such as between a turbine and generator could fail and kill someone. But human safety was mitigated by Toyota in many ways: delaying the floor mat issue, delaying the black box information, no diagnostic fault code for UA, AND no brake override conversions much earlier.
The Lexus model in the Saylor case had one of the worst rate of incidents. Bet you wouldn't trade your Beemer in for the same Lexus model that Saylor's family died in even with proper mats, a repaired CTS sticky pedal if it has CTS instead of Denso (sticky pedals that CTS claims never caused any of the serious crashes), the same stupid 3-second shutoff push button, weird gear shifter, AND without brake override. After you trade it in and drive it 100000 miles with your family in tow, give another assessment. You would know about the 3-second need for the button so you would have an advantage over Mr. Saylor and his brother-in-law who made the 911 call.
I believe you are either a shrill for Toyota or have your head in the sand. It is just not the deaths but the injuries and damage. We will see after the trials - not just the biggie in CA, but the six or so by insurance companies and the one by the Saylor relatives against the dealer (who took issue with the entire floor mat reports including the previous driver.)
Note I am a conservative but take issue with the Republicans in some of the car safety issues, and am for tort reform for lawyers.
IMMHO the floor mat issue did not warrant a recall. In "normal" times I believe it would have been handled via publisizing the problem. Toyota suddenly found themselves behind the eight ball regarding the UA and needed to to show the public "some" action.
At the time Toyota probably had not a clue as to the actual UA cause and maybe don't even today. The recall over the floor mat issue was simply a smoke screen to convince the public that something was being done.
The CTS pedal recall and the shortening of the pedal pad more of the same.
We will see after the trials - not just the biggie in CA, but the six or so by insurance companies and the one by the Saylor relatives against the dealer (who took issue with the entire floor mat reports including the previous driver.)
Evidently, you are a big believer in the "infallible" criminal justice system.
Go ask the folks freed from life imprisonment/death row by DNA evidence how infallible that system is.
Or, read up on how Charlie Chaplin (silent film star) was convicted by a court in a paternity suit, even though the child'd blood type made it impossible for him to be the father... a fact known at the time.
But that "something" would, by necessaty, have to be mechanical.
So now to have a "UA" you would need to have a runaway engine and a clutch mechanical failure simultaneously.
It was 100% possible with a shorted-out cruise control to have a runaway condition with a manual transmission.(Audi and other past cases) Not that this is happening, but a stuck throttle is possible in such a case. An out-of control throttle condition would have been reported as such by someone by now. "My engine also is racing out of control"(happens to have a manual transmission that's at full throttle in 1st)
My guess is that there's a problem aside from the code with the computer control system (probably bad error handling/crash recovery), since there are a few options like stability control and so on that aren't added to the manual vehicles. And also different computer code I'm sure. Oh, and a different wiring harness as well on most models. Got to check that as well - might be a short somewhere/something different if a short happens.
(and on it goes - they haven't even gotten to testing all of the possible failure modes)
Sorry, I don't see how the Audi UA episode applies.
Just what/which driver, knowingly operating a clutch equipped vehicle, would not instantly resort to the use of that clutch even if the right foot was inadvertently on the gas pedal instead of the brake..?
True. But they would still likely file a report or claim it happened. But as a safety issue, manual transmissions do have many redundant overrides that automatics lack
But to some specific problems I've seen with sensors, that have otherwise been working fine and no problem is expected - 1) machine vibration eventually loosens a screw, allowing the sensor to move out of alignment. 2) a piece of fuzz or dust lands on an optical sensor knocking it out, 3) power-failure or surge causing a processor to need to reboot
So, it sounds like, at least in the examples you gave, the cause of the problem is known but nothing has been done to correct it?
If this was part of a mission critical system, then I would: 1. Wasn't the vibration environment known ahead of time? Safety wire or stake the screws so that they cannot vibrate loose. This is SOP on things like satellite systems and avionics. 2. Provide protection for the sensor or even provide a redundant sensor so that a piece of lint cannot knock the who system off line. 3. Require that the power supplies be able to perform properly (within specs) for defined surges and spikes on the input line.
to read the owner's manual to learn how to shut off the car you're driving.
I wouldn't hesitate to drive any Lexus, even Mr. Saylor's, because I know how to stop it.
If a person doesn't have this basic level of competence to push a button or put a car in neutral, then they need to drive something that robotically does it for them I guess.
Better for them, better for the rest of us.
As for science, medieval scientists aren't 'scientists' (the word didn't even exist then), nor are politicians.
REAL scientists spend all their time attacking their own theories. That's what makes science so much more interesting to me than faith, conjecture, politics, etc.
Sure, sceintists can be fallible and even political, but that certainly doesn't trash the actual science.
Whether you believe in gravity or not, or the person espousing it, you still go around the sun once a year.
