Re: What became of the personal robot, and comments on design
Eric Hunting
http://www.youtube.com/watch?v=3qdXkJVQUfM
Though automata as a form of art, puppetry, parlor entertainment, or
stage magic may actually go back as far as Roman times, the general
concept of the robot emerged in western culture at a time when middle
and upper class status was still largely defined by the ability to
afford domestic servants, when machine based industry was still
largely machine-assisted-human production, Victorian notions of class
and eugenics were still strong in the popular culture, society was
first seeing the emergence of a global organized workers rights
movement, and automation as we know it today was nascent with little
common knowledge in society of how such technology worked. If one were
to imagine a machine doing any particular task the only analogs for
how a machine would perform it were derived from the example of the
human being doing that task. It was thus natural that the culture
would arrive at this notion of the ultimate machine being
anthropomorphic. If a machine was going to function as a human worker
or servant, it seemed perfectly logical that its physical form would
have to be very similar in order to perform the same tasks. In fact,
the robots that gave us the term -in the 1921 play Rossum's Universal
Robots- weren't even mechanical. These were biological machines.
Factory-manufactured humans. Frankenstein creatures -if Frankenstein
had gotten it 'right' and then gone into mass production. What in
later terms we would call 'replicants' or 'clones'. Here the author,
Karel Cepek, was drawing a very logical futurist conclusion. Science
of the time was constantly reinforcing the notion of the human body as
a machine -the ultimate machine. Thus the ultimate end-conclusion of
machine age technology would logically have to be something synonymous
with the human body. Would have to be man reinventing man.
The problem, though, is that this idea is completely wrong. The human
body is not the ultimate machine. From an evolutionary standpoint,
it's just a pretty-good generalist. As automation matured we have
found that the most efficient ways to perform most tasks were usually
very different from how a human worker would perform them, since with
automation we can optimize the whole parameters of tasks. The human
body evolved to cope with an environment largely out of human control.
A physiological generalist coupled to a big brain with the critical
power of communication makes for a powerful survival combination. But
it can never do any particular task particularly 'well'. Just good
enough to avoid an evolutionary cul de sac. This is why automation is
always so easily replacing man in the industrial realm. And this is
partly why human beings are forever dissatisfied with their own bodies
and have, since civilization began, been using whatever technology we
had at the time in an attempt to functionally or cosmetically modify
them, using everything from clothing and tattoos to body-building and
plastic surgery. Thus, the android is a While Elephant by definition.
An extremely difficult and sophisticated engineering exercise whose
end result -even where it achieves parity with the human being- is
never going to be all that useful. The anthropomorphic machine has the
same problems as the human being. Any task a relatively simple machine
can do better than a human being is a task it can do better and more
cost-efficiently than an android. Thus there is no practical purpose
to the android. Its form has no particular function. It's aesthetic.
It is made in our image for no other purpose but to look like us.
And I think this is the root of the problem with the personal robot.
From the start the concept was defined anthropomorphically, creating
the persistent expectation of this most-difficult-to-realize form of
machine. But ultimately it's pointless, increasingly anachronistic to
the actual trends in automation. As we actually implemented practical
domestic automation, it took the form of appliances able to do things
more efficiently than us because they weren't mimicking us. And it
took the form of mass production of convenience products -like frozen
TV dinners- that eliminated domestic work altogether within automation
'behind the wall'. The way total automation is depicted in Star Trek.
It's an implied age of Total Automation, but you never see the
machines. They're all somewhere behind the wall, except when they are
perfectly anthropomorphic androids with that mimicry being the point.
A gimmick to deal with limited production budgets, but also a comment
on the real trends in automation across the 20th century. By the end
of 20th century, with the emergence of the personal computer, this had
pretty-much eliminated or radically simplified all the household jobs
the middle and upper classes used to hire domestic servants to do. So
what was left for the personal robot?
