Edmunds dealer partner, Bayway Leasing, is now offering transparent lease deals via these forums. Click here to see the latest vehicles!

Unintended Acceleration - Find the Cause

1383941434446

Comments

  • wwestwwest Member Posts: 10,706
    "..The answer is obvious..."

    No, it isn't...!

    How would YOU go about designing a brake over-ride system that does not interfere with those times the driver might well intend to closely overlap braking and throttle? Then there is the company that claims to have had brake over-ride for many years. I don't remember the exact design wording but I do seem to remember they made provisions in the design for intentional use of the brakes and throttle simultaneously.
  • busirisbusiris Member Posts: 3,490
    The Brake Override System operates when these conditions are met:

    a) The throttle opening is greater than 1/3

    b) Vehicle speed is above 5 mph

    c) Brakes are applied firmly

    d) The accelerator pedal is applied before the brake pedal


    So, in the Edmunds test, we have checkmarks beside A,B,C & D. If the brake over-ride feature was installed in the vehicle, why would the test results under full throttle show a 25%+ longer braking distance?

    Also, when you read reports that Toyota purchased suspect vehicles, and along with the floor mat official recall delays and refusal of access to EDR data at first - just makes you wonder.

    I agree, and you do make a valid point... in that Toyota's resistance to issue recalls/repairs initially suggests a company that is more concerned with profits than its customers and their safety.

    And, there are many other issues besides UA that just haven't gotten the same publicity.... Corolla electronic steering issues, Prius HID headlamp failures, etc.

    IMO, there's little question that Toyota's image is tarnished due to these mis-steps, but unless the company is run by complete idiots, management would certainly be able to figure out the test could be easily duplicated....cheaply.

    All one needs to do is find a used late model Toyota without the over-ride feature , install new rotors/pads, and off to a flat stretch of road he goes...

    Invest $20K or less , demonstrate Toyotas are death traps, and sell your story/test results to the highest bidder.

    A great way to make a living, if one can get the work...
  • busirisbusiris Member Posts: 3,490
    You're grabbing at straws....

    Neither you nor I am designing anything. Its already designed and implemented, and the specifications were shown a couple of posts back.
  • Mr_ShiftrightMr_Shiftright Member Posts: 64,481
    We can always blame the brake-override system for future UAs, too. :P
  • wwestwwest Member Posts: 10,706
    What is the comparative stopping distance starting at 60 MPH for the same car with and without cruise control engaged previous to the beginning of brake application..?

    With CC engaged the engine might still be operating at WOT right up to the instant the brake light switch circuit closes. Without CC the throttle will likely be fully closed once your foot reaches the brake.
  • wwestwwest Member Posts: 10,706
    As currently implemented we can certainly blame it for not preventing UA.
  • wwestwwest Member Posts: 10,706
    edited March 2011
    Listen to the engine NOISE at the 36-38 second point in the clip.

    Now listen for the expected level of engine noise at the 45-47 second mark.

    http://www.insideline.com/toyota/camry/2010/2010-toyota-camry-stuck-throttle-bra- - ke-test.html

    In the first case we must assume that the gas pedal is fully depressed but it is likely the throttle is far from wide open, maybe as little as 10%. Yet we hear LOTS of engine NOISE.

    The engine is operating totally without load so the rev limiter would be preventing the engine from WOT, probably via the DBW system.

    In the second case absent some sort of BOS the throttle would have been WIDE OPEN with the engine RPM being limited only by the transaxle torque converter stall point, 2,000 to 3,000 RPM. No induction noise even...?