REAL scientists spend all their time attacking their own theories. That's what makes science so much more interesting to me than faith, conjecture, politics, etc
Exactly! Science is to a large degree self correcting. Peers reviews of results, and being able to duplicate an experiment and get the same results are another example. Sure, sometimes some looney idea pops up - like cold fusion or the link between vaccines and autism. But those are eventually discredited.
C) Was the firmware code reviewed at the high level language, compiler level, or the actual machine code, assembly language level..? If not the latter your work is useless.
You're kidding, right? You really think that examining the assembly language or machine (binary) code compiled from a 200,000+ SLOC flight control program is going to yield any insights into the correctness of the code?
D) If at the assembly language level, assuming, as would be expected, that the processor was unique and not in widespread use, how did you go about locating the maybe one or two reasonably competent software engineers available worldwide..?
What? You only need one or two SW engineers to analyze the hundreds of thousands (millions?) of assembly language lines of code that is probably in the ECU's computer?
You really do appear to have your head in the sand...
I'm not in the market to purchase a new car at the moment, and might not ever be again, but you can be certain I would not risk my life in a new Toyota or any new DBW vehicle. The first thing I would do if I had to purchase today is add a true, FAILSAFE, brake over-ride.
The point I was trying to make was that it is highly unlikely that NASA even knew at what level the code needed to be analyzed.
"..only need two.."
Again, my point was that there is maybe ONLY two programmers worldwide that would have the knowhow to analyze code unique to a narrow market use NipponDenso processor.
But the answer is yes, I think a focussed look (starting with driving the throttle full open and working backwards) at the code could yeild positive results.
You really do have more to fear from a shark attack than a runaway Toyota. C'mon, you know about risk assessment !!
In short, what *appears* to be most risky is often, statistically, far less risky than the ordinary risks you take every day.
People fear flying more than driving but driving is pure carnage compared to airplane fatalities on major airlines.
Given the NASA analysis of fatalities from UAs, it seems like more people were killed by vending machines than runaway Toyotas. (People bang and pull and shake them in a rage for having been cheated and, being top heavy, the machines fall on the individual)
So, it sounds like, at least in the examples you gave, the cause of the problem is known but nothing has been done to correct it?
Sure if the process is something very important. If you go into any typical factory, say making toilet paper or potato chips, you will not find the equipment having redundant sensors unless there is a serious safety hazard involved. There may be redundant sensors on heaters, but the sensors for the packaging and labelling sections would typically not be redundant. Why because the people who design equipment aren't going to get a sale if their equipment takes up a lot more room and costs more because everything is redundant. And so it goes with autos. Do you have 2 TPM's per tire? Do you have 2 thermostats in the cooling system? ...
If you think it's just a matter of $ and putting in redundant equipment and you then will never have a problem, consider that in the last few days, the highly engineered, expensive high-tech high-speed rtrains have had 2 accidents. I believe one was an Acela train leaving NYC, and the other was a head-on crash in Germany (?). Here you can read some history on the vaunted German engineering and their success with eliminating equipment failure and risk.
I'll trade you my medieval scientists with their data with the Earth at the center of the universe + the (bought) scientists who put out the "smoking is good for humans" science + our government's "WMD in Iraq" scientists, for your MMGW scientists who are at the trough for $20B+ / year.
I dunno - many of those guys were under the thumb of their royal sponsors. Even Newton managed to get a patronage appointment to run the mint and wound up causing a lot of silver coinage to be melted down and shipped out of Britain. Not the most brilliant move (maybe it was the mercury in his system that did it).
If you think it's just a matter of $ and putting in redundant equipment and you then will never have a problem, consider that in the last few days, the highly engineered, expensive high-tech high-speed rtrains have had 2 accidents.
No, $$$ and redundancy can help. But understanding what might cause failures or faulty operation is probably more important.
There are satellite systems I've worked on that were designed for a 5 year life on orbit, but are on going into their 15th year of operation.
There's the Voyager spacecraft, launched what, 30+ years ago, still operating (somewhat, anyway).
There's the Martian Rovers, both of which greatly exceeded their 3 month design life on the surface.
A lot of these successes had to do with understanding the risks, recognizing where the hazards might be (radiation,for instance), and taking the appropriate engineering measure to ameliorate those issues.
I dunno - many of those guys were under the thumb of their royal sponsors.
Yes, and now we've replaced our royalty with corporate and political Thumbs. Either way you have a powerful central authority which controls the $$, the rules, and even more so now - the ability to market their agenda thru the media. You should not underestimate the enhanced power that anything labelled "official" carries when presented in the media. For example when you see a TV news article that the police/government has arrested someone for murder, what is your first thought. "That must be true"? Why? Because you're seeing it on TV and you're conditioned to believe authority/officials.
Go back and consider the whole period where the government presented data and evidence thru different spokesmen, night after night on the news to convince the Congress and public that there were all sorts of WMD in Iraq. Remember all the pictures of the launchers and the biolabs from our satellites being displayed each night? What a great case they made! Create your own news! Snicker at anyone who disagrees with that story, as an obvious nut! They showed you the data! Didn't they?! Believe - Big Brother.