This is the question that confronted consumer robot designers at the
end of the century. We had long cultivated in the culture this
anticipation for the personal robot, had always envisioned it in
anthropomorphic forms, the technology to make it finally seemed to be
close at hand, but we'd obsolesced all its practical domestic uses
with other machines. What was it going to do? Starting in the early
1980s, a hopeful 'consumer automation' industry emerged in parallel
with the personal computer revolution, anticipating a similar
revolution for the personal robot. Though the ideal was still being
defined by visions of androids, late 20th century SciFi had given this
new industry a gift in the form of non-anthropomorphic yet charming
robots that were potentially more easily realized with available
technology. Dozens of 'home robots' hit the market and virtually all
of them failed. Which brings us to Newton, the epitome of the 1980s
concept of a personal or home robot attempted in many similar forms
over that decade. Newton is utterly useless. It has no manipulative
capability, no tools or other functional elements at all as that was
either too difficult to implement at the time or considered too
complicated for consumers, and the crudest of pre-internet digital
connectivity. So it just rolls around, talks, and blinks. It's clearly
a derivative of office/hospital messenger robots that also appeared at
the time but since there weren't a lot of places for a robot of the
sort to go in a suburban home and no way for it to travel outside of
it, they left that capability out. It's a very elaborate, expensive,
and hopelessly limited voice-driven personal computer and home
automation controller for people too dumb to handle a normal PC. Its
design seems to have been justified on the presumed desirability of
its technology alone, independent of any purpose. It's a fetish
object. A futuristic tiki god. This sums up what most attempts at home
and personal robots have produced to date. You know just by looking at
it that, with its high-tech combination of utter uselessness, cheesy
overly conspicuous SciFi styling, and potential for persistent
annoyance, that anyone dumb enough to actually buy one of these things
hacked it or put it in a closet and forgot about it within months. And
you could say that about most of the home robots of the period.
Designers of the time seemed to struggle in vain to find any sort of
practical role for the things when racing against advances in
appliances, consumer electronics, and personal computers -where all
the real home and personal productivity was. By the mid 90s, most
companies that entered the field of consumer robotics were gone and
the concept of the home or personal robot seemed dead -replaced by the
much more rewarding hobby of making robots yourself. And the leading
activity for these machines? Robotic cock-fighting?
And yet there was one personal robot produced in the 80s that, despite
barely qualifying as a robot by contemporary standards and despite a
legacy of some of the most incompetent corporate management in
history, was so successful it sold as many units as the best-selling
personal computers, was revived after corporate failure four times
(most recently in 2005), and is still produced today. And that robot
is Teddy Ruxpin, the famous robotic teddy bear. Teddy Ruxpin opened
the market for robotic toys (as opposed to toy robots that aren't
actually robots) and it succeeded as a personal robot because it fit
and complimented a niche we have, for centuries, filled with simpler
artificial companions animated by children's imaginations. As a
personal robot this succeeds exactly where the much more elaborate
Newton and its counterparts failed, in the only practical role a
personal robot really has; socialization and companionship. It's not
particularly useful but it doesn't try to look like some stereotype of
a robot or try to perfectly mimic humans, and it's personable and
that's what it's all about. Ironically, the developers of Newton did
actually address the role of companionship but they never realized the
actual significance of that role and in that context the design was
still a complete failure because it tried too hard to _pretend_ to be
a robot as we've always expected and as envisioned by contemporary
Hollywood SciFi instead of being something that was genuinely
personable.
An interesting irony of the alter-future depicted in Star Wars was
that, in a universe full of robots, the most useless of them was the
one that looked the most like a human being; C3P0 -a virtual comedic
parody of the robot in Fritz Lang's Metropolis. It's as if George
Lucas was pointing out the functional anachronism of the android we've
so long idealized and saying to us; "y'know, there's only ever going
to be one point to this thing, and it's social." That's really the
only point to having any sort of anthropomorphic machine in our
presence. That form has no other practical purpose but to please our
eye and integrate with our social interactions. Beyond that, it's all
about what's going on in software. And when you get that you realize
that, like so many other things, you can often pull this off with much
simpler hardware because the software matters more to the end effect -
as Teddy Ruxpin demonstrates. Asian companies today are obsessed with
realistic android concierge robots, which always turn out silly or
creepy. You could do a better job with a self mobile video screen
projecting a much more pleasing and capable virtual character. The
Danish government just recently purchased a thousand of those Paro
robot baby seals made in Japan to distribute to nursing homes across
the country for therapeutic purposes. These things are seemingly
useless. They do absolutely nothing but exude personality and an
illusion of attention while being cuddly. The same thing as pets,
without the mess. But that's the point. This purchase comes in the
wake of some recent psychiatric studies comparing the use of robots to
live animals in nursing home pet therapy. They discovered there was
little difference in effect between them. The elderly generally seem
to perceive little difference between living and mechanical animals,
even though they may know, intellectually, the difference. As artists
know well, we are very willing to suspend disbelief when our emotional
cues are properly addressed and there's no age limit to that.
