Readers may wish to review the following recent article which I cited in my original post here. The points made are excellent. The information presented is public knowledge.
How Do We Know Driverless Cars Are Safe? Google Says 'Trust Us'
Posted: 07/01/2014 7:23 pm EDT
Rich Pell, I don't intend to make personal judgements about commenters as you have chosen to do. It serves absolutely no purpose except to attempt to discredit and discount comments. I think it is also against the rules on such sites to participate in global negative generalizations about individual commenters.
I'd appreciate the discussion to center around the vehicle critical safety standards as this was the purpose of my original comment. I specifically cited findings of experts in the field.
Unfortunately, Michael Barr's 300-page findings about Toyota's glitchy ETCS-i are part of *sealed-court documents* in the consumer-victory Oklahoma court case on a runaway 2005 Toyota Camry. Such court-sealed documents prevent the public from learning the truth in a timely manner. This is not only unfortunate but a detriment to public safety. No one but the automaker benefits from hidden key court documents which highlight unsafe situations in vehicles still on the road.
So, I just ask that you not make this about me personally, Mr. Pell. Readers can decide what your motives are for doing this.
Interestingly enough, in many of the sudden unintended acceleration cases covered in the media currently, the drivers are often victims of character assassination before any evidence is revealed. These potentially defaming headlines are published via law enforcement speculation before any *scientific* data is in.
As I mentioned before, I believe thinking like a scientist is essential. Probing questions need to asked and properly answered. Driver character assassination is nothing more than a deflection away for the potential fatal flaws in the affected vehicles.
Mr. Pell, the discussion is about complex electronically-controlled vehicles. In a rather simplistic example, the other day my grandson had a battery-operated bubble-blower working just fine. However, when it came time to turn it off, nothing worked. The machine kept going on its own until I removed one of the batteries.
Obviously, there is no glitchy software there. However, I continue to experience issues off and on with my software-driven electronic devices. I have several I use daily. In many instances, I must reboot them to reset them. This is expected, isn't it, with any electronically-driven device.
The difference is that electronic glitches in vehicles which can result in fatalities is not acceptable under any conditions. The fail-safes have been shown in particular cases to be ineffective and most likely auto consumers are unaware of this possibility. The mantra is that braking will always stop the vehicles if an electronic glitch occurs...that the open throttle will be effectively countered. Unfortunately, drivers have revealed time and again that this isn't the case.
I choose not to discount or discredit what the vehicle owners are telling us. One has only to look at the complaints filed to see the uncanny similarity of the experiences of these SUA victims. This is what NHTSA is supposed to be doing, isn't it...looking for trends in the vehicle owner complaints? The whole NHTSA safety issue is currently under fire in the faulty ignition switch saga. Prior to that, NHTSA came under fire regarding the Toyota sudden unintended acceleration debacle. I'm certain the public is a bit skeptical about the preservation of its safety on the roads based on recent findings of cover-up and continued deceit.
Much more public debate and dialogue is needed. Continued use of the scientific method is needed. The short duration NHTSA/NASA study into sudden unintended acceleration was shown to be flawed. Far more study is needed by experts. Does NHTSA have such experienced embedded software experts? I asked the question before but as yet haven't seen the response.
Its a nice expression to accept the fault. As long as you use software for infotainment and ohter fancy dashboard features in automobile the risk is low the time you start using it to control or even access the basic function of automobile thats is to speed up or stop or turn around or reverse, the risk is way too high.
Most conclude driver error when there's evidence of drunkenness, being asleep at the wheel, girlfriend/boyfriend problems, texting while driving, turning around to see what the kids are screeching about, high school lost the ball game, or any number of other typical causes, Charlene. It's not like I don't see people weaving in and out of their lane, or catatonic at traffic signals, or tailgating, or displaying any number of other boneheaded driving, which have absolutely nothing to do with any potential "unintended acceleration." To pretend there might be a conspiracy by automakers to hide what would explain boneheaded driving behavior is a tad stretching the point.
Traffic fataliities are very well documented, Charlene. I will wager that a good 99.9 percent of them would have been avoided with better sensors and automatic controls in cars, and will instead be repeated if we insist that manual driving is "safer" and that automation should be dreaded.