No, $$$ and redundancy can help. But understanding what might cause failures or faulty operation is probably more important.
Sure. I've made matrices where we plotted and ranked risks vs. probability of occurrence and consequences. And what happens is that the reality of the cost to address each issue, and the need to launch new product quickly, means that not all risks will be tested or addressed.
If you've been working on satellite systems or other aerospace systems that is a different mentality than launching a new vehicle, or any other commercial item. The design cycle of any commercial product has been compressed, pressure is to keep the project rolling, minimize delays and testing,launch the redesign, get to market before the competition. Keep the cost down.
Because there is so much competition and margins are fairly low, you are going to get vehicle designs rushed and testing will be at a minimum. I think most auto engineers do a great job, given the conditions they work under, but they are being pushed by businessmen. Do more with less each year! (I went to grad school with a bunch of Sylvania engineers who supplied the Big 3 with lights, who used that motto).
You probably don't have that sort of environment when working on government projects, though I'm sure it has it's stresses, because there would be more upfront failure-modek work done.
REAL scientists spend all their time attacking their own theories. That's what makes science so much more interesting to me than faith, conjecture, politics, etc.
Except for the MMGW "scientists". They spend their time keeping info. away from their "peers", hiding info, and making exaggerated claims, etc.
No, I'm not a shill for Toyota--personally I'd rather drink ammonia than drive a Toyota every day. But that's just me, and it has nothing to do with safety, but more with driving characteristics.
re: Science -- the neat thing about science is that it's true whether we believe it or not.
If anything, the tragedy of the WMD myth points out quite clearly, in relation to UA, that it is the persistent belief that something that shows no evidence of being there, really is there, somewhere, somehow--- that is the stumbling point.
You probably don't have that sort of environment when working on government projects, though I'm sure it has it's stresses, because there would be more upfront failure-modek work done.
You're right. It's a different environment, but with it's own set of stresses.
Thank God NASA's NESC division did a better job finding the cause of the Challenger explosions - those mechanical problems are easier than those fast electron gremlins.
Sorry - I apologize - no excuse for being a fit of anger as I believe the NASA/NESC team did a bad job - as compared to the space shuttle explosion. I should have taken it out on them - and I will as we pay them. See the full report: http://www.nhtsa.gov/staticfiles/nvs/pdf/NHTSA-UA_report.pdf
"On Jan. 31, Intel executives announced that they had discovered a design flaw in the chipset the week before, and had subsequently found a fix for the problem and were beginning to manufacture new chipsets with the problem corrected. They also had stopped shipments of the problem chipsets, about 8 million of which had been shipped and some of which had been put into systems that had been bought by end-users."
So let's compare this Intel issue to NASA's review of Toyota's UA. Intel probably has some pretty bright people working there right; equal to NASA? And I bet on the design of this chipset Intel had more engineers than NASA, as for NASA this was only a side-job. And I bet each engineer at Intel put more time into the chip project than NASA put into looking at UA. And the people at Intel had the detailed history and knowledge of how everything should work, whereas the folks at NASA were new to the UA-issue when they started.
Yet apparently through the whole design phase, and testing, and well into the manufacturing of the chips, Intel had no idea this flaw was there. Couldn't find it!
Now I'm sure we can find other examples. We could look into Microsoft's history of spending millions of manhours on a new Windows version and overlooking the many flaws they have to patch. Or we can look at nay other software company and the flaws and patches we know occur.
And after ALL these examples, you can sit there and say that because some folks at NASA looked at, that the problem doesn't exist. That just seems like an incredibly illogical conclusion.
Plus as I said before, how do we know if the computer source code supplied to NASA hadn't already been fixed..? Lots of time elapsed between when Toyota would have, SHOULD HAVE, began looking for a coding flaw and when the source code was supplied to NASA.
And the source code probably wasn't even under Toyota's control, more likely than otherwise it was NipponDenso employees who wrote the code.
What seems illogical is asking Toyota to solve a problem that all the experts agree never existed....except for a few misplaced floor mats, a few panicked drivers, and several would be scam artists.
Well I can see you don't consider our modern technology full of flaws and design problems, due to the rush to market products to consumers.
The move to making everything electronic and eventually to having the car "drive itself" is going to be an increasing issue when accidents do happen. Toyota and the other manufacturers might want to consider this as one of the first in many legal problems they're going to have if they continue on this course.
The onus will be on these auto manufacturers and the suppliers of the code to prove that somehow they do not have the design flaws and errors that seem to plague every other product and company. Intel, MS, a lot of companies in aerospace all seem to have hardware and software issues. Recalls, patches, and missiles that miss the targets. Example after example that are public; I'm sure there are several times that many which are never made public!
