[vox] No electronic flaws in Toyotas?

[Shwaine]

On Tue, 15 Feb 2011, Bill Broadley wrote:

> On 02/09/2011 10:11 AM, Shwaine wrote:
>> The one case that gives me pause on calling this purely and completely
>> driver error was the case in Southern California where the driver was a
>> trained officer (I forget if he was CHP or local law enforcement). Now,
>> there was extenuating circumstances there in that the officer did not know
>> that particular car since it was a loaner, so he may not have known that
>> the keyless ignition button needed to be held for 3 seconds to cut the
>> engine. But I think we can be pretty sure he was not stepping on the wrong
>> pedal (particularly since witnesses saw the brakes flame out) and he would
>> have known how to unstick the gas pedal from a floor mat (the attributed
>> cause of the accident) given his law enforcement training.
>
> URL?  I saw some passing references to it, but nothing substantial.
> While CHP or law enforcement do receive some training they are hardly
> driving experts.  In face when I was taking the motorcycle safety course
> they taught us the concept of counter-steering.  There was a simple test
> where you had to approach at 15 mph and then to simulate an emergency
> you can to turn hard right, left hard left, or emergency stop as
> signalled by the instructor.
>

Do a Google search for "officer dies toyota" or "office dies lexus" (it 
was actually a Lexus loaner). The story is in the first page results for 
both queries. And unless you can conduct a seance to contact the officer 
in the great beyond, you can't say from your anecdotal evidence after your 
interaction with one officer in one course that the deceased officer was 
at fault. It is presumptious to do so, and it is very bad to go into a 
scientific or engineering inquiry with presumptions. If you presume it was 
ALWAYS user error, you may overlook some rare logical or transient errors 
in the code or circuitry or a design decision that leads to confused 
drivers.

And actually, according to people who know the model of Lexus involved, it 
does have a confusing design decision. Its system won't let you put it in 
neutral while the engine is engaged without doing a different shifting 
mechanism to let the drive-by-wire system know that you intend to put it 
into neutral and it wasn't an accidental shift. So you could move the gear 
shift into neutral and the system would ignore that directive (and move 
the gear shift back to drive) because you didn't hold the shifter in 
neutral for a couple of seconds to tell it to override the default "ignore 
accidental shift into neutral" reaction by the drive-by-wire system.

Combine that with the fact that it was a loaner (as I already pointed out) 
and drive-by-wire systems were still relatively new at that time, the 
officer very well could have TRIED to put it in neutral and just didn't 
know about the override movement you had to do in that system. Just like 
(as I already pointed out) he probably didn't know you had to hold the 
start button for more than three seconds to kill the engine (at the time, 
people didn't know about this because it wasn't well-publicized; accidents 
like this were what made it well-publicized). There are entire YouTube 
tutorials out there now showing people how to put their drive-by-wire 
systems into neutral or how to cut the engine because it is not the same 
technique as one would use in a mechanically gear shifted car.

So to summarize, it's presumptious to assume this was all "user error" and 
that there were no design issues involved. Not allowing a driver to shift 
into neutral while accerelating without performing an override procedure 
that is not intuitive to a person used to a traditional mechanical gear 
shift is a design issue. Not cutting the engine when the driver hits the 
brakes is a design issue. The design issues may make sense when the 
vehicle is operating under normal conditions, to prevent damage to the 
engine or to deal with people who drive poorly (foot on both brake and 
accelerator) for example, but they severely limit people in emergency 
situations... where all the old driving advice that you listed would be 
hard to actually pull off because the drive-by-wire system would ignore 
half the directives unless you knew the specific override procedures for 
that particular model.

If there are going to be override procedures for shifting into neutral and 
killing the engine, they need to be standardized across all drive-by-wire 
cars, not differing by manufacturer and model, so people can learn the 
"new way" to do these things using one set of procedures, which can then 
be ingrained as emergency responses. Having each company come up with 
their own set of override procedures is going to lead to more accidents 
like this... which I am sure you would chalk up to "user error" if the 
driver did not know the procedures for the model they are driving, but 
which I would chalk up to not having a standardized method of performing 
emergency overrides on drive-by-wire systems. And I would also attribute 
accidents such as this to not teaching drivers how to perform emergency 
overrrides with drive-by-wire systems. Even such simple advice as "try 
holding the start-button/gear-shift for a few seconds and see if that 
kicks in the override" would help avoid serious accidents in the future.