This, then, seems to be the real role of the personal robot, the one
the consumer robotics designers of the 80s seem to have overlooked.
They're social machines. There's an old purportedly yiddish saying
that god created man because he liked to listen to stories. I've long
argued that, should we ever realize sentient AI, it will likely be for
similar reasons. And I see nothing wrong with that because culture is
the highest order of artifice. We otherwise have no need for software
with personality. It gets in the way. This is why user interface
concepts like Microsoft's Bob are such failures. The practical role of
the personal robot seems to parallel this, though it's in competition
with the personal computer and its ability to create virtual
characters which are visually more sophisticated but can't be touched
or interact with the physical environment. What's C3P0's one shining
moment in the whole Star Wars series? When he's telling the Ewoks
stories like an old-fashioned village storyteller. This is his true
purpose. You could sense the smile he could never actually make.
Recently the hobby robot crafters of Kyoto University's Robo-Garage
(perhaps the most sophisticated robot builders today in terms of
design) created a Murisaki-bot; a surprisingly elegant table-top robot
whose only function is to recite the ancient Murasaki Shikibu novel
The Tale of Genji. This is a book embodied as a robot. (http://dvice.com/archives/2008/07/murasakibot_rec.php?p=2&cat=undefined#more
) The android will never have a truly functional purpose even if the
technology does achieve a point of functional parity with the human
body. The point of anthropomorphism in machines is the
anthropomorphism itself. And it doesn't necessarily need to be
'realistic' -in fact, that can often be a detriment if its not total.
It just needs to be aesthetically pleasing, entertaining, and
personable. Once again, the sex industry may be more in tune with the
evolution in technology than the people developing that technology.
There you have companies already anticipating and working toward sex
robots. You can tell when a technology has found a permanent place in
our culture when we start to sexualize it.
I think we will see many specialized personal robots in the future,
though we will probably rarely call them personal robots because, for
the most part, we'll perceive them as tools or toys. The anachronistic
notions of personal robots are going to persist for some time. And yet
while we're dreaming of that cybernetic version of Jeeves (as in the
Jeeves and Wooster novels -which probably could do with a SciFi
remake...) our environment will fill up with all sorts of robots, both
functional and simply entertaining or comforting. I think the robotics
industry overlooks a lot of opportunities because its vision is
clouded by these persistent cultural anachronisms. There are a lot of
really useful applications being missed because they don't fit our
common expectations.
