On 2016-04-05 10:03 Jim Birch wrote:

> Driverless cars have a better record than human drivers.  Their most common 
> accident is being rear-ended by human drivers who are running red lights and 
> expect the driverless car in front of them to do the same.
> 
> A week in a spinal ward might bring home the benefits of driverless cars.  
> It's a no brainer when considered in terms of relative risks, rather than 
> risk elimination.

That's a fairly ideal scenario.

But of course the software will have bugs, as have some engine management 
systems.  I remember one make which was reported to apply heavy throttle 
unpredictably, on one occasion apparently propelling a woman backing out of her 
driveway into traffic at high speed and killing her.  Situations will also 
arise where the software is simply inadequate, perhaps fatally so.  And 
hardware will inevitably fail.

There's a sign on a mountain near Kangaroo Valley warning that the area has 
very high EM radiation levels (because of the antenna farm) and if your car 
doesn't start just roll it down the mountain a bit and everything will be fine!!

Call me over-cautious, but it will be a while before I entrust my nearest & 
dearest to a driverless car.

I think current models allow driver override, and in fact demand occasional 
driver input (like the railway dead-mans handle).  This means there are four 
agents potentially involved in any accident - two computers and two drivers.  
In the event of an accident questions will arise as to whether one or other 
driver should have intervened, and if not, what human agency is liable (you 
can't sue a computer).  I think it would present an interesting legal problem.  
Can you see an Australian insurance company suing Volkswagen or Ford Motor 
Company?


> They are also likely to improve traffic flow and ease congestion.  Especially 
> when human drivers are completely gone.

Experience since the 60's has demonstrated that congestion is never improved 
for long - more cars use the road and pretty soon everything is back to normal.

David L.
_______________________________________________
Link mailing list
[email protected]
http://mailman.anu.edu.au/mailman/listinfo/link

Reply via email to