But auto manufacturers have no electronic problems; they themselves have had their experts and paid experts bless them. And the bright guys over at NASA looked too. There you go. Case closed? Conclusion - there's no proof there's a problem, though every other company and industry has oodles of software and hardware problems.
This is still chasing phantoms....not only has no one ever shown a defect in Toyota's electronics, but they can't even come up with a way HOW these electronics could defeat multiple systems simultaneously and THEN return to normal and THEN never be replicated.
You remember how I talked about the theory of epicycles being used by desperate astronomers in times past, to try and patch up their speculations---and as nothing in their speculations actually made sense compared to observable phenomenon, they kept on coming up with even more elaborate explanations. Soon the epicycles (mini-orbits) grew so complex they couldn't even draw them anymore.
What is interesting to me is that most problems, even in complex machines, are, when finally discovered, not that complicated at all.
Why did Challenger explode? (a machine far more complex than a Prius)_
Why did the Comet, the first jetliner, crash?
Why did Apollo X burn up on the launch pad during testing?
Why do submarines sink?
Inevitably, they trace it to issues, that, once discovered, seem painfully obvious.
It should be, that if your Prius does an UA, you should be able to lend it to me for a few weeks and it should UA on me as well.
If it doesn't happen for me, or for anyone else but you, then what's the answer?
This is still chasing phantoms....not only has no one ever shown a defect in Toyota's electronics, but they can't even come up with a way HOW these electronics could defeat multiple systems simultaneously and THEN return to normal and THEN never be replicated. ****
Sure you can. Crash the machine on purpose. 99.9% of the time, rebooting a computer like they use in cars and other electronics "fixes" it, and unless there is an entirely separate computer system monitoring the first, it will have no idea what happened. The same thing happens with software. Say, I'm using NERO to burn a DVD. And it crashes. Which it is prone to doing. Force-quitting it and restarting always results in a clean slate. Now, true, some applications do have crash logs, but does Toyota for all of the components and sub-components in the chain? From what I gather of OBDII type systems, they are crude 1990s technology at best.
I had a Buick LeSabre years ago with ODBII on it and it came back with a cryptic "do not run" message. Not one code. Engine was dead as a rock. The actual problem was that the timing cog had ground off its teeth and the valves were randomly operating as a result.
No codes, no record, nothing at all. It didn't even record the "do not run" message, either. Yeah, trust me - if you haven't found anything wrong with a car that's exhibiting odd symptoms, the computers likely haven't caught it either.
It should be, that if your Prius does an UA, you should be able to lend it to me for a few weeks and it should UA on me as well.
Why do you think that? The reports I've read indicate 3 things: 1) these vehicles have been on the road for a number of months to years without a problem. and 2) it is a small number of vehicles affected, and 3) the problem does not necessarily continue to happen each and everytime after the 1st noted incident (for those that didn't crash).
For all we know, UA may only happen when a certain series of events happen, such as traction control engages, or stability control while at full throttle. If you don't drive such to trigger this or that condition the vehicle may run fine. As others said it may be some external factor as part of the equation, which you would not drive by.
So you driving it for a few weeks in your manner, and in your environment would only lead you to the false conclusion that your test represents the greater body of possibilities. It doesn't. Your testing would only eliminate a fraction of the possibilities. That's no different than digging holes in the desert, and after your 3rd hole declaring that the other 99% of the desert didn't contain any ruins.
Personal experience: I had a '92 Dodge Spirit RT - remember that special version. I bought it used at a good price, because it was a Chrysler "buy-back" or lemon. I took a chance and they gave me a 1 year unlimited warranty. So I had it nearly a year when the 1 day I went to leave work and nothing - no starter crank, zilch. I had it towed to the Chrysler dealer who the next day looked at it, and it started right up. The mechanics looked for any codes. Zilch. I believe that happened again a month or so later. I traded it in on a new Subaru.
I mean, now really, if my car doesn't start for me, and I pass it around to you, and others, and it's fine for weeks or months, and the minute I get it back, it doesn't start, and then I pass it back to others, and again, it's fine, and then I get it back and it happens to me----now really, wouldn't you kind sorta be thinkin', "you know, I think the problem is Shifty, not the car".
I could hardly think of a *better* test of UA then to take the problem car and give it to other drivers.
Either way you have a powerful central authority which controls the $$, the rules, and even more so now - the ability to market their agenda thru the media.
I'll see your conspiracy theory and raise you one Facebook giving Mubarak heck.
Noncarpetmntis. Feinman is playing bongos somewhere and questioning why the pearly gates squeak.
You are assuming that what you do when driving is the exact same thing as I do when I'm driving. What if the problem is only triggered when you floor the gas-pedal and activate the traction control, and your driving habit is to baby the accelerator. What if the problem is triggered when too many accessories are run in the vehicle, the AC kicks on, and the ECU temporarily is denied power for a split-second; and you never run many accessories? What if the problem is triggered when you drive under a very high voltage power-line, and you don't? or some combination of the above.