I think about this a lot myself because, having long lived with less-
than-ideal health, and particularly as I get older, I've often thought
about tool-oriented robotics as a supplement to personal ability. For
instance, one of the many concepts for acquiring low-toxic housing
I've studied is the idea of robotic home construction based on the
simple premise that sweat equity is difficult for someone of
diminished physical ability. It occurred to me that if I couldn't
build a house readily by myself, or afford the 'market's' labor, maybe
building the robots to do it for me was more attainable -especially if
I could get the hobbyist robotics community to help. (of course, that
was the stumbling block...) I've since imagined a variety of tool
robots that are quite simple, perfectly feasible, and yet remain
unexplored by anyone. For instance, I've long been considering the
potential of a robot in the form of a small ATV akin to the old
military Mechanical Mule with a drive system like the Hallucigenia (http://www.lleedd.com/hallucigenia/index.html
) that does no more than follow me around, go where I tell it to, and
just carry stuff. Basically, a large self-propelled garden cart. I can
put that to ten thousand uses, particularly outdoors. Typical
wheelbarrow and hand cart applications in rough terrain, self-portable
tool boxes and work benches with their own power supply, adjustable
height work platforms, portable stairs, automated lawn tractor tasks,
and the same sort of 'portage' tasks we often still use pack animals
and people for. Imagine going hiking with a Japanese Capsule Hotel
cabin following you as you go. And it doesn't need to talk to me or
have some kind of anthropomorphic 'face'. It just needs to talk to
something like an iPod Touch and use some status lights, maybe an icon
display, and make a few pleasant noises to let me know when I have
it's attention. We've already done this with robotic golf caddies, but
their design is ridiculously over-specialized. Then there's the
InchWorm -a simple multi-jointed arm with modular end-effectors on
either end that alternately plug into power bases, tool pallets,
individual tools, or a socket traverse grid built into a larger
structure and can work individually or in groups. I originally
imagined this as a standard form of modular robot for orbital
telerobotic construction but realized it had many potential
terrestrial uses as well. Again, a very simple form that could do ten
thousand jobs. Plug that into the flat bed of that robot ATV and
imagine what you can do. This hints at the possibility of a modular
robotic tool kit based on the same iPod-Touch-like user interface
where you have these two basic units you just keep elaborating with
plug-ins. A robot like the InchWorm actually already exists, but for
some strange reason its developers haven't quite figured out what to
do with it. I think that sums up the whole robot development problem
right there. They remain solutions in search of problems because we're
not really making them for practical reasons. We're making them for
emotional reasons we never admit to. They're not really engineering.
They're art. Why else put a teddy bear head on a machine intended to
drag wounded soldiers off the battlefield? (http://news.bbc.co.uk/2/hi/health/6729745.stm
)
Eric Hunting
erich...@gmail.com
> Date: Thurs, Dec 18 2008 6:21 am
> From: "Paul D. Fernhout"
>
> From:
> "BBC NEWS | UK | Magazine | What became of the personal robot?"
> http://news.bbc.co.uk/1/hi/magazine/7784509.stm
> "What happened to the personal robot - a domestic servant who would
> never
> tire of being told what to do? When writer Danny Wallace set off to
> find out
> the fate of this long overdue dream, the BBC's Peter Leonard joined
> him.
> Human beings are just brilliant. We can dive to the sea bed in
> submarines,
> fly to the moon in rocket ships, design uber-powerful computers and
> so on
> and so forth. But so far we've failed to deliver one of the most
> tantalising
> promises of the modern age. Personal domestic robots have remained
> irritatingly elusive. Here are some of the reasons that making
> humanoid
> robots isn't as easy as we were led to believe by novelists,
> screenwriters
> and scientists. "
Bryan Bishop
> http://www.youtube.com/watch?v=3qdXkJVQUfM
Thanks for that. :-) Vern Graner and the Austin Robot Group were
playing that on a projector a few months ago when I started attending
their meetings. Part of the "VIDEOS OF INTEREST" segment. Good times.
> And I think this is the root of the problem with the personal robot.
> From the start the concept was defined anthropomorphically, creating
> the persistent expectation of this most-difficult-to-realize form of
> machine. But ultimately it's pointless, increasingly anachronistic to
> the actual trends in automation. As we actually implemented practical
> domestic automation, it took the form of appliances able to do things
> more efficiently than us because they weren't mimicking us. And it
> took the form of mass production of convenience products -like frozen
> TV dinners- that eliminated domestic work altogether within automation
> 'behind the wall'. The way total automation is depicted in Star Trek.
> It's an implied age of Total Automation, but you never see the
> machines. They're all somewhere behind the wall, except when they are
> perfectly anthropomorphic androids with that mimicry being the point.
> A gimmick to deal with limited production budgets, but also a comment
> on the real trends in automation across the 20th century. By the end
> of 20th century, with the emergence of the personal computer, this had
> pretty-much eliminated or radically simplified all the household jobs
> the middle and upper classes used to hire domestic servants to do. So
> what was left for the personal robot?
