thruthefence
Aerospace
Regarding the Toyota troubles currently in the news; any chance of this being a "drive-by-wire" software (or hardware) issue, as opposed to a strictly mechanical, binding throttle linkage,ect problem?
Eng-Tips is the largest engineering community on the Internet
Intelligent Work Forums for Engineering Professionals
-
Congratulations waross on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!
Toyota uncontrolled throttle 6
- Thread starter thruthefence
- Start date
- Status
Back when I first heard the story about the runaway toyota's, I first thought it was like a sci-fi computer take over of the whole car, and the driver was locked out of all the input, such as turing the key off, flipping into neutral, stepping on the brakes. How far off could this be with nowadays computer integration into all operating aspects of the newer automobiles?
MacGyverS2000
Electrical
That assume an automatic, correct? 'Cause I guarantee I can mechanically overrev a manual...ChrisDuncan said:
Dan - Owner
ChrisDuncan
Mechanical
""The wife's LS won't select neutral without pushing the little thingy....Then it quite easily goes all the way to reverse.""
then I should modify my statement to say, the Asian vehicles I'm familiar with go to neutral without the release and reverse only with the release. From a safety standpoint it seems to make sense.
I think the CHP veteran with the Lexus was a paddle shifter auto and he was possibly in manual mode which means neutral is over and up, not just up.
""That assume an automatic, correct? 'Cause I guarantee I can mechanically overrev a manual...""
Yes, and even with an auto it's probably going to go over one time, if it's floorboarded and put into neutral. But it's not going to stay there right?
then I should modify my statement to say, the Asian vehicles I'm familiar with go to neutral without the release and reverse only with the release. From a safety standpoint it seems to make sense.
I think the CHP veteran with the Lexus was a paddle shifter auto and he was possibly in manual mode which means neutral is over and up, not just up.
""That assume an automatic, correct? 'Cause I guarantee I can mechanically overrev a manual...""
Yes, and even with an auto it's probably going to go over one time, if it's floorboarded and put into neutral. But it's not going to stay there right?
GregLocock
Automotive
That's the event they calibrate for, I don't think it'll really hurt the engine.
Cheers
Greg Locock
New here? Try reading these, they might help FAQ731-376
Cheers
Greg Locock
New here? Try reading these, they might help FAQ731-376
"I was hoping this would be the death of DBW, but clearly the legislative reflex of adding yet another layer of crap will prevail"
Luddite! Loom Smasher!
Out of interest whats your issue with DBW?
The cable operated throttle just does not fit in with modern EMS/Engine Hardware concepts and is grossly outdated on passenger vehicle engines.
From my point of view the problem is nothing to do with the actual engine management system per se but rather is a system integration issue.
By this I mean that the ECU was most likely doing exactly what was asked of it - which was exactly what the driver didnt want but the issue was the communication of thie wish ie the pedal mechanism. The issue laying with the pedal linkage sticking and keeping the throttle pot at some setting not desired by the driver.
MS
Luddite! Loom Smasher!
Out of interest whats your issue with DBW?
The cable operated throttle just does not fit in with modern EMS/Engine Hardware concepts and is grossly outdated on passenger vehicle engines.
From my point of view the problem is nothing to do with the actual engine management system per se but rather is a system integration issue.
By this I mean that the ECU was most likely doing exactly what was asked of it - which was exactly what the driver didnt want but the issue was the communication of thie wish ie the pedal mechanism. The issue laying with the pedal linkage sticking and keeping the throttle pot at some setting not desired by the driver.
MS
MikeHalloran
Mechanical
It may well be that the ECU has no bugs, and does exactly what it's been told to do, based on what >it thinks it knows< about what's going on.
The 'system integration' issue is that, while they're much better than they used to be, the connectors and internal interconnects used in automotive systems are not yet up to our expectations.
Mike Halloran
Pembroke Pines, FL, USA
The 'system integration' issue is that, while they're much better than they used to be, the connectors and internal interconnects used in automotive systems are not yet up to our expectations.
Mike Halloran
Pembroke Pines, FL, USA
Most markets require a red E-stop button on motorcycles placed in a standard location. MSF driver courses istruct turning off the engine using this 'kill' switch instead of the key precisely to get riders practiced with it's use and location.