You driving the vehicle does not even sample these conditions/factors. You are thinking very limitedly that the problem is independent of the operator's actions; other than your belief that the operators are stupid drivers. If the vehicle is 100% of the driving equation, then you should be able to win the F1 championship next year if you get the right car!
I don't really see things as conspiracy theories. I believe power and $$$ corrupt otherwise normal, good people. When billions of $$ are at stake, or peoples' or corporate reputations on the line; people will certainly be tempted to act in a borderline ethical or criminal manner. Do you disagree with that?
As wwest pointed out also, the vehicles and code that NASA looked at could have been fixed. Would someone do that for the billions in profit that ride on the outcome? I could see someone arguing that you can't change the past, and what good comes of additional punishment of Toyota for what was a mistake, that they now fixed?
Okay, let me understand you. You're saying that if your car has an UA incident, and then you give it to me to drive for a month, and then I give it to houdini1, and we have no incidents, (we'll buy the gas :P )and then we give it back to you, and you have another UA, and then you give it back to me, and me to houdini, again negative, then back to you and you have ANOTHER UA incident----are you saying that doesn't prove something?
The reason I am reducing the argument to absurdity is because this seems to be a recurring thread in the UA incidents, to wit, no one can reproduce the incident except the victim.
Yes it would prove that the problem is a function of the what the operator is doing to the machine; or a difference in the environment we're running the machine in.
If you and Houdini buy a Weber gas grill and have trouble lighting it and then run it such to set your house on fire, does that mean if I buy a Weber gas grill it will do the same? Of course the operator and conditions of how any machine is operated has an impact on how it functions.
If all of us have the same exact software and hardware configuration on our PC's, what we do, what we're running, has a tremendous impact on the probability of either one of us having the system crash.
Comments
20% of the fleet is producing zero statistical data. That's beyond unlikely if it's just human error alone.
So now to have a "UA" you would need to have a runaway engine and a clutch mechanical failure simultaneously.
Who would the "people" complain to other than the dealer service personnel...?
And if no one was hurt, and why would they be, why would the media be involved, care...?
You won't ever.
All cars have a gear named "NEUTRAL" which WILL terminate any UA event.
The Lexus model in the Saylor case had one of the worst rate of incidents. Bet you wouldn't trade your Beemer in for the same Lexus model that Saylor's family died in even with proper mats, a repaired CTS sticky pedal if it has CTS instead of Denso (sticky pedals that CTS claims never caused any of the serious crashes), the same stupid 3-second shutoff push button, weird gear shifter, AND without brake override. After you trade it in and drive it 100000 miles with your family in tow, give another assessment. You would know about the 3-second need for the button so you would have an advantage over Mr. Saylor and his brother-in-law who made the 911 call.
I believe you are either a shrill for Toyota or have your head in the sand. It is just not the deaths but the injuries and damage. We will see after the trials - not just the biggie in CA, but the six or so by insurance companies and the one by the Saylor relatives against the dealer (who took issue with the entire floor mat reports including the previous driver.)
Note I am a conservative but take issue with the Republicans in some of the car safety issues, and am for tort reform for lawyers.
At the time Toyota probably had not a clue as to the actual UA cause and maybe don't even today. The recall over the floor mat issue was simply a smoke screen to convince the public that something was being done.
The CTS pedal recall and the shortening of the pedal pad more of the same.
Evidently, you are a big believer in the "infallible" criminal justice system.
Go ask the folks freed from life imprisonment/death row by DNA evidence how infallible that system is.
Or, read up on how Charlie Chaplin (silent film star) was convicted by a court in a paternity suit, even though the child'd blood type made it impossible for him to be the father... a fact known at the time.
http://betagenetics.com/paternity-blog/charlie-chaplins-not-so-silent-paternity-- case.html
Just because someone happens to have an opinion (based upon the evidence) that disagrees with yours hardly makes them a "shrill" for anyone...
So now to have a "UA" you would need to have a runaway engine and a clutch mechanical failure simultaneously.
It was 100% possible with a shorted-out cruise control to have a runaway condition with a manual transmission.(Audi and other past cases) Not that this is happening, but a stuck throttle is possible in such a case. An out-of control throttle condition would have been reported as such by someone by now. "My engine also is racing out of control"(happens to have a manual transmission that's at full throttle in 1st)
My guess is that there's a problem aside from the code with the computer control system (probably bad error handling/crash recovery), since there are a few options like stability control and so on that aren't added to the manual vehicles. And also different computer code I'm sure. Oh, and a different wiring harness as well on most models. Got to check that as well - might be a short somewhere/something different if a short happens.
(and on it goes - they haven't even gotten to testing all of the possible failure modes)
Just what/which driver, knowingly operating a clutch equipped vehicle, would not instantly resort to the use of that clutch even if the right foot was inadvertently on the gas pedal instead of the brake..?