Another part of the story, not quite related to the personal robot, is
the home automation scene. I've not really checked too much into it,
but there was a similar slope of interest and commercial products from
the 80s and 90s that I can recall. From an engineering point of view
this is just 'facilities engineering', and we've been making "home"
automation in buildings since the beginning of the industrial
revolution. However, there was a consumer-oriented home automation
trend that started showing up just after personal computers went big,
X10 and the like. These were the types of automation that should have
been called "control systems" like the control of lights, garage
doors, potentiometers, HVAC, etc. Now that we have so many easy
microcontrollers to program with and fun standards like IEEE 802.3
(ethernet), USB, firewire, LTP (parallel ports), RS232 (com ports),
and 802.11g, there's a bit of a comeback because it's now easier to
actually experiment with some of the devices. But the first commercial
stride into this basically failed-- except sprinkler systems
apparently.
I still feel that home automation isn't where it should be. Lawn
mowers are still mostly sold as oil guzzling machines without
programmable controllers and guidance [not just because of the safety
concerns], dish washing is a joke, and laundry, don't get me started
on laundry. At my dorm, we have a large cafeteria, and in the back
there's a dish drop-off point on a conveyor belt, and you place your
dishes, sort your silverware and disposable paper-things, and off goes
the dishes. But if you peak into the window you see that there's a
person in the back, scrubbing all of the dishes. Obviously somebody
completely missed the point when designing /that/ one. Since all of
the plates are of a standardized size, it wouldn't be too terribly
hard to build some metal guides to stack the plates (and straighten
them into the guide rails and cage), and then drop the plates down
through the stack (FIFO (first in first out)) and load each plate to a
specific slot in a dish washing machine - which I admit might look
like a complicated mechanical arm. Actually, even better, most
dishwashers have a removable rack for dishes, so just have people load
that up when they are placing their dishes on the conveyor belt.
Anyway, the point is that it's far from where it should be. As for
laundry, throwing the clothing into a giant wad is trouble for
processing, and individual container/bag-based washing is the only
idea I've been able to come up with, which would seem to require
significant overhead.
Paul D. Fernhout
> Though automata as a form of art, puppetry, parlor entertainment, or
> stage magic may actually go back as far as Roman times, the general
Eric-
I'm always impressed with your command of the history of technology and
related analysis. I agree with you about the anthropomorphic and
entertainment issues. Thank you for sharing that.
I surprisingly liked the Newton video (never seen it before)
http://www.youtube.com/watch?v=3qdXkJVQUfM
and can see that they were still ahead of their time, because even without a
manipulator, it was essentially a security watch dog (but with personal
computer functions like education thrown in). Everything they hyped there
could really be done now in a solid way as far as I can tell. There is also
a big market for home and office security devices now, for good or bad. I
see several other security robot designs these days -- this is my favorite
based on looking just the right amount of scary and walking:
"Banryu Guard Dragon Home Robot"
http://www.technovelgy.com/ct/Science-Fiction-News.asp?NewsNum=952
A cheap current item with similar functionality in many ways (but not as
intimidating :-):
"Erector Spykee Robot"
http://www.youtube.com/watch?v=7j0TCvfnOpA
Now, personally, I was more interested in the 1980s in robots with a
manipulator (the Silent Running Drones have always been my ideal), and I
never would have bought something like the Newton. But now I can see the
potential appeal of a mobile sensing platform which does not much else.
Although, following your theme, it probably is cheaper and easier just to
put web cams and smoke detectors everywhere in a small house. But, it is
still nice to have multiple approaches to a situation, so it could provide a
nice mobile backup sensing system. And combine that with the right
personality entertainment software, and maybe we will yet see a Newton-like
robot. :-)
On the mobile tool platform you mention, you may not believe this, but I
thought of something related about ten years ago, but not quite at the level
of detail you did. I might have been inspired by this Star Trek episode with
robots called "exocomps":
http://memory-alpha.org/en/wiki/Exocomp
"The exocomp consisted of a micro replicator, a boridium power converter and
axionic chip network. This axionic network gave the exocomp formidable
computational power. The micro replicator not only created tools which the
exocomp could use to solve problems but also created new circuit pathways in
the exocomps memory when it performed new tasks. This mechanism gave the
exocomp the ability to learn. The more tasks it had to perform, the more
pathways were formed in its memory. Problems to be solved could be entered
through a command-pad, after which the exocomp decided what kind of tool it
had to use and then replicated that tool."
