Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > sci.physics > #691860

Re: Self drive cars

From Trevor Wilson <trevor@SPAMBLOCKrageaudio.com.au>
Newsgroups sci.physics
Subject Re: Self drive cars
Date 2018-06-26 08:30 +1000
Message-ID <fpd8ovFpum2U1@mid.individual.net> (permalink)
References <6fa5e49b-ff56-41b2-aaab-7f128b538e30@googlegroups.com> <fpbeqmFepbkU1@mid.individual.net> <pgrnjj$1v7c$1@gioia.aioe.org>

Show all headers | View raw


On 26/06/2018 7:42 AM, Sergio wrote:
> On 6/25/2018 1:01 AM, Sylvia Else wrote:
>> On 24/06/2018 12:26 PM, john wrote:
>>> Even if a driver is there supervising, things happen so quickly that
>>> unless a person is feeling the road- actually driving- there is no
>>> time to react.
>>> Nevergoing to work on roads- maybe in Space
>>>
>>
>> Driver supervision is only being used during development. The intent is
>> that the end product will be fully autonomous, with no driver, only
>> passengers, and the latter will not be criminally liable for accidents
>> or breaches of the road laws.
>>
>> Sylvia.
> 
> fully autonomous vehicles will always have accidents,
> 
> sensors fail, age, lose calibration

**Easily dealt with, by self-calibration systems. In the case of a 
catastrophic fault, the system would simply shut down. In the same way 
that modern fuel injection system failures cause the engine to either 
stop, or work in 'limp home' mode.

> probability of detection is never  100%,

**True, but the detection systems are already vastly superior to average 
drivers and certainly better than very young, or very old drivers. In my 
case, I have a relatively modern car, fitted with a large array of 
automated systems. In the 3rd week of ownership, I was set to reverse my 
car from a parking space. My driver's side view was completely 
obliterated by a large structure in the guise of a Toyota Land Cruiser. 
I put the car in reverse, checked the rear view mirror and the rear view 
monitor. I began to reverse, when an urgent noise was emitted from the 
dash. I stopped reversing. I then noted two small children running from 
my blind spot, across the path I was about to take. I remarked to my 
partner that the safety systems had potentially saved a bunch of people 
form having a really bad day. In fact, I am betting that the particular 
system (Subaru call it: Rear Cross Traffic Alert) will be a mandatory 
fitting on all cars within 5 years.

> probability of false detection is never  0

**True enough, but, again, the systems are already better than the 
majority of drivers.


> one cannot identify or characterize all road hazards

**Correct. Humans are still better than most automated systems. My 
Subaru employs a total of 6 cameras in various locations. Cameras and 
computers are cheap. Software and integration of various systems is 
improving significantly. Since Subaru rely totally on cameras, there are 
potential problems with heavy rain and the detection systems. That said, 
my partner suffers with poor vision (though she can still legally drive 
in any nation on the planet, with corrective lenses), the Subaru 
detection systems are superior to hers. The Subaru systems are not 
superior to my own vision. Yet. I understand the next generation of 
Subarus are due to be fitted with an upgraded Eyesightâ„¢ system, 
supplemented by radar and highly accurate GPS. The systems are improving 
very rapidly.

> plugs corrode, disconnect

**Not usually for many decades. Long after the vehicle has exceeded it 
useful life.

> the system has an overall failure rate

**So do humans. Consider a human suffering a common cold or 'flu. Or a 
human has not enjoyed adequate sleep. One could argue that a sick, sleep 
deprived, alcohol or drug affected human is vastly more dangerous that a 
computer system. In any case, feedback systems will usually detect 
faults on any system. That includes corrosion and CPU or other faults.

> 
> the courts, lawyers, owners, victims, hospitals, will all demand
> liability be identified.

**And THAT is the big sticking point right now. When autonomously driven 
cars can be shown to be better and safer than the average human driver, 
then the bean counters at insurance companies will agitate for change. 
My tip is that, within 20 years, autonomously driven cars will be 
mandatory on certain roads. It may even be within 10 years. Technology 
improvements are occurring very rapidly.

