toyota electronic throttles vindicated by NHTSA

Javelin wrote:

Osterkraut wrote:

Javelin wrote: If you *really* want to get an idea of the types of losses "drive by wire" control produce, look into the aircraft industry. Even with redundancy and seperate systems there have been more computer-error crashes than hydraulic failures in older aircraft. The NTSB has some really interesting reading on the subject.

Cite NTSB sources, please.

My father is a Professor at Embry-Riddle Aeronautical University and one of the heads of Safety at NASA for the Orbiter. I took in one of his classes for fun. A lot of it was above me as frankly I'm no Aerospace engineer, but the reading is out there. Buy a book on NTSB accident investigation and fly by wire controls. The point on failure rates of the FBW vs hydraulics was in the lecture and notes as that particular unit was on why everything has to be so redundant. I was trying to tie the discussion into sources other than OEM auto manufacturers to show the problem is not unique or unheard of.

Cite your NTSB sources, please.

I for one welcome the fact that I can effecitvely have different throttle ramp rates in the same vehicle at the push of a button. Like say the new grand cherokee... If you put it in 4 low throttle ramp rate becomes slower. This would have been awesome in my old jeeps. Awyone who wheels knows that with the vehicle bouncing around you sometimes stab the gas by accident, and a feature such as this would have helped eliminate it..

It's OK though the reactions of the Anti folks are typical of those who are afraid of new technology. The same arguments proposed against DBW were propsed against pneumatic tires. ZOMG they are full of air and there is no solid link to me and the ground. If they explode we will all die and children will no longer be sweet and succulent for us to eat.

Give the technology a bit of time. The big issue is I think people are afraid of something that is non mechanical and they cannot understand. The failures are then unkonwn and scary. In my experence working in many plants, I'll always take an electronic item over an mechanical one. Anyone remember the old relay cabinets all over plants, no flexibility, all wasted space and energy.. Nope. Noone wants those again, but the again the Anti arguments were the same and the PLC won the day.

Give in to your electronic overlords people.. They're better.

In reply to Ignorant:

There is no flaw in your logic about different throttle ramps for different modes. I think the issue in this argument is the lack of predictability. With cables you always know you are going to get X throttle for X pedal application. With DBW you may get A, B or C throttle for X pedal application. It's not because you flipped a "mode" switch, but because it's what the computer wants to do at that time for varying unpredictable (to the driver, not the engineer Alfadriver) reasons. It is one more thing that disconnects you from feeling like you have direct control of what all parts of the car are doing. Heck, this crowd hates the simple throttle hang that occurs when you are BETWEEN gears, removes the "Idiot Valve" from BMW clutch lines, avoids cars with overboosted power steering, stiffens up factory suspensions, discusses the "Feel" when considering brake pads or tires, removes half their interiors for a minimal improvement in performance on a daily driver, and recommends a 20year old lightweight 2 seat sports car for everyone's perfect transportation. We basically want to be able to control what happens when we drive and pride ourselves on doing so. A throttle that doesn't predictably do what you tell it to is a pretty big lack of control. Calling a group of people who think a cool DD would be a Bi-Turbo with a wiring harness swap and a Megasquirt, technophobes is bordering on ridiculous.

Ignorant wrote: ........Give in to your electronic overlords people.. They're better.

Not completely. Not yet. I think the issue is we are early in the application of DBW and the companies are either not concerned about feel or just haven't mastered it yet. I predict it will be another thing that you tend to know about certain brands. Mazda has better feel than Chrysler, Ford has better than Chevy, etc...

Honest question for Alfa: do you truly believe that Toyota's electric throttles were not to blame in any of the incidents?

As an engineer, I'd like to point out that most of you are making a mistake about drive by wire throttles. You're actually complaining about the complex system the drive by wire throttle is wired into.

A drive by wire throttle does not need to be wired into the computer. It can be rigged as a dedicated stand alone system consisting of nothing but the input potentiometer (pedal) and the actuator (throttle plate). Done this way, it will act just like a mechanical throttle linkage or cable. And if it fails, it will fail just like a mechanical throttle would. And you would use a conventional throttle return spring. And all would be perfectly happy with it. Think pertronix for throttles if you will.

But that is not what is done. Instead, it's wired into a complex computer, taking multiple other inputs, processing it, and then the computer comes up with some sort of answer that it sends to the throttle.

THAT is where the problem lies. The complexity of the system. That computer can come up with some serious goofy answers to give the throttle plate. Some that are terrible for performance (elimination of power braking for example), some that are just plain dangerous (holding the throttle open).

