Aren’t All Technical Service Bulletins Public?

March 23, 2010

CNN reported today that a Technical Service Bulletin (TSB) issued by Toyota in 2002 calls for a software patch to fix a calibration table on Toyota Camrys.    You are probably aware that the 2002 Toyota Camry was the first Camry to use Electronic Throttle Control.  In addition, there was a spike in reported sudden unintended acceleration incidents between 2001 and 2002.

One may rightfully wonder about why the existence of this TSB is not evidence of Toyota knowing about possible software problems in its electronics.   I wonder more about how such a TSB is not publicly available.  Interestingly, NHTSA had seen the TSB and has a short report on it.

Given that the TSB was not public, it would seem that only those 2002 Camrys, which were taken to a Toyota dealer, could have been patched.    Others could still be using the mis-calibrated tables in the firmware on the engine control module.


Event Data Recorders of a Different Kind

March 22, 2010

On its website, Toyota has a video clip providing its analysis of the runaway Prius in Southern California.  They obtained data from the Prius, demonstrating that the driver of the Prius was alternating between braking and accelerating 250 times.  They point out that the data were not obtained from an event data recorder.   They mean that the data did not come from the vehicle’s ‘black-box’ which in their vehicles stores only 2-5 seconds of data, can be read only by Toyota and customers can get that limited data only with a court order.

That’s a distinction without a difference!   Any logger on the vehicle that stores events from the automobile is an event data recorder (recall that many planes have two black boxes).   It would appear without additional information that they had read the non-volatile memory on the Engine Control Module or other ECU (Electronic Control Unit).    Is it possible to read such data from other Toyota vehicles as well?  Any afflicted parties and NHTSA ought to get a complete listing of what data get stored by each Toyota vehicle.  NHTSA should also be given the means to read that data.

As I have argued in a different posting, Toyota should release what information is stored in each vehicle before future incidents are analyzed.

Probabilities and Assumptions

March 22, 2010

A recent article in the New York Times noted that

[…] Driving one of the suspect Toyotas raises your chances of dying in a car crash over the next two years from .01907 percent (that’s 19 one-thousandths of 1 percent, when rounded off) to .01935 percent (also 19 one-thousandths of one percent)….  It’s not worth losing sleep over.

Intriguing analysis.   Some items to note that may alter the above math significantly:

  1. Why compute the probability over the next two years as opposed to 10 or even 20 years?  (According to Toyota, about 80% of their vehicles that are 10-20 years old are still on the road).
  2. Why compute only the probability of dying?  What about the probability of being injured which is about 10-20 times higher?  Do we want our cars to be unsafe and to add to the uncertainties of the world we live in?
  3. An unsafe vehicle does not just pose a danger to the occupants of the vehicle.  It could pose greater dangers to pedestrians and passengers in other vehicles.
  4. The NHTSA database only contains reports that were filed by pro-active complainants.  Many incidents may not have been reported; one estimate is that only 20% of incidents get reported – people may just sell a car with problems and move on.   It is also at least conceivable that there have been accidents, injuries and deaths that have been wrongly attributed to driver error when in fact the vehicle is to be blamed.
driving one of these suspect Toyotas raises your chances of dying in a car crash over the next two years from .01907 percent (that’s 19 one-thousandths of 1 percent, when rounded off) to .01935 percent (also 19 one-thousandths of one percent).

Some Good News for Toyota

March 19, 2010

Toyota has to deal with the fallout of its admission of electronics-related stalling in 2005-2007 Corolla and Matrix cars.   Amidst this is also a silver lining for Toyota.   NHTSA has reported very briefly that a recent incident of SUA (sudden unintended acceleration) reported by a New York driver looks to be an outcome of pedal mis-application (the gas pedal being pressed while braking was intended).     This is good news for Toyota though it does not automatically all the other reported cases of SUA.

I also interpret this report as confirmation that NHTSA does have a black-box reader from Toyota.   That’s good news for all.

Electronic Gremlins in Toyotas?