So, it sounds like, at least in the examples you gave, the cause of the problem is known but nothing has been done to correct it?
If this was part of a mission critical system, then I would:
1. Wasn't the vibration environment known ahead of time? Safety wire or stake the screws so that they cannot vibrate loose. This is SOP on things like satellite systems and avionics.
2. Provide protection for the sensor or even provide a redundant sensor so that a piece of lint cannot knock the who system off line.
3. Require that the power supplies be able to perform properly (within specs) for defined surges and spikes on the input line.
Next you take it to the dealer service center and get the typical response.."symptom cannot be duplicated".
And now it NEVER happens again, NEVER.
What is there to report, document..?
I wouldn't hesitate to drive any Lexus, even Mr. Saylor's, because I know how to stop it.
If a person doesn't have this basic level of competence to push a button or put a car in neutral, then they need to drive something that robotically does it for them I guess.
Better for them, better for the rest of us.
As for science, medieval scientists aren't 'scientists' (the word didn't even exist then), nor are politicians.
REAL scientists spend all their time attacking their own theories. That's what makes science so much more interesting to me than faith, conjecture, politics, etc.
Sure, sceintists can be fallible and even political, but that certainly doesn't trash the actual science.
Whether you believe in gravity or not, or the person espousing it, you still go around the sun once a year.
Exactly! Science is to a large degree self correcting. Peers reviews of results, and being able to duplicate an experiment and get the same results are another example. Sure, sometimes some looney idea pops up - like cold fusion or the link between vaccines and autism. But those are eventually discredited.
You're kidding, right? You really think that examining the assembly language or machine (binary) code compiled from a 200,000+ SLOC flight control program is going to yield any insights into the correctness of the code?
D) If at the assembly language level, assuming, as would be expected, that the processor was unique and not in widespread use, how did you go about locating the maybe one or two reasonably competent software engineers available worldwide..?
What? You only need one or two SW engineers to analyze the hundreds of thousands (millions?) of assembly language lines of code that is probably in the ECU's computer?
You really do appear to have your head in the sand...
I'm not in the market to purchase a new car at the moment, and might not ever be again, but you can be certain I would not risk my life in a new Toyota or any new DBW vehicle. The first thing I would do if I had to purchase today is add a true, FAILSAFE, brake over-ride.
"..only need two.."
Again, my point was that there is maybe ONLY two programmers worldwide that would have the knowhow to analyze code unique to a narrow market use NipponDenso processor.
But the answer is yes, I think a focussed look (starting with driving the throttle full open and working backwards) at the code could yeild positive results.
In short, what *appears* to be most risky is often, statistically, far less risky than the ordinary risks you take every day.
People fear flying more than driving but driving is pure carnage compared to airplane fatalities on major airlines.
Given the NASA analysis of fatalities from UAs, it seems like more people were killed by vending machines than runaway Toyotas. (People bang and pull and shake them in a rage for having been cheated and, being top heavy, the machines fall on the individual)
Sure if the process is something very important. If you go into any typical factory, say making toilet paper or potato chips, you will not find the equipment having redundant sensors unless there is a serious safety hazard involved. There may be redundant sensors on heaters, but the sensors for the packaging and labelling sections would typically not be redundant. Why because the people who design equipment aren't going to get a sale if their equipment takes up a lot more room and costs more because everything is redundant. And so it goes with autos. Do you have 2 TPM's per tire? Do you have 2 thermostats in the cooling system? ...
If you think it's just a matter of $ and putting in redundant equipment and you then will never have a problem, consider that in the last few days, the highly engineered, expensive high-tech high-speed rtrains have had 2 accidents. I believe one was an Acela train leaving NYC, and the other was a head-on crash in Germany (?). Here you can read some history on the vaunted German engineering and their success with eliminating equipment failure and risk.
http://www.france24.com/en/20110130-germany-train-crash-kills-ten-head-on-collis- ion-hordorf
Technology is very fallible, no matter what you do.
I dunno - many of those guys were under the thumb of their royal sponsors. Even Newton managed to get a patronage appointment to run the mint and wound up causing a lot of silver coinage to be melted down and shipped out of Britain. Not the most brilliant move (maybe it was the mercury in his system that did it).
Post Toyota Recalls, Brake Override Could Become Law (AutoObserver)
But we don't throw out his good work based on his errors in other areas, which is my whole point about science, UA, climate change or anything else.
Facts are still facts, even if spoken by a hippo wearing a pink tutu.
Flying...I wouldn't except for business requirements.
Chance of dying from a car accident: 1 in 18,585
Odds of being murdered: 18,000 to 1
Chance that Earth will experience a catastrophic collision with an asteroid in the next 100 years: 1 in 5,000
Odds of fatally slipping in bath or shower: 2,232 to 1
No, $$$ and redundancy can help. But understanding what might cause failures or faulty operation is probably more important.