A whole lot depends on the coding. How buggy is it, how robust, how complex. A good bit depends on the quality of the parts, and what happens when those parts fail. Lots of unforseen problems arise here, and are darn hard to isolate.

It would be convenient to blame it all on the engineers at this point. And a good bit of it can rightly be blamed on the engineers. But other folk contribute to the problem. Bean counters who want to use cheap parts and cut testing short. Marketing folk who want the product out now, before all the bugs are worked out. Lawyers who want liability protection for the company, etc. These things all contribute to product that is overly complex, ever increasingly complex, and riddled with problems. In this case, a car with drive by wire throttle, but we could just as easily be talking about a toaster.

Is it all therefore bad? No. Look at the incredible power coming out of modern race car engines using very complex computer codes and zillions of inputs. It can work great! So what's the difference? Market, focus, resources, etc. The Honda F1 racing team is focusing on a handfull of cars, with the single objective of going fast, and all resources are dedicated to this one task. This team has a much better chance of making a better product than the vast Honda auto group. That group is split up into many cars with all kinds of different objectives. From cup holder locations to headliners, and "oh yea", throttle control. This bunch is far more likely to run into trouble because they are not as focused as the racing team.

MrJoshua wrote:

Ignorant wrote: ........Give in to your electronic overlords people.. They're better.

Not completely. Not yet. I think the issue is we are early in the application of DBW and the companies are either not concerned about feel or just haven't mastered it yet. I predict it will be another thing that you tend to know about certain brands. Mazda has better feel than Chrysler, Ford has better than Chevy, etc...

I agree with you. I think the technology is in its infancy, but in the next 5 years will prove to be better than what we are currently used to... From a failure perspective, I'll bet they'll prove to be more reliable than any mechanical device.

MrJoshua wrote: In reply to Ignorant: There is no flaw in your logic about different throttle ramps for different modes. I think the issue in this argument is the lack of predictability. With cables you always know you are going to get X throttle for X pedal application. With DBW you may get A, B or C throttle for X pedal application. It's not because you flipped a "mode" switch, but because it's what the computer wants to do at that time for varying unpredictable (to the driver, not the engineer Alfadriver) reasons.

The Grand Cherokee has the ability like I mentioned. It is a cool but of kit.

But, with some of the comments in this threads, particularly centered around "safety", I will say that a great deal of us are technophobes and they have tipped that hat themselves.

Ignorant wrote:

MrJoshua wrote:

Ignorant wrote: ........Give in to your electronic overlords people.. They're better.

Not completely. Not yet. I think the issue is we are early in the application of DBW

From a failure perspective, I'll bet they'll prove to be more reliable than any mechanical device.

That will work until the zombies get HERF guns.

MrJoshua wrote: Honest question for Alfa: do you truly believe that Toyota's electric throttles were not to blame in any of the incidents?

Honestly, I don't know. All I am getting is the same news you are, so it's quite hard to tell what is causing what.

There's some appearence that ideas were missed- assuming that the button doesn't shut off the car, or the shift to neutral doesn't work. But I can't say that I really belive that ETC is the cause. (I also stopped working specifically on ETC about 15 years ago, currently I just deal with the interface).

But I do believe that whatever Toyota's issues may be, that does not say that the technology is bad for the rest of the industry. That can be said for a lot of things. And it's not really as if we are that early in ETC's life span- they been on the road for well over 10 years.

Typically, failures happen due to implemetation issues, not specifically due to the technology. Just like you can't hang ETC on Toyota's issues, you can't hang direct injection on BWM's issues when many other are able to do it (including Toyota) with out the same problems.

Rustspecs13 wrote: I hate DBW. Mainly for the fact that I believe that you should have control over things that could kill you, or cause huge safety problems. My friends brother had a 02 civic SI. Or whatever had the K20 first. It would randomly accelerate and bang off the limiter. Thankfully it was one of those crazy manual normal cars. So it was an easy option to push in the clutch, turn the key and turn off the car. STRANGE! IMO a solid connection between the driver and throttle body (or fuel injection pumps on diesels) and driver to front wheels should never be taken away. There are just too many ways to fail and cause more problems then they solve. I really think there should never be a disconnection between the steering rack and driver. I bet its going to happen on a large scale, but I dread that. Some things just shouldn't be done when it comes to safety. ~Alex

Ok- here's a situation:

Your throttle sticks open.

With a cable throttle, what can you do? Well, it's going to have to be the drive who does something, as the computer can't tell if it's stuck, or the driver has a very steady foot. Desire and output are the same sensor.