    This 2010 Camry had a BOS system installed, puzzling part is why Dan Edmunds did not note same.
  • busirisbusiris Member Posts: 3,490
    It would appear that some folks simply don't know when to stop beating a dead horse.
  • frankok1frankok1 Member Posts: 56
    Wow if you are correct the others who did tests on braking (remember others - but not sure) then you have a hit on your hands. I would send a separate note suggesting a retest
    http://support.edmunds.com/contactus.aspx?type=error&altref=http%3A%2F%2Fwww.edm- unds.com%2F
    https://www.ideas.edmunds.com/Page/Home

    As for busiris replying to me:
    "Invest $20K or less , demonstrate Toyotas are death traps, and sell your story/test results to the highest bidder. "

    I was hoping NASA would do that for RARE UA cases with at least 200 times $20k - but it appears to be a poor job just on the basis of the vehicles and pedals they selected - and not using MISRA for computer code checks.
  • wwestwwest Member Posts: 10,706
    If that statement is meant to be somehow a rebuttal of my position on the UA stopping distance (no engine noise = no WOT) then you missed your target by a wide margin.
  • Mr_ShiftrightMr_Shiftright Member Posts: 64,481
    edited March 2011
    After careful consideration of everything I've read here and elsewhere, I'm pretty much satisfied that the case is closed, and no electronic malfunction is at fault for the UA incidents.

    I suspect the floor mats + driver inability to cope with the emergency is the reason, and in some cases, pure driver error.

    I do however, think that malfunctioning cruise control does happen in cars, but that it is easily shut off by tapping the brakes in all cases.

    So I guess I'm done with hashing this around personally. It's been most enjoyable sparring with everyone and exploring all the possibilities.

    Naturally I know that some of you are not satisfied, so carry on. Should I see anything that re-opens the case I will let you all know!

    Shifty, Visiting Host
  • wwestwwest Member Posts: 10,706
    edited March 2011
    "..easily shut off by tapping the brakes in all cases.."

    And if the brake light switch malfunctions, then what...?

    How many of us are so inured to "tapping" the brake to shut down CC that in a panic situation if it didn't work would we know to resort to the CC stalk...?
  • houdini1houdini1 Member Posts: 8,327
    I guess we all assumed that you would be completely up to date on "toe tapping". :P

    2013 LX 570 2016 LS 460

  • wwestwwest Member Posts: 10,706
    I'm so up to date I have a CC patent application pending.

    On I90 in ID I would often shut CC off as I was a bit nervous about entering those mountainous terrain curves with the CC set at a fairly HIGH speed.

    Lack of being in control, if you will.

    So I modified the system in my Porsche C4 (DBW) such that if I put foot pressure on the gas pedal the CC goes into "neutral" mode. Once I release foot pressure the CC automatically returns to normal operation.

    So now "I" can drive those curves even above the set speed if I wish but be much more comfortable that if I want/need to let off the gas slightly the car will instantly respond.
  • busirisbusiris Member Posts: 3,490
    After careful consideration of everything I've read here and elsewhere, I'm pretty much satisfied that the case is closed, and no electronic malfunction is at fault for the UA incidents.

    I suspect the floor mats + driver inability to cope with the emergency is the reason, and in some cases, pure driver error.

    I do however, think that malfunctioning cruise control does happen in cars, but that it is easily shut off by tapping the brakes in all cases.

    So I guess I'm done with hashing this around personally. It's been most enjoyable sparring with everyone and exploring all the possibilities.

    Naturally I know that some of you are not satisfied, so carry on.


    I'm it total agreement.

    More and more, at least for most rational examinations, it appears that the UA ship has sailed...
  • killer007killer007 Member Posts: 5
    All existing sudden unintended acceleration (SUA) investigations focus on the question of how the throttle is opened when drivers do not press the gas pedal. They fail to reveal a convincing cause of SUA. This submission here reports a novel investigation, in which we explored and confirmed a new possibility that the sudden unintended acceleration can be trigged when the throttle is closed. We find that the Evaporative Emission Control System (EVAP) can cause the sudden unintended acceleration when the throttle is closed.

    The nature
    There are two fuel-air supply paths in a modern vehicle. One is the main path including the throttle and fuel injectors, which are controlled by the vehicle computer and gas pedal, and the other is the Evaporative Emission Control System (EVAP), which is controlled solely by vehicle computer. The second fuel-air supply path is ignored by industry so far in investigating the reason of SUA. The EVAP system deserves a careful examination as a possible cause of the SUA, because existing vehicles lack an effective system to monitor the functioning of the EVAP system.

