DPRG List  

[DPRG] [Fwd: RE: Dallas News inquiry]

Subject: [DPRG] [Fwd: RE: Dallas News inquiry]
From: David Peterson robodave1 at comcast.net
Date: Thu Jul 7 22:47:06 CDT 2005

Hello Jim,

I believe that Japans push towards humanoid robotics is a great natural 
progression to what people really would like to see as helpers. Perhaps 
in their own culture there have been items of interest (cartoon super 
robot heros) that have also contributed, where in our own culture robots 
have been a bit more threatening. Also simple, single purpose machines 
may seem like an appropriate route to go for the tasks of elder care 
because we can more easily see a single task being achieved. Perhaps the 
quality of care is what is being pursued by the Japanese. When was the 
last time you had a conversation with your microwave? Have you developed 
a fondness for your VCR? What I think would be better to realize is that 
people generally prefer other people to care for them, not some 
featureless box that hums around. Elder care is more likely to be 
quality care if it appears to be given from something that even just 
appears to be interactive and even more so if it appears "like us". 
There may be some difficulties in achieving all the tasks that we 
percieve are needed of humanoid robotics, but even limited interactivity 
can achieve results in giving people new life.
    I believe our own assumptions of what humanoid robots should do is 
coloring what value they can have, even at a less than all-inclusive 
functionality. I would imagine that a person that is staying in a 
nursing home would be willing to tell you the value of "someone" simply 
coming by to say hello (speech synthesis, already achieved) , to note 
who they are (face/object recognition, already achieved) and to be able 
to summon help if requested or noted that it is needed (temp/heart rate 
monitoring). Is it possible to simulate a conversation, at least on the 
level many such interactions might be held? Yes, even with a program as 
simple as Eliza ro Alice. Is it possible to do more, like carrying a 
person? Probably (see Kawada Industries HRP-2), but there comes some of 
the difficulties that others have noted. It could take some study, but I 
believe that anthropomorphic care-givers could initally provide some of 
the more simple tasks. Then as technology improves more complex and 
difficult tasks can be achieved, such as lifting people to wheelchairs 
or beds, cleaning tasks, medicinal adminstrations, meal preparation.
   As far as power concerns are measured, I do not see a problem with 
"tethered" operation until such point as better mobile power technology 
is achieved. Currenlty Asimo achieves approximately 30 minutes operation 
on 10.5 Ah of 38.4 volt NiMh batteries, weighing about 17 lbs (7.7 kg). 
This could nearly double witha similar weight of LiPoly cells, but that 
still isn't a long runtime. However, we don't expect our floor buffers, 
microwaves or other "appliances" to operate off of batteries, so this 
objection could be significantly cured by simply plugging in the robot 
where ever it is needed.
   Safety issues are something that can be brought up with any form of 
technology. Those LiPoly cells I suggested can be somewhat dangerous if 
improperly used, but how many thousands of cell phones already use them? 
Laptops? Power levels are generally monitored and even in these devices 
a graceful shutdown is made once power levels are read to be a bit low. 
Issues such as proper performance of designated tasks involve testing 
through the FDA. I can only imagine the tests involved in that 
interesting balancing wheelchair, the Ibot 3000 before it was finally 
approved. Many "what-ifs" could be proposed, but companies with the 
ability to tackle them (Honda, Sony) are well aware of them, as they 
have net with liabilities in other markets. (automobiles, computing).
    Interactivity has to be the key value of a care-giver robot. 
Personality can provide that, even in non-anthorpomorphic robots, such 
as R2-D2..  I recall a study in which Aibo(s) had been given to some 
elderly people for a period, and the results were decreased heart rate, 
and other factors that indicated a new sense of well being had been 
achieved even with this simple non-humanoid form. Even researchers here 
in the US have worked with "Gizmo" for an interactive pyschological 
study, and apparently to develop a methodology for robot learning. We, 
as people, do thoroughly enjoy interactions with animals, but ultimately 
people will look towards other "robot people" as perhaps things that can 
be trusted to be helpers.
   The value delivered by a care-giver robot must then be measured by 
more than it's ability to bench-press grandma. It has to include the 
ability of that robot to provide "entertainment" and a "quality of life" 
for the people it is intended for. Our world has been developed by us to 
fit human size and shaped people, so the platform most likely to achieve 
the greatest variety of tasks will ultimately be humanoid. And the first 
task of "Hello World" has already been achieved.


