[Dprglist] Tesla

[Rud Merriam]

The legal landscape will be defined by the laws and regulations, 
overall. In the meantime we will see litigation defining the parameters. 
That's where it will be a challenge. But we'll also see expert witness 
making the points I just made and IMO prevailing. Tesla and other auto 
makers will have their legal teams crafting all the protections against 
litigation into the contracts.

I saw a mention of one auto company requiring hands on the wheel with 
some response needed to a wheel wiggle or something. If the car didn't 
get a response it shut down, presumably safely. I started blasting a 
loud noise inside the car, honk the horn, and flash every light on the 
vehicle with an announcement "Possible medical emergency. Unresponsive 
driver. Calling 911 for assistance."

I was revising my rovers early this summer and digging into ROS. Then I 
heard about the NASA Space Robotics Challenge that started in the fall. 
Been working on qualifying for it and that wrapped up last Saturday. 
Supposedly they will judge the simulation files submitted by the 31st. 
Twenty teams get $15,000 and can compete in the main competition. I've 
seen results from some other teams and kind of feel I'll either squeak 
in or just miss getting in.

The challenge is a simulation of the Johnson Space Center's R5 / 
Valkyrie robot that is 6' 3" and weighs 300 lbs. The main competition is 
fixing damage from a 'storm' on Mars like "The Martian" movie. Prizes 
are in the $100,000 range but also at _least_ two weeks with the actual 
robot.


-73 -
*Rud Merriam K5RUD*
/Mystic Lake Software/ <http://mysticlakesoftware.com/>
/Hackaday Contributor/ <http://hackaday.com/>


On 01/27/2017 08:01 PM, David Anderson wrote:
> Hi Rud,
>
> Not to belabor the point, but I think perhaps the point was missed.  
> You are talking about technology.   Not the legal landscape.  I find 
> it somewhat amazing that furturists such as yourself, as roboticists 
> tend to be, don't see what's coming down the road.  (pun intended)
>
> How is the Lunar Sample Retriever robot going?   Are you still 
> designing for those scenarios?  (ok, worse pun).
>
> cheers,
> dpa
>
>
> On 01/25/2017 01:29 PM, Rud Merriam wrote:
>>
>> Any driving system, even a human, is going to be incomplete. People 
>> fall asleep, which causes accidents.
>>
>> Your statements imply that the robot system has to be perfect before 
>> deployment and use. Or even nearly perfect. If that's true you'll 
>> never get automatic driving on the road.
>>
>> The other point missing is the trade offs. This report or others said 
>> that Tesla's have a much lower, like 50%, accident rate than other 
>> vehicles because of the auto driving capability. The trade off is 
>> death from this scenario versus deaths from other scenarios that the 
>> vehicle avoided.
>>
>>
>>
>> -73 -
>> *Rud Merriam K5RUD*
>> /Mystic Lake Software/ <http://mysticlakesoftware.com/>
>> /Hackaday Contributor/ <http://hackaday.com/>
>>
>>
>> On 01/25/2017 10:21 AM, David Anderson wrote:
>>>
>>> Dave Ackley sends along the following link.
>>>
>>> http://spectrum.ieee.org/cars-that-think/transportation/safety/feds-call-teslas-autopilot-safe/?utm_source=TandM&utm_medium=Newsletter&utm_campaign=TM0124201 
>>> <http://spectrum.ieee.org/cars-that-think/transportation/safety/feds-call-teslas-autopilot-safe/?utm_source=TandM&utm_medium=Newsletter&utm_campaign=TM01242017>
>>>
>>> The key line, at least for me, is:
>>>
>>> "It concluded the automatic emergency braking system (AEB) was not 
>>> to blame because it hadn’t been designed for such a scenario."
>>>
>>> So, that's how they're going to weasel out of it.
>>>
>>> Take hope!  Your robot's failure at the contest was not a failure at 
>>> all.  It was simply not "designed for such a scenario" and therefore 
>>> was actually a success!
>>>
>>> And that baby carriage the car ran over? Not a problem.   Not 
>>> designed for that scenario.   The fault is the driver's, for putting 
>>> the car in that position.  Or perhaps for not purchasing the "baby 
>>> carriage avoidance package."
>>>
>>> Anyway, all those millions of edge cases we were worried about turn 
>>> out not to be a problem!   Or at least not Tesla's problem...
>>>
>>> onward,
>>> dpa
>>>
>>>
>>>
>>>
>>> _______________________________________________
>>> DPRGlist mailing list
>>> DPRGlist at lists.dprg.org
>>> http://lists.dprg.org/listinfo.cgi/dprglist-dprg.org
>>
>>
>>
>> _______________________________________________
>> DPRGlist mailing list
>> DPRGlist at lists.dprg.org
>> http://lists.dprg.org/listinfo.cgi/dprglist-dprg.org
>
>
>
> _______________________________________________
> DPRGlist mailing list
> DPRGlist at lists.dprg.org
> http://lists.dprg.org/listinfo.cgi/dprglist-dprg.org

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.dprg.org/pipermail/dprglist-dprg.org/attachments/20170128/97be0a0c/attachment.htm>