    In a typical existing EVAP system showed in Figure 1, the gas vapor is temporarily trapped in the Charcoal Canister, when the vehicle stops or works in a condition that is different from the predetermined criteria. When the vehicle is running in a ‘good’ condition that fits the criteria stored in the vehicle computer (PCM), the PCM actuates the Purge Solenoid Valve, and then the trapped gas vapor is sucked by the vacuum of the intake manifold into the engine and is burned there. By design, the EVAP system has serious potential safety problem, That is, if the liquid gas in the gas tank enters in the Charcoal Canister for whatever reason, the EVAP system will automatically supply the engine with highly dense gas-air mixture once the Purge Solenoid Valve is actuated, and then the vehicle will suddenly experience unintended acceleration, ineffective brake and other unwanted consequences.

    The nature of the unintended acceleration problems is the defective design of the Evaporative Emission Control System (EVAP) shared by the vehicles of all automakers. All vehicles have been required of equipping with the EVAP system since 1971. Therefore the SUA is a common problem for all automakers.

    The Plausibility

    It is plausible in realistic vehicle running environments for the liquid gas in the gas tank to enter the charcoal canister of the EVAP system. Following is a partial list of possible scenarios:
    • People over-refill the gas tank, either intentionally or accidently;
    • Vehicle experiences heavy bump;
    • Gas tank is squeezed by accidents;
    • Fuel in the tank expands due to the temperature effects, because there is big temperature difference between the outside air and the underground storage of gas station in some days and some places.
    • The valves supposed to prevent liquid gas from entering into charcoal canister are broken, like the fuel shut-off valve.
    • Other cases.
    Both the vehicle computer and the driver do not know whether the liquid gas enters into the charcoal canister of the EVAP until the vehicle experiences the unintended acceleration. Since the unintended acceleration is so violent that many drivers cannot deal with it appropriately and promptly.

    BY Zhongkui Wang
    PHD in Syracuse University
  • wwestwwest Member Posts: 10,706
    Okay, you have a good case for how fuel might enter the intake manifold, but from where does the air/oxygen come to run the engine at such great torque levels....?
  • wwestwwest Member Posts: 10,706
    edited March 2011
    "..The EVAP system will automatically supply the engine with a highly dense gas-air mixture once the PURGE solenoid valve is actuated..."

    Where does the air come from if the throttle plate isn't WIDE open...?

    And with the throttle closed how do we produce enough heat via compression to ignite a DENSE mixture...?

    Pour all the liquid fuel you wish into the intake manifold and thereby into the cylinders but absent enough air/oxygen the spark plug will not be able to ignite the COLD fuel.

    Inject fuel into a diesel engine without the "throttle" FULLY open and you get NO "ignition".
  • frankok1frankok1 Member Posts: 56
    Did you submit it to Edmund's million dollar contest that required proof with test data that they would then confirm? (they must have relied on Mr. Shiftright's opinion to take a chance with the 1-million)

    Did you send the idea to NASA? - they reviewed a few others' ideas

    If not contact the other group investigation of all UA causes
    http://www.trb.org/PolicyStudies/UnintendedAccelerationStudy.aspx

    They just had meetings last week but there is contact info
  • steverstever Guest Posts: 52,454
    A bit over my head. :shades:

    "How might such a rare event occur? The National Academy of Sciences panel heard one theory Thursday from Ronald A. Belt, an electronics expert who holds a doctorate in physics from the University of Notre Dame.

    Belt, who works at Honeywell but presented his report independently, suggested that a kind of short circuit called a "latch-up" could cause a Toyota's electronic throttle to open and stay open - without the gas pedal depressed and outside the control of the vehicle's electronic control unit (ECU). Another electronics expert likened a latch-up to "a faucet getting jammed in the open position."