March 19, 2010

The Detroit Free Press is reporting that Toyota is considering a fix to the 2005-2007 Toyota Corolla and Matrix for engine stalling problems due to a problem in the electronics.  Apparently, there are several tens of complaints about the stalling problem.  I don’t mean to be snide, but I wonder why this is not  ‘reverse pedal mis-application’ where people meaning to press the gas pedal end up pressing the brake!   There have been several hundreds of complaints about sudden unintended acceleration and every such driver is accused of pressing the gas pedal when the brake pedal was intended.   There is certainly more to come on this story…

The Black Box Mystery Continues…

March 17, 2010

In several posts on this blog, I have harped on the topic of black boxes being key to both Toyota establishing its innocence and for diagnosing an electronics problem if it exists in its vehicles.

Most people have a strong opinion now on whether Mr. Sikes staged a hoax of a runaway Prius in Southern California last week.   Amidst the flare-up came another seemingly minor twist on the black-box front that Toyota doesn’t seem to want anybody else to understand.     For a long time, Toyota has maintained that their black-boxes (event data recorders) could store only 2-5 seconds of vehicle data for (experimental) use by them but not by others.    Furthermore, only they could read the data, and a consumer can get the data only through a court order.   And Toyota has said that they had exactly one laptop in the US that could access said data.    Recently, there was a report that NHTSA was given 3 such devices but that report has not been confirmed as far as I know.   (Data from cars made by the Detroit Three can be read directly by consumers today).

Then, the Southern California event with a Prius (supposedly) runs away, afflicted with a claim of sudden unintended acceleration.    Toyota is now confidently reporting that their analysis of the vehicle data showed that the brake and accelerator pedals were being alternatively pressed *250* times, i.e. over a duration of several minutes.  ( The implication is that Mr. Sikes was just faking the situation.   He may have well been, but it could not have been because the situation could not be repeated – SUA incidents have not been repeatable posing a major challenge for those trying to find the problem source).  So, when the data are in Toyota’s favor, they are able to read several minutes’ worth of data about the status of braking and accelerating.    Now, please watch this video clip on CNN.   No braking and acceleration data available even for a few seconds!     Toyota would very likely say “The features and recording time depend upon model year and other ‘variables'”.

Here is what Toyota should do: Release to the public a table that shows for various model years (and VIN ranges if applicable), which data and how much data are recorded.  So, the next time a SUA event happens, everybody knows beforehand what can be expected.  That sounds fair.  Right?

The Easy (and Wrong) Way Out

March 16, 2010

There are many reports over the past couple of days (please see here and here for example) implying that it’s all the driver’s fault if SUA (sudden unintended acceleration) happens.  The driver “mis-applies” the pedal, they say – pressing the accelerator pedal when intending to press the brake pedal.  The rather suspicious version of events by the driver involved in the Toyota Prius SUA incident add another layer of skepticism.

Three things to note:

  1. Readers of this blog know that we have always acknowledged that pedal misapplication certainly does happen but it does not address all SUA incidents.  Life is not simply binary: either all electronic problems or all operator error.    There is no law against there being some of each.   One of the articles highlights the fact that senior citizens are involved in more than 55% of SUA-related deaths/incidents.   Am I the only one to note that this still leaves at least 45% of incidents that cannot be just waved away by using the older age of the driver as the explanation?   A psychologist is quoted that, unlikely as it may seem, all SUA problems can really be attributed to drivers – he cites the Audi incidents from the 80s as his base.   Hello!  Does his PC or laptop work perfectly today?  It perhaps did in the 80s – simply because he had none.    Electronics is complex – just because you do not understand how it works does not mean that it is bug-free or cannot go wrong.   The blame is being pinned on the hapless driver when these writers do not want to take the time to understand what can indeed go wrong.   The drivers are sadly victimized twice: once by the vehicle, and then by these allegations.  Even NASA, whose spacecraft are operated by astronauts who have trained for their entire careers, places non-zero failure rates on their electronics when they have an obscene amount of redundancy and use the best technology/people available.   Comparably, automotive electronics only has an “affordable” level of redundancy.
  2. SUA incidents spiked in the 90s when electronic cruise control was introduced.  They spiked again when electronic throttle control was introduced in the early ’00s.   Did people suddenly become less smart when electronics got introduced?   Would/did they even know that there was electronics involved?
  3. If drivers were to be blamed, the SUA rate will be, statistically speaking, roughly uniform across all years, makes and models.  Toyota cars, after ETCS-I introduction, had more than 45% of the share of SUA incidents while having 20% of the market share.   That’s statistically significant – meaning, it is very likely that “pedal mis-application” fails to explain the discrepancy.