(Dale, are you really more frightened of a keypad than you are of a 

----- Original Message ----- 
From: "Dale Wheat" <dale at dalewheat.com>
To: <dprglist at dprg.org>
Sent: Wednesday, July 06, 2005 8:46 PM
Subject: [DPRG] [Fwd: RE: Dallas News inquiry]

> His response to my response, my response, then his initial inquiry, in 
> that order.  Feel free to contribute to the discussion.
> Thanks,
> Dale Wheat
> http://dalewheat.com
> (877) DALE WHEAT
> (972) 486-1317
> -------- Original Message --------
> Subject: RE: Dallas News inquiry
> Date: Tue, 5 Jul 2005 17:18:22 -0500
> From: Landers, Jim <JLanders at dallasnews.com>
> To: 'dale at dalewheat.com' <dale at dalewheat.com>
> Dale,
> Thanks for your good insight. I look forward to the responses from 
> other
> Dallas robotics fans.
> Jim
> -----Original Message-----
> From: Dale Wheat [mailto:dale at dalewheat.com]
> Sent: Tuesday, July 05, 2005 6:14 PM
> To: Landers, Jim
> Subject: Re: Dallas News inquiry
> Jim,
> Thanks for writing.  I can't say that I'm an expert on either (1) the
> Japanese
> social mindset or (2) Japanese state-of-the-art robotics, but I can 
> give you
> a 3
> minute course in what you might need to know to ask the right people 
> the
> right
> questions.
> I will also pass this letter along to our member mailing list, which 
> is
> always
> guaranteed to get interesting if not always useful responses.
> The upshot, however, in my most humble opinion, is that the entire 
> concept
> of
> the android/humanoid robot as household servant, companion, et c., is
> fraught
> with difficulties.  Just as we don't build really big machines with 
> really
> big
> legs to transport people or material (we use trains, planes and other
> machines
> that bear no resemblence to us or other naturally occuring beasties), 
> it
> makes
> more sense (to me) to build specially-designed equipment for specific
> purposes.
>   Does your home alarm system look like a Rottweiler?  No, it looks 
> like a
> thermostat with a keypad.  Why?  Rottweilers are complicated - 
> thermostats
> are
> not.  The same holds true for the future keepers of Japan's (and the
> world's)
> aged.  The same machine that assists you in the restroom need not be 
> the
> same
> machine that handles your postal mail or prepares your meals.  The 
> list of
> analogies goes on.
> The Japanese perception that humanoid machines will be more easily 
> accepted
> by
> society is terribly short sighted.  What would Ford-truck-driving
> great-grandpa
> think of the Escalade with voice-activated GPS mapping?  Although it 
> should
> be
> possible to model or mimic humanoid motion (ala Asimo), it's still 
> many
> years
> away from practical application.  Asimo can walk around for about 20 
> minutes
> before needing a recharge.  In this 20 minutes he/it can walk around 
> quite
> impressively, identify faces and even climb stairs, but cannot carry 
> any
> significant payload or react to any non-programmed events in the
> environment.
> This level of acheivement has taken many years and much work.  I was
> fortunate
> enough to see Asimo at SMU and came away with a great respect for the 
> work
> being
> done.  It's still a long way from being a practical, producible 
> commodity,
> as it
> needs to be to fill the described need.
> The topic becomes more complex when you look at how far apart the 
> different
> schools of thought are on solving this exact problem.  Compare Marvin
> Minsky's
> work in artificial intelligence with Rodney Brooks.  Compare the 
> various
> team
> strategies used in the recent DARPA Grand Challenge to the Roomba.
> Have fun in Japan!
> Thanks,
> Dale Wheat
> http://dalewheat.com
> (877) DALE WHEAT
> (972) 486-1317
> Landers, Jim wrote:
>> Hello Dale Wheat,
>> This is Jim Landers, a Washington correspondent for The Dallas 
>> Morning
> News.
>> I'm going to Japan next week to look at ASIMO and QRIO and to talk 
>> with
> the
>> Japanese about whether robots like these will be providing home 
>> health
> care
>> for a rapidly aging population. I'm an international affairs 
>> reporter, not
> a
>> science writer, and what interests me as much as anything else in 
>> this
> story
>> is how the Japanese have maneuvered their society into a predicament 
>> where
>> robots seem like a solution (low birth rate, negative views on
> immigration,
>> etc.) But I gather the Japanese focus on humanoid robots is widely
> discussed
>> in robotics circles around the world. ASIMO's visit to SMU is one 
>> local
>> point of departure, and there are probably others. I'm hoping your or
>> someone else in the Dallas society can offer views on how close the
> Japanese
>> are to a product that can play this home health care function, and 
>> whether
>> it can be done at a reasonable cost. Would you be willing to discuss 
>> this
>> over the phone?
>> Thanks
>> Jim Landers
>> Washington Correspondent
>> The Dallas Morning News
>> (202) 661-8407

More information about the DPRG mailing list