[SystemSafety] How Many Miles of Driving Would It Take to Demonstrate Autonomous Vehicle Reliability?

[Chris Hills]

-----Original Message-----
From: systemsafety
[mailto:systemsafety-bounces at lists.techfak.uni-bielefeld.de] On Behalf Of
Derek M Jones
Sent: 18 April 2016 16:35
To: systemsafety at lists.techfak.uni-bielefeld.de
Subject: Re: [SystemSafety] How Many Miles of Driving Would It Take to
Demonstrate Autonomous Vehicle Reliability?
Pater,
>> How about a prescribed solution to the "trolley problem" ethical 
>> dilemma
>> (http://www.philosopherstoolkit.com/the-trolley-problem.php) and 
>> strong evidence that it has been implemented correctly.
>
> There are known quasi-acceptable (but not perfect) solutions, and some 
> known unacceptable solutions, that have been published over the last fifty
years.
This philosophical stuff is all very interesting.  But...
What I want to know (and probably most customers) is can, I choose a car
configuration option that considers my life to be more important than
everybody else's? 
 
[CAH] would that therefor be premeditated homicide if your car then killed
someone to save you? 
[CAH] Just asking but when ever asked the vast majority (from memory I have
no reference to hand) of people, like Derek (and Me :-) want a car that
saves them as a primary objective. 
 
Or are the government going to mandate that the software treat the life of
everybody equally?
[CAH] The government has to say everyone is equal?
It is just that Government Ministers are more equal (obviously) :- )  
There would then be a black market  for hacked SW to save the driver.....

As a human driver I have the option, given enough thinking time and fast
responses, of deciding whose life is worth more.
[CAH] no your honour I did not decide to hit A instead of B  it was just an
instinctive reaction.......
I recommend you watch this which asks similar questions
http://www.boreme.com/posting.php?id=44255#.VxUG3__ruUl