I’m disputing it, for one
3 Likes
I’ve only read his 2012 report on An Electronic Cause for Sudden Unintended Acceleration – 4/11/12
But he’s been writing about the subject ever since, last post was in Jun '23. You can read and judge for yourself. Might take awhile but i discovered him from a web search and found his writing to be incredibly well presented.
Dr. Ronald A. Belt’s Sudden Acceleration Papers - The Center for Auto Safety
Deleted by me
Yeah… Most people with “unintended acceleration” that I have read about drive Ford Escorts, Nissan Altimas, and the like. I don’t think I have seen one story when an 80 year old man crashes his 420 hp Mustang into a building due to unintended acceleration.
The only thing that came close was the issue with the Audi 5000 back in 1984 that started the awareness of unintended acceleration. The 130 hp Audi 5000 had many incidents of unintended acceleration that ultimately were attributed to a unique pedal placement that put the accelerator closer to where Americans were used to finding the brake pedal. How a car with brakes that could generate almost 600 hp of “stop” could overpower 130 hp of “go” from the engine was apparently a mystery to many in the media that wanted a sensational story of killer cars.
3 Likes
I read the entire article.
The investigations and deconstruction of control systems was very thorough and based on my experience accurate in how these systems are designed and the failsafe mechanisms they include. The biggest issue I see is the reliance on IF. In many cases, the possibility of failure requires multiple design mistakes, errors in fabrication and/or lack of proper design verification.
So we have this list of potential contributors and a speculation that the brown out condition exceeds the design criteria. Why didn’t they simply TEST for that condition and look for fault conditions? The supporting evidence is all relying on speculation and witness testimony (which if you ask any law enforcement people is notoriously unreliable, especially in stress environments). This is easy to empirically test. In fact, the test regimes required to pass compliance certification require this type of testing to be performed under the worst case conditions. Voltage sag, spikes etc they are all empirically tested and verified that the designed protections work.
So then if there is design verification, under worst case conditions, and it passes, what is left?
Exactly what the author pointed out- component out of tolerance or other processing variables leading to components not meeting original design specifications. There is heavy reliance in industry on initial compliance testing with some level of ongoing production testing to verify key design elements are not compromised. I highly doubt they are repeating the entire suite of brown out and reset trigger functions on an AQL level let alone every vehicle.
So that brings us back to the definitive evidence. What prevented them from taking a car that was supposedly involved in one of these incidents and subjecting it to actual testing to verify that the failure mode they propose actually occurs? Seriously, here’s where the entire argument falls apart IMHO-
Supporting evidence for this explanation is provided in the following sections of this paper. The supporting evidence
includes testimonials describing how battery replacement cured sudden acceleration in several vehicles, and
discussions of how the above explanation explains many driver observations associated with sudden acceleration
The first situation is a correlation argument. I changed something and the problem went away. That is not definitive.
The second argument relies on witness testimony which I already discussed above.
My suspicion is that someone or some organization has done a more thorough analysis and concluded they couldn’t get it to happen. There is no better failure analysis situation possible than having multiple examples of the hardware available for empirical testing. There should be no need to conclude the analysis with more speculation or correlating evidence. A real root cause analysis would reveal the cause and be repeatable…
There’s a whole host of issues I could continue delving into but at the risk of boring people to death, unless the discussion is ongoing, I’ll leave it at this…
2 Likes
+1
The media loved their “killer Audi” message for quite a while, back in the '80s–until it was pointed-out that pedal placement in those cars was the most likely cause of “sudden uncontrolled acceleration”. Ergo–human error, aided by an unorthodox placement of a purely mechanical device.
Really George? Some have not? Try like 99.999999% have not.
2 Likes
Do you happen to know the actual statistics?
And much more media coverage of these types of things.
3 Likes
Folks, let’s use Occam’s Razor here. The simplest explanation is usually correct. You can make all sorts of Rube Goldberg type solutions to problems that are complicated and difficult or impossible to replicate. In this case the giveaway to the simple solution is the quote we often hear after an unintended acceleration incident: “The harder I pushed the brake, the faster the car went.”
The prosecution rests.
I tend to agree with your assertion. The witness statements to that effect seem pretty convincing as to the cause.
However, real engineering is based on scientific methods and sound reasoning. In the case of a failure, ALL potential contributors are evaluated. They are given an estimated value for probability and scored accordingly. Then there is an effort to either prove or disprove any of them as a cause. If they cannot be ruled out, they are still in. In this case, if no evidence exists of a hardware issue, then the highest probability is attributed to human error.
There has been a lot of effort put into searching for hardware causes and so far, none have really been shown to be definitive. So they took the highest probability causes and put in place measures to address them (e.g. mat hooks to prevent them from interfering, additional protections in engine control computers to override the gas pedal input if the brake is also applied (not universal AFAIK), etc.
I personally like to see the actual failure analysis so that I can see what level of diligence has been put into it. Often, companies want to sweep stuff like this under the rug as quickly as possible. To see the thinking, rationale and steps taken is reassuring. However, the report cited has a glaring omission- the effort to definitively verify the proposed cause…that is a huge red flag for me…
