Safety group claims Toyota's acceleration problems caused by electronics

[Omaha Steve]

Source: Washington Post

By Frank Ahrens With the words “Toyota Truth” written repeatedly on a large red banner as a backdrop, four non-company safety experts gathered at the National Press Club on Tuesday afternoon to offer their theories on what is causing the Japanese automaker’s problems, including runaway acceleration, The Post's Dana Hedgpeth reports.

Bottom line to this group: It IS the electronics, despite Toyota's repeated claims to the contrary. Toyota says the acceleration problems are caused by mechanical issues, which are being addressed by its two big recalls.

The panel was moderated by former National Traffic Highway Safety Administration head Joan Claybrook, who has been a strong critic of her former agency, saying it is a "lapdog, not a watchdog" for the auto industry.

Keith Armstrong, an electronics expert, and his fellow panelists argued that there needs to be a push by NHTSA to have better design safety regulations for the electronics in vehicles. Toyota – or any other automaker – couldn’t possibly do enough testing to try out all the scenarios that can cause problems, they say. That, Armstrong has calculated, would require driving one car about 200 million miles or “test 36 vehicles, 24-7, for 10 years.”

Read more: http://voices.washingtonpost.com/economy-watch/2010/03/safety_group_claims_toyotas_ac.html

[KILL THE WISE ONE]

1. now they are getting close

there is a Gremlin in the electronics and not an AMC Gremlin.

[truedelphi]

2. Back in Mid march I had a major over acceleration problem w/ my Ford van.

Due to a comment that someone had made on DU maybe four or five days earlier, I knew that the fix for the van would be simply dis-connecting the cruise control. (We never use the cruise control and it wasn't supposed to be in operation at the time of the acceleration.)

I had the van towed home, and my spouse disconnected the cruise control, and the van was fine. (You could tell when it was fixed, because the awful noise that the van assumed when over-accelerating and that made the van sound like a 747 jet imediately ceased.)

So thanks to whichever person here said that this was the fix.

[Atman]

3. Why would want to DRIVE a computer?

Especially when you can't even rely upon one to deliver your e-mail on a daily basis without crashing?

Mechanical devices weren't/aren't perfect, but they had them pretty much down to a science. Now, our cars' engines are a science experiment. Run by software. Software which offers up the "Blue Screen of Death" on our desktops and freezes up our mobile phones and craps out our cable boxes. But we're relying upon them to control our cars while we're careening down the highway? And people are surprised that there are problems?

The only people that surprise me in this situation are the deniers. They say "Well, it only happened a small amount of time in thousands of cars." Same as with computers. EVERYONE'S computer doesn't crash daily. But EVERYONE's computer crashes now and then. And if you called in tech support, they couldn't replicate the crash just by sitting there pressing buttons for a couple of hours.

This is one of those "SO FUCKING OBVIOUS" things that amazes me. I'm guessing some PR asshole weighed the pros and cons, the idiots vs. learned, and decided there were more idiots so they should play to them instead of those with cognitive thought processes.

Fuck Toyota. Karma bears heavy chains.

.

[530jonathan]

7. computers

I am an certified auto mechanic very familiar with drive by wire and computer systems. The sad truth of the matter is that the PRIMARY reason why these systems exist, especially the dbw systems, is to pass the ever stringent emission standards that the EPA puts on vehicle manufactures. Case in point, with the dbw, on a normal car, when you let you foot off the accelerator, ie, when coasting, when the throttle slams shut, the pollution increases by a fraction. Same effect is when you slam the pedal to the floor to accelerate. The dbw dampens these things resulting in a lower fuel consumption and decreased emissions. I'm all for clean air to breathe, and for the environment, but as long as these standards keep decreasing, expect to see car company's using more and more of things like this in the future. And, expect to see your repair bills go higher and higher as well, as cars get more complex. I work on german cars alone, and we've been using dbw since around 1999, and never have problems like toyotas though.....

[Atman]

8. Welcome to DU!

:hi:

Good response! Thanks.

[IDemo]

4. Haven't heard Joan Claybrook's name in a while

She was considering requiring motorcycles to have rear-wheel steering and rollbars during her term at NHTSA. Before someone had the kindness to point out that motorcycles don't behave nicely with rear wheel steering.

I'm also of the opinion that at least some of the reported incidents are genuine; not driver error or floor mat related, and that the only realistic point of failure lies within the ECM (engine control module). But Toyota does do a great deal of testing on the electronics, as seen http://pressroom.toyota.com/pr/tms/2010-toyota-electronic-throttle-154266.aspx?ncid=12036">here. Maybe no amount of testing is enough for a system that holds lives in its hands, though.

I read somewhere two days ago that upwards of 25% of a modern car's cost is in the electronics. That would make for a great deal more time in the test chamber than a laserjet printer undergoing FCC regulatory testing, for instance. Time for another look at the drive-by-wire concept, maybe..