> 
> that June Bug your car hit at 60 mph took out the front right radar
> unit, which missed that pedestrian you hit.

**Fortunately, there will be multiple back-up systems. My car doesn't 
use radar. It uses stereo cameras (though I understand the next 
generation will employ both cameras and radar). In fact, a mate has a 
similar vehicle. He doesn't bother to use his brakes when driving 
anymore. Personally, I don't trust the systems that much, though I have 
found the autonomous braking works quite well.


-- 
Trevor Wilson
www.rageaudio.com.au

Back to sci.physics | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread


Thread

Self drive cars john <johnsefton288@gmail.com> - 2018-06-23 19:26 -0700
  Re: Self drive cars Alan Folmsbee <omnilobe@gmail.com> - 2018-06-23 20:44 -0700
    Re: Self drive cars benj <benj@nobody.net> - 2018-06-24 01:09 -0400
      Re: Self drive cars Alan Folmsbee <omnilobe@gmail.com> - 2018-06-24 03:49 -0700
    Re: Self drive cars Double-A <double-a3@hush.com> - 2018-06-24 13:24 -0700
  Re: Self drive cars Sylvia Else <sylvia@not.at.this.address> - 2018-06-25 16:01 +1000
    Re: Self drive cars Alan Folmsbee <omnilobe@gmail.com> - 2018-06-25 01:59 -0700
      Re: Self drive cars "reber G=emc^2" <herbertglazier0@gmail.com> - 2018-06-25 19:12 -0700
        Re: Self drive cars "hanson" <hanson@quick.net> - 2018-06-25 19:34 -0700
    Re: Self drive cars Sergio <invalid@invalid.com> - 2018-06-25 16:42 -0500
      Re: Self drive cars Double-A <double-a3@hush.com> - 2018-06-25 14:46 -0700
        Re: Self drive cars Trevor Wilson <trevor@SPAMBLOCKrageaudio.com.au> - 2018-06-26 08:31 +1000
          Re: Self drive cars Double-A <double-a3@hush.com> - 2018-06-26 15:05 -0700
            Re: Self drive cars Trevor Wilson <trevor@SPAMBLOCKrageaudio.com.au> - 2018-06-27 08:30 +1000
            Re: Self drive cars "reber G=emc^2" <herbertglazier0@gmail.com> - 2018-06-26 16:03 -0700
              Re: Self drive cars Double-A <double-a3@hush.com> - 2018-06-26 16:44 -0700
                Re: Self drive cars Sergio <invalid@invalid.com> - 2018-06-26 18:52 -0500
                Re: Self drive cars john <johnsefton288@gmail.com> - 2018-06-26 18:45 -0700
                Re: Self drive cars Double-A <double-a3@hush.com> - 2018-06-28 13:53 -0700
                Re: Self drive cars Alan Folmsbee <omnilobe@gmail.com> - 2018-06-28 16:04 -0700
                Re: Self drive cars Double-A <double-a3@hush.com> - 2018-06-29 13:22 -0700
                Re: Self drive cars "reber G=emc^2" <herbertglazier0@gmail.com> - 2018-06-29 13:44 -0700
                Re: Self drive cars Leo Sgouros <adravirgo@gmail.com> - 2018-06-29 13:50 -0700
                Re: Self drive cars "reber G=emc^2" <herbertglazier0@gmail.com> - 2018-06-26 19:57 -0700
                Re: Self drive cars "hanson" <hanson@quick.net> - 2018-06-26 22:12 -0700
                Re: Self drive cars Double-A <double-a3@hush.com> - 2018-06-28 13:57 -0700
              Re: Self drive cars "hanson" <hanson@quick.net> - 2018-06-26 22:10 -0700
      Re: Self drive cars Trevor Wilson <trevor@SPAMBLOCKrageaudio.com.au> - 2018-06-26 08:30 +1000
  Re: Self drive cars "reber G=emc^2" <herbertglazier0@gmail.com> - 2018-06-25 19:02 -0700
    Re: Self drive cars "hanson" <hanson@quick.net> - 2018-06-25 19:33 -0700

csiph-web