There are satellite systems I've worked on that were designed for a 5 year life on orbit, but are on going into their 15th year of operation.
There's the Voyager spacecraft, launched what, 30+ years ago, still operating (somewhat, anyway).
There's the Martian Rovers, both of which greatly exceeded their 3 month design life on the surface.
A lot of these successes had to do with understanding the risks, recognizing where the hazards might be (radiation,for instance), and taking the appropriate engineering measure to ameliorate those issues.
Yes, and now we've replaced our royalty with corporate and political Thumbs. Either way you have a powerful central authority which controls the $$, the rules, and even more so now - the ability to market their agenda thru the media. You should not underestimate the enhanced power that anything labelled "official" carries when presented in the media. For example when you see a TV news article that the police/government has arrested someone for murder, what is your first thought. "That must be true"? Why? Because you're seeing it on TV and you're conditioned to believe authority/officials.
Go back and consider the whole period where the government presented data and evidence thru different spokesmen, night after night on the news to convince the Congress and public that there were all sorts of WMD in Iraq. Remember all the pictures of the launchers and the biolabs from our satellites being displayed each night? What a great case they made! Create your own news! Snicker at anyone who disagrees with that story, as an obvious nut! They showed you the data! Didn't they?! Believe - Big Brother.
Sure. I've made matrices where we plotted and ranked risks vs. probability of occurrence and consequences. And what happens is that the reality of the cost to address each issue, and the need to launch new product quickly, means that not all risks will be tested or addressed.
If you've been working on satellite systems or other aerospace systems that is a different mentality than launching a new vehicle, or any other commercial item. The design cycle of any commercial product has been compressed, pressure is to keep the project rolling, minimize delays and testing,launch the redesign, get to market before the competition. Keep the cost down.
Because there is so much competition and margins are fairly low, you are going to get vehicle designs rushed and testing will be at a minimum. I think most auto engineers do a great job, given the conditions they work under, but they are being pushed by businessmen. Do more with less each year! (I went to grad school with a bunch of Sylvania engineers who supplied the Big 3 with lights, who used that motto).
You probably don't have that sort of environment when working on government projects, though I'm sure it has it's stresses, because there would be more upfront failure-modek work done.
2013 LX 570 2016 LS 460
Except for the MMGW "scientists". They spend their time keeping info. away from their "peers", hiding info, and making exaggerated claims, etc.
2013 LX 570 2016 LS 460
No, I'm not a shill for Toyota--personally I'd rather drink ammonia than drive a Toyota every day. But that's just me, and it has nothing to do with safety, but more with driving characteristics.
re: Science -- the neat thing about science is that it's true whether we believe it or not.
If anything, the tragedy of the WMD myth points out quite clearly, in relation to UA, that it is the persistent belief that something that shows no evidence of being there, really is there, somewhere, somehow--- that is the stumbling point.
You're right. It's a different environment, but with it's own set of stresses.
See the full report:
http://www.nhtsa.gov/staticfiles/nvs/pdf/NHTSA-UA_report.pdf
and this lawyer site gives some serious issues:
http://kansascity.injuryboard.com/defective-and-dangerous-products/what-nasas-re- port-said-about-toyota-sudden-acceleration.aspx?googleid=288272
http://www.eweek.com/c/a/Desktops-and-Notebooks/Intel-Resumes-Chipset-Shipments-- - for-PCs-Not-Impacted-by-Flaw-737197/
So let's compare this Intel issue to NASA's review of Toyota's UA. Intel probably has some pretty bright people working there right; equal to NASA? And I bet on the design of this chipset Intel had more engineers than NASA, as for NASA this was only a side-job. And I bet each engineer at Intel put more time into the chip project than NASA put into looking at UA. And the people at Intel had the detailed history and knowledge of how everything should work, whereas the folks at NASA were new to the UA-issue when they started.
Yet apparently through the whole design phase, and testing, and well into the manufacturing of the chips, Intel had no idea this flaw was there. Couldn't find it!
Now I'm sure we can find other examples. We could look into Microsoft's history of spending millions of manhours on a new Windows version and overlooking the many flaws they have to patch. Or we can look at nay other software company and the flaws and patches we know occur.
And after ALL these examples, you can sit there and say that because some folks at NASA looked at, that the problem doesn't exist. That just seems like an incredibly illogical conclusion.
And the source code probably wasn't even under Toyota's control, more likely than otherwise it was NipponDenso employees who wrote the code.
2013 LX 570 2016 LS 460
The move to making everything electronic and eventually to having the car "drive itself" is going to be an increasing issue when accidents do happen. Toyota and the other manufacturers might want to consider this as one of the first in many legal problems they're going to have if they continue on this course.
The onus will be on these auto manufacturers and the suppliers of the code to prove that somehow they do not have the design flaws and errors that seem to plague every other product and company. Intel, MS, a lot of companies in aerospace all seem to have hardware and software issues. Recalls, patches, and missiles that miss the targets. Example after example that are public; I'm sure there are several times that many which are never made public!