So you pull back on the pedal- which won't do anything, since the system is designed to pull, and not push- all of the slack in the cable will be taken up. The only systems I've seen that will do anyting are Alfa's and Mercedes, who had a very expensive rod system from the pedal to the throttle. Had being an important distinction. You are left with turning off the car.

Now, you can belive it or not, but cable systems are less reliable than wires. Even including the complex systems.

With an DBW/ETC system, it won't take too long for the computer to figure out that your foot is off the pedal- looking at the reduntant sensors and comparing it over reduntant wires, compared to the throttle- which also has redudant sensors and wires. So it can do a few things very quickly- cut power so that the car will slow down, and ALSO drive the throttle shut. IF the throttle does not shut, you will still drive the car- if the engineers decide that you safely can. Or if the second processor that is looking at the ETC decides thats a bad idea, it will shut the system down.

You'll note that it will take a lot of things failing in the ETC system for it to do nothing- and that's not including turning off the car. Most major electrical faults that people like to come up with will also shut the entire car down, ETC or not.

And that's just my ideas- I really am not up to date on the state of the art of ETC and it's failure modes.

So.

Actually, my argument isn't anything like the EFI ones. I could care less if I could fix it myself. I just want to have a direct link between me and the car so when something fails (not if, when) I can safely the control the car and bring it to a stop. Once again, I have no issues with assisted systems, like an electric power steering pump or ABS, just 100% electron replacement, like DBW.

Javelin wrote: Actually, my argument isn't anything like the EFI ones. I could care less if I could fix it myself. I just want to have a direct link between me and the car so when something fails (not if, when) I can safely the control the car and bring it to a stop. Once again, I have no issues with assisted systems, like an electric power steering pump or ABS, just 100% electron replacement, like DBW.

So, reading the thread just above yours, what can you do with the cable if the throttle sticks that makes it so much "safer"?

Pushing a rope isn't going to move much.

Considering the P71 had a sticky throttle, I'd do what I did then. Pump the pedal to try and get the return spring to bring it back (which worked), and if it didn't I'd use the ignition key to turn the engine off and pull over (which I had to do on a 62 Rambler once).

In the same situation, what would I do with a DBW? The computer is racing the engine for some reason and my foot inputs do nothing, and there's no key to turn off. Now my only option is to slam the brakes, shift to Neutral (if the auto box will let me), and hope I can get the thing turned off completely before the motor blows.

How is that so hard for you to understand?

My '58 Pontiac has a rod-type throttle linkage, clearly it's not rocket surgery to make one.

Drive a manual transmission, if all else fails, stand on the brakes and stall that thing.

Why is neutral being locked out of any transmission? Seems like an unsafe situation to me.

Shawn

You know, I had this whole series of observations typed up, and then it occurred to me that folks may be arguing different facets of the topic:

Are we just talking about different systems? I'm thinking of something like contemporary systems as I understand them (possibly incorrectly): Totally DBW throttle, hydraulic brakes with computer-controlled ABS.

The concept of all-electronic brakes (no physical link from pedal to calipers) gives me a bigger case of the heebie-jeebies than DBW throttle, but I wonder if I'm being rational: ABS already has the power to say "no brakes for you", though this doesn't seem to be a problem outside of the poor/older systems which are stymied by slick surfaces.

Ignorant wrote: Give in to your electronic overlords people.. They're better.

Fascinating how your political/economic opinions tend to color your engineering ones as well. At least you're consistent. ;-)

alfadriver wrote: So, reading the thread just above yours, what can you do with the cable if the throttle sticks that makes it so much "safer"? Pushing a rope isn't going to move much.

Javelin wrote: Considering the P71 had a sticky throttle, I'd do what I did then. Pump the pedal to try and get the return spring to bring it back (which worked), and if it didn't I'd use the ignition key to turn the engine off and pull over (which I had to do on a 62 Rambler once). In the same situation, what would I do with a DBW? The computer is racing the engine for some reason and my foot inputs do nothing, and there's no key to turn off. Now my only option is to slam the brakes, shift to Neutral (if the auto box will let me), and hope I can get the thing turned off completely before the motor blows. How is that so hard for you to understand?

My point is DBW has far FAR more failures to date, and always will. One of my friends throttle cables went bad at over 200K miles. replaced with one off a parts car with probably 150K on it. Cable failure is super rare, easily fixed.

Nissan has a bad cable quality control day, and 20 years later one or two cars needs to replace their cables at 150-250K miles. No one dies, no accidents happen, or any varying degree of really bad stuff.

