Frank Ahrens: Why it's so hard for Toyota to find out what's wrong

The author has a degree in engineering while Norton is a respected lawyer/politician.
That itself says a lot.

 

As a software programmer for decades myself, I deal with that lack of understanding routinely.

It's a monumental challenge trying to accurately recreate an error when the user has no idea what they did to invoke it in the first place and is painfully unaware of the circumstances of the environment in which it occurred. To make matters worse, they cannot even recreate it themselves.

Troublesome of not, that's the reality of the situation.
.

 

A couple of techniques are logging of the failures, and failure analysis. Since toyota has refused to use these techniques, I would suspect that one reason it is taking so long is toyota has been actively avoiding some diagnosis methods. Toyota has now said it is stopping this behavior, so I have hope the problems will be diagnosed.

 

I do think that's a pretty fair article with what has unfortunately been a rare media viewpoint.

But the headline is still alarmist, combining "Toyota" and "Wrong" in bold type once again.

Basicly that has been what I have found most troubling the past few weeks. Politicians, and/or "experts" with questionable motivation and agenda have speculated or suggested that Toyota has a problem with software and/or the electronics of the system.

But as an automaker Toyota cannot "fix" a suggestion or an accusation. And Toyota like any and all auto manufacturers shouldn't be expected to do so. All Toyota can do in the face of specualtion and accusation is investigate.

There are probably multiple fingers being pointed at Toyota lately, but there are at least two connected but distinctly different fingers. #1-Toyota hasn't earnestly and correctly and transparently investigated reports of "unintended acceleration". Combined with allegations that the investigation when implemented has been substandarded or flawed. ....This is debatable.

#2-Toyota has a problem....it's software or the electronics of the throttle system and they should fix it. This is where the above article is applicable. That isn't fair, because whatever your position is on #1, nobody has repeatable proof or definition of #2. Without solid evidence of a definable, repeatable flaw expecting or holding Toyota to a definition of negligence is unfair.

So in summation, while I think Toyota has gone #1, I think it unfair for anyone to go #2 on them.

 

The evidence is the number of reports of unintended acceleration where floor mats were not involved. Proof is a weasel concept when not investigating customer complaints.

For example toyota first said the only problem was floor mats, then a whistle blower came forward and said Toyota knew about sticky throttles. Given your standard of proof, the cts pedals should never have been investigated, since dealers in the US had not looked at the pedals to see any proof. It was internal documents and tests in europe that were some how compartmentalized away from the united states.

Having evidence is not proof, but reason for a good faith effort at investigation to either find the problems or rule them out. There is ample evidence that toyota has not until this point put forth a good faith effort. They have put the burden of proof on drivers. Do you think that is fair? Do you think drivers have access to the information?

 

Discovery is always a process. It's unfair for people to expect Toyota or anyone to be able to jump to instant resolution without the process of discovery and definition.

Your example is incorrect. Toyota never said the only problem was floormats. They said it was a potential problem and moved towards resolution but it would be foolish and near impossible for Toyota to advance it as the "only" potential factor. You can fix one problem and then discover others....fixing one does not negate the discovery of others and Toyota never advanced this idea, they couldn't, it would be childishly foolish.

As far as standard of proof? I offer no standard of "proof" only that repeatable, definable results should be available before politicians or others accuse toyota of having a problem they should fix. So far, I hear a lot of accusation but imo nobody has clearly defined the source of the problem. Nobody, Toyota or outside sources.

When reports of unintended acceleration come in, in abnormal numbers everything should be investigated. Pedals, Electronics, floormats, user error, enviromental conditions...everything.

As far as dealers looking into problems? I think "dealers" or dealership service departments try to fix cars to Toyota specs. They have specific ways of reporting problems to Toyota. But it isn't a "dealer" that should be in the business of scientific investigation and ultimate problem definition. They should communicate up the ladder and leave that to Toyota.

As far as the discovery that some accelerator pedals had a defect and that discovery evidently coming from testing in Europe? Who cares? Toyota is a global company...Global. I don't care where the results are discovered, be it a lab in the US, Japan, Europe or Pandora....

You can debate, as it has been debated, until your blue in the face whether Toyota has acted quickly enough or as you say with a "good faith" effort, but I disagree with your automatic assumption that there is ample evidence that they have not.

I think there is just as much evidence that once a problem was defined, as evidenced with the pedal, that Toyota has reacted swiftly and in MASSIVE scope, at great cost to themselves to rectify the problem. But I admit that Toyota's reaction is debatable and more information as to timeline and reaction is always evolving.