Says that in 2012, 10,322 deaths were attributed to drunk driving, and 3,328 to distracted driving (which also injured 421,000), out of 33.561 total fatalities. Both drunk driving and distracted driving, in that one year, drastically outnumbered any unintended acceleration incidents, and both of those types of accident could have be avoided with automatic controls, or reduced with driver assistance.
These are huge numbers, even just in one year. And to be clear, I'll repeat, this is not to excuse Toyota or anyone else. This is only to keep a reasonable perspective on the subject matter of driver assistance or even autonomous driving. Let's not let ourselves get carried away with hyperbole. Software glitches and mechanical glitches can be resolved. Boneheaded behavior by human drivers cannot.
The "look back" is a rather vague reference to an "idle stabilization" mechaniosm, i.e. not a full-open throttle system, which may have startled drivers into planting their foot on the accelerator pedal, thinking it was the brake pedal, because the pedal placement was slightly shifted compared with other cars.
That latter part is what I remembered.
The article isn't sure whether this Audi system was firmware operated or not. Which isn't that important. I had a similar experience in a Fiat, with a mechanical cable-operated hand throttle, which bound up the accelerator pedal linkage at one point, keeping the idle speed up to above 2000 RPM when I was attempting to slow down for a traffic signal. The immediate remedy was push clutch, push brake, shut off ignition. Then disconnect the hand throttle altogether, to keep this from recurring.
Whatever the catalyst may have been in the Audi case, the sudden full throttle acceleration was caused by driver error. Only pointing out that human drivers can be unpredictable, and that we need better systems, rather than nothing at all, to assist, if not take over.
Bert you might wish to review the following item.
A Look Back at the Audi 5000 and Unintended Acceleration
Friday, March 14th, 2014
@Bert, please cite the study which shows your "bottom line" premise. Has someone actually studied the occurrence rate of the electronic software glitches and the aftermath? OR, has this glaring omission been minimized through a clever and very calculated PR campaign?
Since no one is currently tracking and properly investigating vehicle crashes into buildings, storefronts, and homes because they occur on private property, the evidence that the glitches aren't as deadly is yet to come in. Drivers experiencing a sudden unintended acceleration event in parking lots or from stops is ever-increasing. Try this...do a Google Alert using the words "plowed," "barreled," "careened," "slammed," etc. along with other key words like "building," "home," "storefront," etc. and see what you come up with. There are simply too many vehicle incidents to keep up with. Most all conclude with driver blame when a MECHANICAL cause is ruled out. No one is looking at the complex electronics and driver statements are discounted or discredited. Drivers with concomitant variables are doomed before any evidence is in. Just take a look at the Carol Fedigan Chrysler Jeep case to see what I mean. Outrageous!
I am unwilling to take responsibility for learning how to mitigate a potentially deadly consequence of owning a poorly-designed vehicle. Such vehicles should not be on the road until the critical safety standards are designed into them. It is absurd to ask the driver to take the potential deadly roller coaster ride and then try to stop it...or not. There is no logic behind this way of thinking. By "teaching" vehicle owners to react to a potentially deadly sequence of events, you are in effect minimizing the matter. This is the same as asking GM owners to go ahead and drive their defective ignition switch vehicles while they wait for the repair. Maybe they will encounter certain death...or not...based on their reaction to the failure. WHO should be asked to take such a risk and WHY?
Excellent points, @Antony Anderson. It bothers me greatly to see actual sudden unintended acceleration accounts discounted as if in an effort to shield the automaker from close scrutiny in the matter. Good scientific inquiry includes ALL the recorded information, especially that given by the driver. I'm seeing in so many of the media articles, this critical information is either being omitted altogether or discounted with the assumption or speculation that the driver is confused.
My Mom the Radio Star Max MaxfieldPost a comment I've said it before and I'll say it again -- it's a funny old world when you come to think about it. Last Friday lunchtime, for example, I received an email from Tim Levell, the editor for ...
A Book For All Reasons Bernard Cole1 Comment Robert Oshana's recent book "Software Engineering for Embedded Systems (Newnes/Elsevier)," written and edited with Mark Kraeling, is a 'book for all reasons.' At almost 1,200 pages, it ...