Honda has a bad QC day, and something's off. Anything could contribute. The T/B its self, something in the ecu, the gas pedal sensor, or the wiring. Multiple failure points. Something made my friends brothers car (not that I like him any way..) fail. Randomly it would start accelerating, and then he'd push the clutch in, turn off the car, turn it back on and go down the road. This car was about 8 years old at the time. Cool, sounds like its going to be a fun car down the road for a cheap broke tuner.

My point with DBW, is so many more things contribute to a failure. All of them electrical. All much harder to pin point.

IMO there are some things where the driver should have DIRECT control over things. Throttle and brake are the main ones, with steering a close third. ABS is nice- as long as it works right. New cars at least are getting better with that. But this is about DBW.

Oh and for who ever brought up testing in super cold and hot temps. Thats great, thanks for testings stuff.

But then your employer goes and gets it done for the cheapest bid. Now its down to PERFECT quality control to ensure no failures. Like thats going to happen.

~Alex

Javelin wrote: Considering the P71 had a sticky throttle, I'd do what I did then. Pump the pedal to try and get the return spring to bring it back (which worked), and if it didn't I'd use the ignition key to turn the engine off and pull over (which I had to do on a 62 Rambler once). In the same situation, what would I do with a DBW? The computer is racing the engine for some reason and my foot inputs do nothing, and there's no key to turn off. Now my only option is to slam the brakes, shift to Neutral (if the auto box will let me), and hope I can get the thing turned off completely before the motor blows. How is that so hard for you to understand?

Why don't you have a key with DBW? That's got nothing to do with it. All of my DBW vehicles have a key. I think you're simply complaning about electronic controls in general, and assuming that they're all present in all conditions.

The big difference between DBW and cable throttles is that when a DBW system fails, everyone runs around with their hands in the air screaming. When a cable system fails, people just shrug and say "ah, that happens". Alfadriver has access to less sensationalist statistics (for Fords at least) than we do, so I'd listen to him on failure rates.

Rustspecs13 wrote: My point is DBW has far FAR more failures to date, and always will.

How did you get there? And where do you get goggles to see the future of "always will", because I want a pair.

I spent twenty minutes searching NTSB and NHTSA and general Googling, and I can't find any statistics at all. Not yet, anyway. If it's there, it's all buried under sixty-three hundred syndications of alternating complete vindication and demonization of DBW.

I am also extremely wary of claims that fly by wire fails on airplanes more often than the traditional stuff. The FAA is extremely conservative with safety issues like that, and right now is an amazingly safe time period in aviation reliability. Safer than any other time in history. Yet we've moved more and more to fly by wire, to the point that I believe almost all current airliners are fly by wire. So the claim that it's more failure prone than traditional controls - in an airplane setting at least - would seem to be false to me.

I believe Keith's got a handle on it when it comes to car stuff. People are scared already of new technology so they highlight the failings of the new and cover over those of the old. Look up confirmation bias - this is a classic case of it. A failing drive by wire system will typically fail to "off" more readily than a traditional cable throttle. One anecdote of a specific car going to redline with no post-incident investigation is not at all damning of the entire technology.

Rustspecs13 wrote: Oh and for who ever brought up testing in super cold and hot temps. Thats great, thanks for testings stuff. But then your employer goes and gets it done for the cheapest bid. Now its down to PERFECT quality control to ensure no failures. Like thats going to happen. ~Alex

my employer does the specification, design, development, validation, and manufacturing. but that doesn't matter, because successful completion of all that testing defines the window of component (and condition) variability in which the assembly will meet specifications. i have no idea what you mean by "PERFECT" quality control, perhaps you could elaborate. i'm not sure what your experience is, but it's apparently quite different from mine.

Given that DBW systems are installed of bazillions of vehicles now, there must be data data on real-world failures leading to accidents, right? Have there been any major recalls for bad DBW software? In the last 10-15 years there have been a half million deaths on American roads. Were any the result of a DBW failure, or is that unknowable?

wcelliot wrote:

Ignorant wrote: Give in to your electronic overlords people.. They're better.

Fascinating how your political/economic opinions tend to color your engineering ones as well. At least you're consistent. ;-)

Right, and by that logic your political beliefs put you squarely into the, Relay cabinets were good for my forefathers and therefore they're good for me.. camp.

absolutely dreadful.

Or you could effectively say, Through my extensive experience of working in extremely crappy plants in maint, manufacturing, and product design.. I've experienced 100000x more mechanical failures than electronic ones.