But auto manufacturers have no electronic problems; they themselves have had their experts and paid experts bless them. And the bright guys over at NASA looked too. There you go. Case closed? Conclusion - there's no proof there's a problem, though every other company and industry has oodles of software and hardware problems.
You remember how I talked about the theory of epicycles being used by desperate astronomers in times past, to try and patch up their speculations---and as nothing in their speculations actually made sense compared to observable phenomenon, they kept on coming up with even more elaborate explanations. Soon the epicycles (mini-orbits) grew so complex they couldn't even draw them anymore.
What is interesting to me is that most problems, even in complex machines, are, when finally discovered, not that complicated at all.
Why did Challenger explode? (a machine far more complex than a Prius)_
Why did the Comet, the first jetliner, crash?
Why did Apollo X burn up on the launch pad during testing?
Why do submarines sink?
Inevitably, they trace it to issues, that, once discovered, seem painfully obvious.
It should be, that if your Prius does an UA, you should be able to lend it to me for a few weeks and it should UA on me as well.
If it doesn't happen for me, or for anyone else but you, then what's the answer?
Exactly.
****
Sure you can. Crash the machine on purpose. 99.9% of the time, rebooting a computer like they use in cars and other electronics "fixes" it, and unless there is an entirely separate computer system monitoring the first, it will have no idea what happened. The same thing happens with software. Say, I'm using NERO to burn a DVD. And it crashes. Which it is prone to doing. Force-quitting it and restarting always results in a clean slate. Now, true, some applications do have crash logs, but does Toyota for all of the components and sub-components in the chain? From what I gather of OBDII type systems, they are crude 1990s technology at best.
I had a Buick LeSabre years ago with ODBII on it and it came back with a cryptic "do not run" message. Not one code. Engine was dead as a rock. The actual problem was that the timing cog had ground off its teeth and the valves were randomly operating as a result.
No codes, no record, nothing at all. It didn't even record the "do not run" message, either. Yeah, trust me - if you haven't found anything wrong with a car that's exhibiting odd symptoms, the computers likely haven't caught it either.
NASA is smart but maybe the problem isn't there.
Why do you think that? The reports I've read indicate 3 things: 1) these vehicles have been on the road for a number of months to years without a problem. and 2) it is a small number of vehicles affected, and 3) the problem does not necessarily continue to happen each and everytime after the 1st noted incident (for those that didn't crash).
For all we know, UA may only happen when a certain series of events happen, such as traction control engages, or stability control while at full throttle. If you don't drive such to trigger this or that condition the vehicle may run fine. As others said it may be some external factor as part of the equation, which you would not drive by.
So you driving it for a few weeks in your manner, and in your environment would only lead you to the false conclusion that your test represents the greater body of possibilities. It doesn't. Your testing would only eliminate a fraction of the possibilities. That's no different than digging holes in the desert, and after your 3rd hole declaring that the other 99% of the desert didn't contain any ruins.
Personal experience: I had a '92 Dodge Spirit RT - remember that special version. I bought it used at a good price, because it was a Chrysler "buy-back" or lemon. I took a chance and they gave me a 1 year unlimited warranty. So I had it nearly a year when the 1 day I went to leave work and nothing - no starter crank, zilch. I had it towed to the Chrysler dealer who the next day looked at it, and it started right up. The mechanics looked for any codes. Zilch. I believe that happened again a month or so later. I traded it in on a new Subaru.
I could hardly think of a *better* test of UA then to take the problem car and give it to other drivers.
The easiest problem to fix, yet the hardest one to find, is the problem that never existed in the first place...
I'll see your conspiracy theory and raise you one Facebook giving Mubarak heck.
Noncarpetmntis. Feinman is playing bongos somewhere and questioning why the pearly gates squeak.
You driving the vehicle does not even sample these conditions/factors. You are thinking very limitedly that the problem is independent of the operator's actions; other than your belief that the operators are stupid drivers. If the vehicle is 100% of the driving equation, then you should be able to win the F1 championship next year if you get the right car!
As wwest pointed out also, the vehicles and code that NASA looked at could have been fixed. Would someone do that for the billions in profit that ride on the outcome? I could see someone arguing that you can't change the past, and what good comes of additional punishment of Toyota for what was a mistake, that they now fixed?
The reason I am reducing the argument to absurdity is because this seems to be a recurring thread in the UA incidents, to wit, no one can reproduce the incident except the victim.
This is highly suspicious, n'est pas?
If you and Houdini buy a Weber gas grill and have trouble lighting it and then run it such to set your house on fire, does that mean if I buy a Weber gas grill it will do the same? Of course the operator and conditions of how any machine is operated has an impact on how it functions.
If all of us have the same exact software and hardware configuration on our PC's, what we do, what we're running, has a tremendous impact on the probability of either one of us having the system crash.