Art161

I wonder how reliable the black boxes are in our cars. For example, my Owner's Guide states that the "event data recorders" are capable of collecting and storing data during a crash or near crash event. Included in that is how far (if at all) the driver was depressing the accelerator and/or the brake pedal.

If the black boxes merely record what the sensors think is happening, then I say that a lot of "driver error" cases may not be that at all.

Example: I was in a parking space with my foot only on the brake pedal, not on the accelerator pedal. The car was in Drive and idling normally. When I shifted into Park, with my foot still only on the brake pedal, the engine went WOT. I didn't crash into anything, since the gear shift lever was in Park, and that operated correctly.

If I had been in gear and crashed, would the black boxes have recorded that I was depressing the accelerator pedal? Or do the black boxes get their data from different sources than the sensor(s) that told the car to go WOT?

BTW, Ford Motor Company said the event could not have happened the way I described because of all the safeguards built into the car. But they weren't there at the time of the event. Ford engineering, Ford customer service, and NHTSA were absolutely useless.

jz78817

sensors don't "think." they output a signal based on whatever it is they're supposed to monitor. the control module they're connected to monitors those signals, and if the signals they're receiving don't make sense (even if it's within the sensor's normal output range) it'll set fault codes.

Only if you were.

the "Black boxes" aren't separate modules, just logging capability built into the control and safety-related modules already in the car (PCM, BCM, RCM, etc.) the PCM is already monitoring all of these sensors to keep the engine running; for example since you have electronic throttle control (ETC) the PCM has to know how far you are pressing on the gas pedal in order to know how far to open the throttle, and then uses the throttle position sensor (TPS) to make sure the throttle opens the intended amount. If the two disagree you'll get a check engine light, an ETC Fault warning light, and it'll go into limp-home mode.

so what the "black box" functionality means is that these modules keep a running log of the sensors they're monitoring, and store the last x minutes of data in the case something happens which causes airbag deployment. So then the logged data can be downloaded and investigators can see things like if any modules had DTCs logged, or "how far was the accelerator pedal pressed? How far was the throttle open? Do these agree? Was the brake pedal being pressed?" Then they can look at things like steering angle sensor data to see how much and at what rate the steering wheel was turned, and get accelerometer data from the RCM to figure out what direction the car was going and how hard the driver was braking (if at all.)

what do you expect them to do? if they ran diagnostics on the car and didn't find anything which would indicate support for your claim, there's not much they can say. if there truly was an uncommanded throttle opening then there should be evidence something went wrong. The pedal assembly has redundant Hall effect sensors which are what tells the PCM how far you have the pedal pressed, if one disagrees with the other(s) then you get an ETC fault. the likelihood of multiple Hall sensors giving the same erroneous data is pretty low. And NHTSA's not going to get involved based on one report with no evidence. If the engine revved up as you say (I'm curious how you know it was at WOT) then the overwhelming probability is that something was pressing or holding the pedal down. Whether it was sticking, or something in the footwell was blocking/binding it (floor mat) or you were unknowingly pressing down on it with the side or heel of your shoe. I point out that back when Toyota was going through the "unintended acceleration" stuff, there was this guy who went on either Dateline or 60 Minutes claiming he could make some Toyota car accelerate without stepping on the gas. He was full of crap; in order to get the car to do that he had to hack up the wiring harnesses and induce a quadruple- or quintuple-fault in the control system. The probability of that actually happening in the real world is so small as to be practically nonexistent.

Chromeshadow

This is true but it does exist. Both Hall sensors get their 12 power from the same source. I can make a 12 V pulse or a large current pulse in the wire next to the sensor wire that makes them both give the same error. All car manufacturers do extensive tests to try to make sure this does not happen in real life.
There is no way to prove that a data logger is 100% accurate. They are considerably more reliable than people, but they are not always correct. When they do make a mistake, you will be on the defensive trying to prove that they made the mistake and not you, and you will lose.

Art161

Thanks for your comments. (Also, thanks to Chromeshadow. Yeah, people believe computers more than they believe other people.)

I expected the Ford engineers to analyze what could cause such an event and fix it.

Check the NHTSA website. There are multiple reports of incidents similar to the one I reported.

I could tell it was at WOT from the sound of the engine and the tach needle going up to redline.

When the engine went to WOT, I lifted my foot off the brake pedal, so my foot was on neither the brake pedal nor the accelerator pedal. I "blipped" the accelerator pedal and the rpm dropped to normal idle. The accelerator was not physically all the way to the floor; it appeared to be in its normal position for not being depressed. The mat was not a factor. It is a factory mat, lies flat on the floor and the retainer was in place.