As more newer MCs are including DBW throttles, one hopes the E-stop switches remain hard-wired rather than sending a meek 'cease & desist' to the ECU!
As more newer MCs are including DBW throttles, one hopes the E-stop switches remain hard-wired rather than sending a meek 'cease & desist' to the ECU!
I guess I'm "hard wired" <snicker> and, too old to change...
I was not aware the modern motorcycles were DBW, too. Just as well I probably will never own another one after I sell my '48 Norton this summer.....The "kill" switch (sic) is the compression release!
Rod
I was not aware the modern motorcycles were DBW, too. Just as well I probably will never own another one after I sell my '48 Norton this summer.....The "kill" switch (sic) is the compression release!
Rod
Yes, in addition to the 'drivers' for cars, MCs apparently benefit in FI modulation at part throttle with some form of electronic butterfly control. Preliminary applications featured dual butterfly valve throttle bodies with one cable/driver actuated and the other ECU. Now full DBW, mostly sports bikes so far, but no doubt will cascade.
For those interested the David Gilbert preliminary report is available here at the attachment below. In summary he was able to connect the two Hall outputs together with a resistor without the ECU noticing and then by connected one of those signal lines to the particular Vcc for the backup sensor he was able to get full accel.
It implies that two near shorts would have to occur.
It implies that two near shorts would have to occur.
Yes you can shift to neutral, but, who is really commmanding the shift valves? You or the computer system?
When you turn off the ignition switch, who is commanding fuel and ignition off? You or the computer system?
When you push on the gas pedal who is commanding it? You or the computer system?
When you push on the ABS brakes, who is commanding it?
You or the computer system?
Some nice detailed information on the systems would be nice.
When you turn off the ignition switch, who is commanding fuel and ignition off? You or the computer system?
When you push on the gas pedal who is commanding it? You or the computer system?
When you push on the ABS brakes, who is commanding it?
You or the computer system?
Some nice detailed information on the systems would be nice.
JayMaechtlen
Industrial
One comment on dbw that I haven't noticed elsewhere:
Makes it super-easy to do cruise control- gets rid of the vacuum motor and controls. That has to be a nice part of the cost savings.
Regards
Jay
Jay Maechtlen
Makes it super-easy to do cruise control- gets rid of the vacuum motor and controls. That has to be a nice part of the cost savings.
Regards
Jay
Jay Maechtlen
thomasthethird
Computer
This is a far more serious problem than a lot of people may think. Just fixing the gas pedal linkage is not the complete answer. What happens when we have a similar problem with steer-by-wire?
There are two areas requiring additional discussion.
Surely there should have been high level behaviour boundary software present? Given the number and variety of accidents it is obviously there was not. ( I have worked on embedded sysems design).
Next why did not any of our departments of transport safety specialist catch this problem. They have known for years that software was going into cars. When new software is used for drive-by-wire, steer-by-wire or stability control then their is a public safety issue. Transport safety specialists MUST develope a capability to certify this software. They should have emulators/simulators that enable them to exercise in-boundary and out-of-boundary driving conditions.
As I said this is not just a Toyota problem. It is an early warning of major downstream problems unless changes are made now.
There are two areas requiring additional discussion.
Surely there should have been high level behaviour boundary software present? Given the number and variety of accidents it is obviously there was not. ( I have worked on embedded sysems design).
Next why did not any of our departments of transport safety specialist catch this problem. They have known for years that software was going into cars. When new software is used for drive-by-wire, steer-by-wire or stability control then their is a public safety issue. Transport safety specialists MUST develope a capability to certify this software. They should have emulators/simulators that enable them to exercise in-boundary and out-of-boundary driving conditions.
As I said this is not just a Toyota problem. It is an early warning of major downstream problems unless changes are made now.
GregLocock
Automotive
a) steer by wire is currently illegal, and is 10-15 years away at least IMO
b)no electronic problem has been confirmed to date- most incidents have been pilot error, some have been mats. Very few are unexplained, those might be hardware/software related, we just don't know.
c)apparently your expertise allowed you to miss point b, ie your expertise is borderline irrelevant
d) They don't have that expertise. In most cases they don't even see a new car before it is launched. You would need to certify the system, not just the software. They don't have the emulators. They don't have the people. They don't have the cars. Apart from that they are all set. Would this certification process include the effect of jamming the pedal with aftermarket mats? or drivers pressing increasingly hard on the wrong pedal?