    Lyons dismissed Belt's report via e-mail, saying Belt "admittedly guessed at the design of Toyota's ECU and, as a result, has made a number of incorrect assumptions that have led him to draw erroneous conclusions."

    Philly.com
  • wwestwwest Member Posts: 10,706
    edited March 2011
    First a definition: SCR, Silicon Controlled Rectifier. Semiconductor device wherein mostly undesireable substrate "latch-up" has advantages. SCRs are commonly found in relatively high power switching circuits.

    Take any +5 VCC volt logic IC (Integrated Circuit) and reverse bias the chip substrate via any input pin and the circuit will "latch-up", allowing the negative circuit current to flow continously until the negative voltage source is removed.

    Provided the reverse current flow is somehow limited such that no permanent damage results then after a power cycle the circuit might return to fully operational status and remain so for the next 100 years.

    Remember that I theorized that the Haggerty engine UA resulted from a short internal to the gas pedal hall effect sensor..? Look at page 21 of Belt's presentation.
  • srs_49srs_49 Member Posts: 1,394
    edited March 2011
    Latchups can also be caused by high energy particles. They are part of a general class of upsets known as Single Event Effects (SEEs). These are things that satellite and spacecraft designers have had to worry about for over 40 years.

    As wwest said, Latchups (Single Event Latchups, or SELs) in general can be destructive or non-destructive, usually depending of the amount of current that can flow when the parasitic SCR is triggered. A non-destructive latchup goes away, or resets itself (the SCR is turned off) when power is removed. Destructive latchups permanently damage the device.
  • wwestwwest Member Posts: 10,706
    "...high energy particles.."

    The more transistors we are able to put on a given chip size the more the parts will be subject to SEL due to such things as ALPHA particle radiation.

    PING.....!

    Back maybe 30 years ago Intel spend TONS of money on R&D trying to find out why their memory chips were so subject to SEF, Single Even Failures. It turned out to be totally random RADIATION effects.

    So today my company only ships systems with FULL ERCC, memory fault error detection and correction. I am currently of the undestanding that even the "X86" cache memory structures now have "on-board" ERCC.

    But what's to be done about, with, the random logic structures surrounding these processors as the geometry grows smaller and smaller...?

    People ask why the "spat" of UAs and then nothing....strong storm on our sun???
  • srs_49srs_49 Member Posts: 1,394
    But, we have been at a sun spot minimum for the last 4 or 5 years, right about the time as the bulk of the UA incidents. Lower sunspot count = quieter sun. Of, course, that doesn't allow for solar flare events, which can occur any time.
  • busirisbusiris Member Posts: 3,490
    Its a bit off topic, but I thought those from the scientifically-oriented crowd might enjoy this article that came from today's USA Today...

    http://www.usatoday.com/news/opinion/forum/2011-03-09-column09_ST1_N.htm
  • srs_49srs_49 Member Posts: 1,394
    Doesn't surprise me.

    Lawyers for plaintiffs, particularly those involved in lawsuits want jurists who are going to be receptive to their arguments - malleable, to some extent. Engineers, scientists, and the like tend not to fit that mold.

    That's why in many cases there is so much shopping around for the right venue in which to file a lawsuit, or to try and get a case moved to a more plaintiff-friendly setting.
  • busirisbusiris Member Posts: 3,490
    My thoughts exactly...
  • killer007killer007 Member Posts: 5
    Thanks for reading.
    As you know, there are three hoses connected with the charcoal canister of EVAP, one to the engine, one to the gas tank, and another one to the outside air. When EVAP purges the charcoal canister, the outside air will go into the charcoal canister and bring the fuel there into the engine. Therefore, the EVAP system can supply the engine the fuel and air simultaneously.

    Your further feedback will be appreciated.
  • killer007killer007 Member Posts: 5
    Thanks frank.
  • wwestwwest Member Posts: 10,706
    We're talking about a FWD vehicle operating with WOT engine that can overcome FULL braking and still sustain speeds as high as 80-100 MPH.