Chromeshadow

Art,
This actually sounds pretty serious.
Do you know if there was anything else running that could effect it? Were you close to a garage door opening, receiving a cell phone call, near an airport, fire truck, ambulance, police car or medical office?
Opening your car window? Were you near underground utility power wires? ( if there are no power lines in sight, they are underground). Do you have any aftermarket items in your car...a different radio, speakers, amps etc? Cheap cell phone charger?

Art161

Up to now, I have been factual in what I stated. Now we are getting into the realm of "I don't know, I'm not sure, I don't remember." I don't remember if either the radio or the a/c were on. I'm pretty sure the windows were closed. I had them closed during my trip, and was going to exit my car after I turned the ignition off, so there would be no reason for me to open the windows when I stopped. I was at a rest stop alongside I-80 in the Sierra Nevada, so it's pretty rural. There was probably a California Highway Patrol car somewhere in the vicinity because I saw a CHP officer (not in his car) talking with someone. I am sure I was not using a cell phone, because I don't have one. I don't remember one way or another about power lines. There is a place where the big rigs park, but it's probably a few hundred yards from where I was. I wouldn't think that any of their radios caused my issue. I also didn't notice if there was an airplane overhead. My car is stock except for replacing the cheap plastic covering for the dash power point with a nice piece from MGW.

Thanks for your comments.

moosestang

If it went WOT in park, it would have immediately bounced off the rev limiter, did that happen? Did it only last for a second?

jz78817

have you done so, or are you just presuming you could do so? do you know the source and load impedances of the sensor circuit? how would you source enough current to do so yet not blow a fuse?

Art161

The tach needle went up into the lower part of the red area of the tach, but I don't know if it was bouncing or holding steady. It was definitely in that area for more than a second. I don't know how long it took me to think to blip the throttle, which got the needle down to idle rpm, and the engine wasn't roaring any more. Maybe this was five seconds. IIRC, of the drivers who filed a similar complaint with NHTSA stated that he also blipped the throttle to get back to normal rpm.

tbear853

I was a Trooper for over 31 years, retired a few years ago. For the last 20 of those years I was also assigned to a reconstruction team as part of my duties and attended a great deal of training and taught some. In all that, I also attended a couple week long courses a few years apart at our academy taught by W. R. "Rusty" Haight. http://www.reddit.com/r/IAmA/comment...est_dummy_ama/ Also, his company:
http://www.collisionsafety.net/
You might recall him appearing on Jay Leno a couple times and doing crashes in the back lot.

Rusty has spent a lot of time testing "black boxes" and the reliability of their decisions and recordings, he sets up crashes, knows the answers before the crash, then does the crash and looks at the black box to see if it performed correctly. Last I talked to Rusty, he had over 700 crashes and zero failures and now he's north of 900.

I also, over the years, had many many opportunities to compare what was downloaded to statements, skid marks, yaw marks, vector analysis, using drag coefficients, weights, etc .... even one where the crash occurred while the "crasher"s speed was being checked by radar.

You should know that no reconstructionist worth his salt is going to pin it all on a few seconds long record down loaded off a vehicle's EDR. He is going to do the tests, weigh things, make the evaluations of damage, the scale diagrams, grade measurements, skid and yaw mark evaluations, and vectoring and apply math, he's going to interview people who were involved or saw something, and it builds a picture, and a download from a EDR is only a small part of it. Standing alone, EDR data means might near next to nothing.

Air bags were mandated by the Federal Govt. The EDRs were not so mandated and their inclusion in air bag equipped cars was not for LEOs to use. There are two main reasons for them, one being simply the manufacturers covering their butts so that after a deployment, if questioned in a civil proceeding, a record exist that showed why a air bag or belt pre-tensioner was fired ... or not fired. The other reason was the manufacturer's wanting a way to judge performance over time in real world accidents to aid them in development of better versions as technology progressed.

However, once it was known that the info often existed, it was only a matter of time until the means were made available to LE to obtain that info as after all .... it is "evidence" which has a "value" in a criminal court (traffic court is such, even if the violations are classed as misdemeanors or infractions .... they are not civil violations unless specified in code) whose "value" is determined by the court (judge usually). Some circumstances may require a search warrant, sometimes not so.

EDRs constantly write info into volatile memory, it's only locked when a deployment decision is reached. Any deployment will almost always, with few execptions, also require a replacement of the EDR simply because the old one is no longer able to record again. Car get's fixed, it gets a "new slate".

To answer the question, I have no doubt but that a EDR is more reliable than a human witness .... but I also will say the next one that records may be wrong. They aren't biased, they just "do it". They aren't subject to suggestion, they just preserve (usually, there are some circumstances where a part of the record can be screwed up.) what is read by them as sent to them.

As to a Mustang's engine going to part throttle or wide open when the driver shifts from D to N in a rest area when there is no impact and thus, no deployment .... there's likely nothing to download and read.