[colug-432] Great moments in computer science

[Tom Hanlon]

Some thoughts..


>> Most of the cars on the road are not entirely drive by wire. Your point
>> about drive by wire is well taken. I was surprised the people driving
>> those corollas never thought of either:
>>
>> Turn off the ignition
>> Put it in neutral - let the engine blow up.
>> Engage the hydraulically actuated clutch
>> Throw on the parking brake

I am not surprised..

It can be hard to think rationally and creatively when the one thing
you expect to work and has worked for all of your driving life, fails
to work.

So you take your foot of the gas.. and the car accelerates.

The brake pedal would be a natural choice, almost muscle memory and I
am sure that is what most drivers probably do.

The other options you propose.. I doubt that many folks would come to
them in a panic.

I am not  "surprised" that folks do not turn off the ignition, or put
the car in neutral or pop the clutch. I think when you try what has
always worked and is more or less muscle memory, then your lower brain
kicks in, not your higher brain.

By that point they are frightened, and in a speeding car, rational
thought about how to stop this speeding car in a non traditional way
are not the usual pathways for human thought while panicked.

When was the last time you turned off the ignition to stop a moving car?
When was the last time you put a car in Neutral to stop a moving car?
The clutch is used to slow a car down at times, so I give you that as
a potential response.
When was the last time you stopped a moving car with the parking brake?

So those solutions although they would work, are not something you
have the luxury to figure out. None of the above are reflexive
reactions to the problem of a speeding car.

Just like a table saw accident, or grabbing a breaking pane of glass,
reflex is the response rather than logic.

It is easy to think outside the box while sitting at a computer, much
harder to think creatively when in a speeding car and your mind is
more into thinking reactively rather than creatively.

As for an earlier comment about insurance rates for self driving cars
most likely being higher. I would credit that mostly to the small
sample size to judge the proper insurance rate. I assume insurers when
faced with an unknown bet high, once they have enough data to make a
safer bet for them, then the rates go down to the appropriate level.

In general I assume the rise of cheap sensors and cheap CPU's has
taken all of the wonderful mechanical engineering out of cars. I do
not know much about clutches and powertrains, but I have to think that
an analog controlled system was somewhat elegant.