[HopeHoops]

5. I've been saying that all along! Fucking placebo shim my hairy ass.

[Mopar151]

6. The thing is, other car companies have done "drive by wire"

A lot better. As a dealer tech posted here, and my friend the wizard programmer confirms - this has to do with issues like default conditions, redundant sensors, and system architechture. Most automakers use open source code, and code readers are readily available. In fact, aftermarket programs are available from multiple sources.
Toyota, in contrast, has been very secretive about their programming, and, until very recently, had ONE computer in the US that could download crash data recorder info - and this has been a requirement since OBD-2 was introduced in the '90's

[madokie]

9. I worked for a decade as an electrician

converting the machinery in the local foundry from being operated by a shitpot of relays to a process controller and it would always take months to get all the little kinks out of the code. Somethings they never did get the program to where there wasn't a glitch here or there. Most times yes they did but there were a couple of the machines they never did. Get them to a point where you can live with it was the way it pretty much was done in those cases. We would always have to do the work as the machines were in use as neither system uses the same wiring or input/output devices so sometimes I'd work on a project for months. Then take a long weekend like holidays to make the big change over, brain transplants is how we referred to it. One project alone included over sixty thousands feet of control wire, that was only the first order with several made following that as conditions warranted with the bulk of the wires cut and used at less than 30 ft long, every color you can imagine. Lots of terminations over many weeks but my brother and I was who they wanted to do the work and they planned their schedule around that. It was great work but I got burnt out after 11 years. Oh and it was real stressful

What I'm really saying is that toyota has a code problem and it will take them down if they don't address that. I gave my big blow there to give you some idea as to maybe I might know what I'm talking about on this as I've seen it all before, the denial, the fighting and in a few cases the never admitting. I've also seen some machines hum like a kitty kat too so this toyota problem can be fixed. I say that because I've seen it done with lot bigger machines than an auto can ever wish to be with whatever electronics they want to put on it or control with.

[Mopar151]

11. I've had similar experiences

From a different angle - I'm a machinist and mechanical tech, who has done the mechanical bits, including sensor placement and running wire/cable, for various sorts of machinery - and I'm pretty good at fixing the crashes when things don't go according to Hoyle.
And brotha, I hear you about the stress! There's always a bunch of naysayers around, who don't trust anything new - even if the old one never worked! And you can do a lot with code - but there are some things that cannot be programmed around. Knowing the difference is a BIG deal. That's how you get to the "hummin along" stage.

[madokie]

12. For a good part of that time the programmer was the kind that it couldn't be in his program

so I'd have to go out and trace everything down to make sure I hadn't made a mistake, which never was the case btw, then he would work on his program. Finally the powers to be figured out the biggest problem was the guy who they had doing the programming and ran his ass off and hired another and then things started looking up for all of us. After we finished all the changeovers, about 10 years worth, to all the machines I checked out and went back to finishing concrete. For some reason I'll never understand i just loved finishing concrete and finished out the last 15 years my working life as a finisher.

[Mopar151]

13. There's a lot to be said for a job that stays done! n/t

[Tesha]

10. Please don't generalize from your PC experiences to how car computers work.

PCs contain billions of lines of computer instructions (code)
coming from hundreds or thousands of vendors, many of whom
are low-skilled at design, coding, integration, and/or testing.
No one in the world can predict how your particular mash-up
of Operating System, Applications, and Network is going to
behave and given the lax standards of PC programming and
testing, it's no surprise at all that the complete system represented
by your PC frequently misbehaves.

By comparison, the engine control module in your car is a
"purpose built" computer with a MUCH smaller collection of
software, mostly all designed by a relatively few engineers
who are probably of at least reasonably-high quality and
skills. The range of environments in which this software
must function is also many orders of magnitude smaller
than for your PC. The operating system that controls the
overall operation of the car's computer is similarly small,
purpose-built for so-called "realtime" operation, and itself
is much more reliable than (say) Windows.

These systems are small enough that if a vendor wanted,
they could probably be mathematically "proven" to be
correct. That's certainly within the realm of technical and
economic feasibilty for at least portions of the system (such
as the throttle control subsystem) and the Operating System
can provide good isolation between the various subsystems,
ensuring that others keep working even if one fails.

These systems are also designed to detect and recover from
faults. For example, if a subsystem stops processing, a
"watchdog timer" can detect this and restart the failed
subsystem or, if necessary, the entire computer. And this
can happen in a very few seconds or less; none of this
"my computer takes five minutes to boot" stuff.

So please don't assume that, like your PC, the engine
management computer "must" have fatal faults just because
"you know how computers are"; this sort of computer is
in a whole different domain of computer engineering from
your PC.

Tesha