    I sincerely doubt that the EVAP system can provide enough fuel flow for that effort let alone an A/F mixture close enough for normal spark ignition.
  • killer007killer007 Member Posts: 5
    I agree your doubts--all those conditions that you are aware of must happen simultaneously. That is why SUA is a rare event.
  • steverstever Guest Posts: 52,454
    A bit off-topic, but the proposed computer breach is interesting.

    "The new report is a follow-on to similar research these experts conducted last year, which showed that cars were increasingly indistinguishable from Internet-connected computers in terms of vulnerability to outside intrusion and control. That project tried to show that the internal networks used to control systems in today’s cars are not secure in the face of a potential attacker who has physical access to the vehicle."

    Researchers Show How a Car’s Electronics Can Be Taken Over Remotely (NY Times)
  • srs_49srs_49 Member Posts: 1,394
    That's not surprising. But the operative phrase there is "physical access". Also, in the case of the experiments run last year, it took those researchers weeks to come with a set of commands to cause the system to do their bidding.

    There are not many systems that are truly secure if someone has physical access to the network. Maybe something that comes out of NSA...

    The history of security leaks shows that the weakest point is usually the human. The Wiki Leaks of all the state department cables is a good example.
  • steverstever Guest Posts: 52,454
    edited March 2011
    Good point about the weakest link. But the "direct physical access" to the network can be wireless if you get within range and crack the Bluetooth or, say, remote start module.
  • steverstever Guest Posts: 52,454
    edited March 2011
    I guess the point is that when these systems can be "physically" hacked, who's to say the same "unprotected" systems can't be adversely affected by stray electrons or a whisker short or coding issues?
  • kernickkernick Member Posts: 4,072
    The history of security leaks shows that the weakest point is usually the human.

    Not just security, but design and quality defects. Rushing to stay on schedule, using the lowest cost parts and materials, plain human error, and disgruntled employees are major sources of problems. I think the 1 fictional character we all remember who helped turn a highly structured system into chaos, was in the book & movie - Juraissic Park.
  • kernickkernick Member Posts: 4,072
    That's exactly been 1 of my points. Show me a piece of software or hardware that is fault-free or not prone to failure. Where is that PC, server, or program? I don't see Intel, MS, or the Pentagon having such fault-proof hardware or software.

    In life it's not about IF something will fail; it's WHEN. The improbable and unplanned happens everyday; I still like the recent problem only-Mazda has with spiders; that caused them to have a recall.

    To say that Toyota or NASA thought of all the possible failure modes that causes UA is statistically invalid. Like archaeologists they dug some holes in the Sahara desert where they thought they thought they could find something. And if Toyota did have some specific information, and did find something, they might never had made it public, and just fixed it. Several billion $'s will make people do unethical things.
  • steverstever Guest Posts: 52,454
    Apparently Hondas have spider problems too, but Honda just did a TSB instead of a recall.
  • frankok1frankok1 Member Posts: 56
    edited March 2011
    I got on line to post the same thing. It's been going on for a while - just Google: hacking "automobile computer" for some - 12,000 webpages.

    Hope this wasn't the real reason for RARE Toyota cases that no doubt had to be electronics. I just changed my original SYNC password in my (great) 2010 Ford Flex as I haven't since I got it. It is a Microsoft system - darn. I accessed my account and it says there is an update to download from my computer and install it. More work for a non-GEEK. I'm having trouble setting up my new printer.
  • kernickkernick Member Posts: 4,072
    And we're seeing that man's machines aren't quite so "perfect" when you expose them to some vibration and water, eh? Apparently Godzilla (chaos/Nature) has something against Japanese machines.

    "History shows again and again how Nature points out the folly of man".
  • frankok1frankok1 Member Posts: 56
    NASA gave a summary of their testing to – 03/11
    http://onlinepubs.trb.org/onlinepubs/UA/030311Kirsch.pdf

    Last page shows that they found errors giving smaller than full throttle openings - they just couldn't find the cause of "large unintended accelerations".