It is not a Toyota problem, it is a combination of operator error and the media enjoying a beatup, admittedly aided and abetted by an approach to PR that seems comically inept in retrospect.
Cheers
Greg Locock
New here? Try reading these, they might help FAQ731-376
b)no electronic problem has been confirmed to date- most incidents have been pilot error, some have been mats. Very few are unexplained, those might be hardware/software related, we just don't know.
c)apparently your expertise allowed you to miss point b, ie your expertise is borderline irrelevant
d) They don't have that expertise. In most cases they don't even see a new car before it is launched. You would need to certify the system, not just the software. They don't have the emulators. They don't have the people. They don't have the cars. Apart from that they are all set. Would this certification process include the effect of jamming the pedal with aftermarket mats? or drivers pressing increasingly hard on the wrong pedal?
It is not a Toyota problem, it is a combination of operator error and the media enjoying a beatup, admittedly aided and abetted by an approach to PR that seems comically inept in retrospect.
Cheers
Greg Locock
New here? Try reading these, they might help FAQ731-376
thomasthethird
Computer
Hi Greg
Did not mean to imply it was a sortware problem, ie your point (b).
We can take advantage of the embedded software to read sensors. Lets say we do a dynamic read of speed, brake presure clutch position and gas pedal pressure, and compare the data to recent past data. From the results we can determine if a driver/car is/are exibiting signs of acceptable behaviour.
The point is with sensors, correctly located, and appropiate software I think we can increase the safety of the car by detecting some mechanical problems. It gives us a capability we never had with 'just mechanical contrivences'.
I think my experience is relevant. GM obviously thought so. I gave a talk to a boardroom full of their execs in Detroit. The topic, the future of computers in the automotive and appliance industries.
Thanks for your comment.
Did not mean to imply it was a sortware problem, ie your point (b).
We can take advantage of the embedded software to read sensors. Lets say we do a dynamic read of speed, brake presure clutch position and gas pedal pressure, and compare the data to recent past data. From the results we can determine if a driver/car is/are exibiting signs of acceptable behaviour.
The point is with sensors, correctly located, and appropiate software I think we can increase the safety of the car by detecting some mechanical problems. It gives us a capability we never had with 'just mechanical contrivences'.
I think my experience is relevant. GM obviously thought so. I gave a talk to a boardroom full of their execs in Detroit. The topic, the future of computers in the automotive and appliance industries.
Thanks for your comment.
MacGyverS2000
Electrical
Isn't that the current intention? Appropriate sensors, appropriate software, etc.? It's obviously not working as well as many had hoped, mostly because determining what's "appropriate" is partially a moving target and partially unknown or untestable.thomasthebird said:
Dan - Owner
patprimmer
New member
The software attached to my feet, hands, middle ear and eyes tends to indicate to me how my mechanical systems are working, but yes, cars are getting safer for the average shmo.
I still resent that I cannot legally use a 6 point full harness because it is not self adjusting just because some other dopes do not adjust their belts correctly.
Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &
for site rules
I still resent that I cannot legally use a 6 point full harness because it is not self adjusting just because some other dopes do not adjust their belts correctly.
Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &
for site rules
GregLocock
Automotive
If you mean there should be an override such that hitting the brake cancels the throttle signal, um yes, that does seem an obvious thing to do in retrospect and is already fitted to many cars.
It is not quite as simple as that, since simultaneous application of brake and throttle is a valid control state for some drivers in some conditions, and all drivers in one condition, unless they live in Holland.
Cheers
Greg Locock
New here? Try reading these, they might help FAQ731-376
It is not quite as simple as that, since simultaneous application of brake and throttle is a valid control state for some drivers in some conditions, and all drivers in one condition, unless they live in Holland.
Cheers
Greg Locock
New here? Try reading these, they might help FAQ731-376
- Status
Similar threads
- Locked
- News
- Locked
- Question
- Locked
- Question