Does your vehicle experience SUA?

March 14, 2010

If YOU own a Toyota and experience SUA, please consider the following.   Switch to neutral, come to a complete stop by the side of the road, and then if you have the right frame of mind AND you are in a SAFE location, please do NOT switch off the engine.   Under SUA conditions, the engine is expected to be revving at high RPM (or perhaps going up and down) even if the gas pedal is not being pressed.  Take pictures of the dashboard with your cellphone, make a video clip if your phone supports camcorder operations, ask other people to verify the situation,  get a police officer to note the condition, and request their identities so that they can be witnesses to the occurrence.   A nearby friendly mechanic with an OBD-II device can be asked to read the following information:  read accelerator pedal data, the throttle position data and vehicle speed information.   Ask the mechanic to store/record the data.  Take a printout or make copious notes of the data (I would love a copy!).    Or, if you are in the Pittsburgh region, give us a call (you can find my coordinates by searching “Rajkumar CMU” in your favorite search engine) and we would like to take some readings while the engine is in that abnormal state.    Even if a dealer recommends turning off the engine to read stored Diagnostic Trouble Code (DTCs), please say no and ask them to read the accelerator/throttle position data BEFORE the engine is shut down.   DTCs are not stored under SUA conditions!  The end-result is that the dealer will very likely say “There is no problem with the vehicle!”.

The rationale is that if the engine is shut off and then restarted, the problem seems to go away (but may reappear later).  Diagnosing the situation is significantly easier when data are collected while in that abnormal state.  Again, the key is to get these data while the engine is still in the SUA state (albeit in neutral gear).

Alternatively, if you want to run some tests on your car after it has experienced SUA, please consider lending us your Toyota for a couple of weeks.    We are currently testing a Lexus IS250 in this mode.   Many Toyota vehicles do not encounter SUA, but some do.   Tests are expected to be much more likely to find a problem source by looking at the latter category of vehicles.

Please be safe under all cases.   If you know of somebody encountering SUA issues, please let them know of this as well.

More Drama Ahead!

March 13, 2010

The Prius incident in Southern California trumped Toyota’s PR move to discredit Dr. Gilbert’s test on a Toyota Avalon and his congressional testimony.  Now, it increasingly appears that the Prius incident may have been a hoax.  (I also believe that that a good number of SUA reports after the “fixes” made  by Toyota may not be legitimate and have hinted at that in earlier posts).   Now, every time a SUA incident is reported, the question would automatically come up, as it should.     However, based on many different credible accounts, there certainly appears to be a sudden and sustained unintended acceleration problem.   We now unfortunately have to peer through the fog of misleading reports.    While Toyota may be happy in one sense to see some reports being seen as hoaxes, this would keep the story alive in the media, which is not good for them.

Should we go back to mechanical systems to ensure the safety of cars?

March 10, 2010

A reporter asked me yesterday whether there are any ‘independent’ organizations or groups that test the safety of car electronics.   An independent entity would be one that does not work with any carmaker, automotive supplier or plaintiffs in a car accident.   Unfortunately, I had to answer ‘No’.   The reason for this absence of independent entities who can offer “unbiased” feedback is simple: how will they support themselves?   Automotive electronics is complex; one needs the services of experts in mechanical engineering, electrical engineering, control systems, electronics hardware, embedded real-time software, fault-tolerant systems, sensors, actuators,  EMI and ESD.  There are hundreds of models sold *every* year.   The cost of sustaining such a testing operation will be enormous, and unless one has a service contract with one of the automakers, or looking at specific issues for a plaintiff, it is very difficult to sustain the operation.  Let’s look at the landscape and how we can help the situation.

Read the rest of this entry »