    Those smaller ones could explain many of the low crash incidents in parking lots etc. There is no reason why older drivers are misapplying pedals in Toyotas, but not in many others such as GM's that have had very very low numbers of cases per vehicle sold.
  • plektoplekto Member Posts: 3,738
    Sure there is. I mentioned this more than a month before this "contest" was started. Hall Effect sensors produce errors and fail in fundamentally different ways than potentiometers. They also are not nearly as easy to design a monitoring system around, mostly because a potentiometer has only two conditions to really worry about - a short or a dead connection, both of which are dead-simple to detect.

    GM uses potentiometers. As expected, UA or throttle "issues" are extremely low even though the parts do "wear out" more often.(I'd consider a Hall Effect sensor that's giving bad data to be "worn out" as well, but I guess they mean breaking apart "worn out" for a recall)
  • wwestwwest Member Posts: 10,706
    Potentiometers are like brake pads, the frictional contact surface(s) will eventually wear/erode away. Hall effects sensors have the advantage that they do not require a frictional contact between the two sensor elements.

    GM uses potentiometers for the same reason Chrysler switched the Jeep from front CV's to U-Joints......CHEAP.....!!

    But then it does take a decent design engineer, HIGH COST engineer, to come up with a good working Hall Effect sensor set.
  • plektoplekto Member Posts: 3,738
    edited March 2011
    Everything will eventually fall apart. The cost of both parts (GM vs Toyota throttle sensors) are nearly identical as well. GM isn't doing it for cost, but because the system is nearly foolproof. They certainly looked into Hall Effect sensors and made the correct decision that it would be horrendously hard (read - lots of time and money) to design a 100% guaranteed fail-safe system for it.

    A first year employee with a EE degree could design a foolproof system for a potentiometer by comparison.

    Potentiometers provide a constant and easy to monitor voltage output. If it drops to zero resistance or goes to maximum and stays there, then they know that there's a problem. In simpler terms for readers who aren't up to speed on electronics and such, it would be like your circuit breaker blowing or shorting out and sending 240V throughout your house - either scenario is *instantly* known by the occupants every time.(no lights or every bulb blows out after becoming 4x brighter for 1-2 seconds)

    The instances where the contacts are getting worn also give months or weeks of warning to the driver. This small-scale erratic and/or sluggish behavior (often without the driver feeling it yet) triggers the "service mode" of the vehicle with nearly 100% accuracy. Replace it and you're good for another half dozen years. Nobody gets hurt or killed, either.

    note - CV joints are far less robust than U-Joints and cost far more to replace as well.
  • srs_49srs_49 Member Posts: 1,394
    I have never seen a potentiometer used in a piece of equipment I was associated with to sense angular position - not a one. The reason being is that devices that relied on physical contact (like potentiometers) were deemed to be less reliable than non-contact solutions. To be fair, we were not under the same cost constraints as a consumer product (like a car).

    Today we use Hall-Effect sensors and optical encoders for rotational measurements. Heck, even 43 years ago when I was first getting started, we were using devices called synchros and resolvers to measure angular position. No, they weren't cheap - much more expensive than a simple pot. But the inherent reliability improvement was deemed to be worth the cost.
  • plektoplekto Member Posts: 3,738
    Sure you see them - your game console controllers have them in them. As does a gaming steering wheel. So do many other places that you'd not initially suspect, especially in consumer electronics. Because they are very robust and much easier to design something as simple as a throttle or position around.

    True, high-end aircraft systems and so on wouldn't use them, but then again they have proper designs as well, with multiple backups and fail-safes. A typical car apparently has only some code and a secondary sensor at best to verify that everything is OK. Given a half-baked fail-safe that isn't one at all, I'd chose the potentiometer based design or better yet a traditional throttle cable every time.
  • wwestwwest Member Posts: 10,706
    "...game console controllers.."

    Last I knew there was NOTHING "mission critical" about a game controller...

    Lowest common denominator insofar as cheap, CHEAP, CHEAP...!
Sign In or Register to comment.