It is simply however unfair to ask Toyota to react beyond investigation, to a "problem" that exists only as accusation by media and politicians born out of what they "think" might or could possibly be a problem.

Yes, as an automaker, never ending investigation of potential problems must exist...but the difference is an accusation is not a definition of a problem, and an accusation cannot be tangibly recalled or fixed. That is what is unfair of the media or politicians to expect.

 

Logs are limited. The amount of detail and how often it is recorded can vary profoundly. Yet, that isn't getting attention.

Our diagnosis depends heavily upons those logs. But how much data can actually be stored and how much processing power it takes to collect it is a really big deal. So, there's always a balance of resources & cost.

There's no behavior or avoidance if what you have to work with is limited in the first place.
.

 

It took me 6 weeks to undercover one particular problem. I cleaned up coding and adding extra logging. During that discovery process, I ended up finding other glitches.

Ulimately, I ended up finding the cause of that particular problem. Turns out, it only occurred when 25 other unrelated problems did between application resets. That was just awful to find the pattern. But as soon as I recognized it, we discovered there was a parameter with a threshold set to 25. Warnings counted, even though they weren't actually failures and they didn't impact data. So, we just increased the threshold and put monitors in places to confirm the count tally was from them and not something worse than just a warning.

In short, eliminating the exposure of floormats was essential step in the debugging process... not the final or only step.
.

 

Debugging is as much an art as science and puts extreme value on:

• Solid background - you need to understand not only the systems but also the physics behind it.
• Statistics - it is how we avoid chasing something that really isn't there.
• Intermittent problems accumulate - until the customer blows up. Often we have first level responders who are not so skilled and if they guess based upon their 'cookbook.' Many times they are right and life is good. But usually one intermittent problem will mask a second and sometimes even a 3d.

Reproducible problems are easily solved. Intermittent problems are often a learning experience.

Bob Wilson

 

I agree. I'd say more, but I've been accused of not being brief...

 

This may have been your shortest post.

 

I totally agree here, and with tony. I don't think I made my point well. By 2006 there was ample evidence of some problems that toyota should have been interested in tracking down. I would expect that if they were making a good faith effort, they would by 2008 be logging accelleration and braking data in the erd's and have had software widely distributed to be able to read the devices. I think this situation is finally changing, and they will at least be reading the information logged as soon as april. The next step is to verify that software is spitting out proper error codes and logging good diagnostic data.

 

"OH NO" . . . the Toyota nay-sayers will proclaim. "Toyota could'a should'a put 500 terabytes worth of logging capability in each and every car". Knowing nothing makes it extremely easy to fix cars & people's behavior, and cultural differences, etc. Don't you think? That's why they can't figure out why this "fixing thing" is taking so long. Dumbness means that one is used to experience all galactic problems getting solved within the span of a 30 minute TV show.

.

 

You have no idea what they're doing in-house in-lab at Toyota, nonewhatsoever. Your only inputs are what you read 3rd or 4th hand in the press. That puts your perspective in proper context. Your views on this then are like those of the representative from DC.

 

One of the points made by Toyota that bob wilson referenced is that a good number of these 'post-fix' complaints are unverifiable because they left no references such as a VIN. That complaint could have been lodged by anyone - even one herein. That complaint might be real or fabricated.

So as john pointed out and as many have noted simply identifying the problem, if there is one, can be a bear but then the level of volume can be turned up by numerous unsubstantiated reports.

 

But the data logs on Toyotas that crashed can't be accessed because Toyota has encrypted the data and there is only 1 notebook computer in the entire U.S. that must be operated by a Toyota technician (probably only 1 trained technician in the entire U.S. as well) before data can be extracted? That is constructive obstruction of justice.

 

Actually I have sworn testimony from toyota. So yes I have evidence.

You on the other hand on the braking problem misdiagnosed it, said you were an expert and were told you were wrong by toyota and the nthsa and the japanese government. If you continue to put forth information that you know is false why should anyone listen to you.

I on the other hand have been called in to get approval for the faa and fda on devices and understand the processes. If you have something to contribute fine. But large number of false posts can not be constured as productive.

 

You do understand that the auto insurers are desperate to obtain the ERD information? And that the ACLU and other privacy advocates have a lot of problems with installing "black boxes" for cars? It's not because the former want to save lives or the latter don't. The former want to save money by proving you were speeding by 2 mph and then denying your claim. The latter think protecting privacy (like where you've been just before the accident that you shouldn't have been) for millions is worth something too.