Anyway, my thoughts had been about a robot that could follow me around the
house and, like R2D2 holding Luke's lightsaber in one movie, this "toolbot"
would carry all my hand tools somehow, and I could ask for a tool (by voice)
and it would hand the right one to me. Then when I was done with a tool, I
could hand it back and it would store it away. And if it could do that, then
why not have it store common parts, like nuts and bolts and wire? As I
reflect on that, that idea was a predecessor to the general idea of an
internet of packets of things but on a small scale.
I totally agree that we have not yet tapped the full potential for
specialized devices around the home. But, I can also wonder what the overall
effect of more and more devices are as they begin to interact -- for
example, getting a robot to clean the dog hair out of our Roombas. :-)
Or, as another example, I'd use one of several juicers more often if I had a
robot that cleaned them easily and without the worry of cutting my fingers
by accident on the blades. A self-cleaning food processor or self-cleaning
juicer is a different variation on that, or a self-cleaning Roomba -- that
would go more with the single appliance model, but somehow I feel a separate
device might work better for each one.
On your concept, have you seen this?
"Robotic mule"
http://www.youtube.com/watch?v=y3W8dm5JxFc
About thirty seconds in, someone even kicks it and it stands upright. When I
hung around the CMU Robotics Institute in the 1980s, I met people in Marc
Raibert's lab in passing (great people), and I believe that is an extension
of his lab's work.
http://en.wikipedia.org/wiki/Marc_Raibert
My wife, a keen study of animal behavior, immediately suggested that it
looked like two people carrying a litter. Related:
http://en.wikipedia.org/wiki/BigDog
So, these kind of things are getting more and more doable.
The open manufacturing relevance? Well, maybe, like you suggest, such
devices in partnership with people, may make it easier to make all sorts of
structures in outdoor locations more quickly.
Actually, on this theme, a major time cost in using machine tools is
positioning clamps to hold materials worked on. Even back in the 1980s I
know there was work to automate that. So, devices that hold things easily
will also lead to breakthroughs in lowered cost and greater flexibility in
producing things is small batches (or single items) in a machine shop.
Or to try to take your idea one step further (maybe you've thought of that
already) by putting flexible material holders (and you mention related
things) on the "Big Dog", that might lead to one person being able to do
more carpentry quicker. Another thing to add is laser measuring tools and
maybe ever a smart but "safe" circular saw,
http://dethroner.com/2007/07/26/how-a-safe-circular-saw-works/
and before you know it, you have a truly all terrain mobile carpentry buddy.
Maybe it has a display screen for plans or checklists. Give it a mild and
humble "good mule" personality, and carpenters might love it. And since many
carpenters need to bring in a generator anyway to a work site, that's a cost
savings right there to make the economics of the device a little better,
since usually these outdoors robots use big generators.
Bryan's absolutely right in his suggestion it another followup that
redesigning the dishwashing task in a cafeteria would save a lot of time and
effort. I've seen lines like that with belts for ten or twenty years -- why
has this not happened? Even with perhaps a dedicated dishwashing machine for
each item?
Anyway, to address this systematically, we would need to make a detailed
list of activities people really do around the home (even tiny things) or on
the job site, and think in detail about how each little thing might be
automated alone, redesigned to be unneeded, be automated as part of a
package, or be just made really fun to do, so doing it is a non-issue.
From:
http://www.whywork.org/rethinking/whywork/abolition.html
"What I really want to see is work turned into play."
So, with that last point, we may even see a convergence of the utility
aspect of automation and their entertainment aspects. :-)
Maybe someone would want their helping robots to act like the three stooges,
for fun. :-)
"Number 5 and the Three Stooges"
http://www.youtube.com/watch?v=kC1LSSL-d50
In that video, anyone can see what a difference personality can make in a
robot. :-)
--Paul Fernhout
Bryan Bishop
> Anyway, my thoughts had been about a robot that could follow me around the
> house and, like R2D2 holding Luke's lightsaber in one movie, this "toolbot"
> would carry all my hand tools somehow, and I could ask for a tool (by voice)
> and it would hand the right one to me. Then when I was done with a tool, I
> could hand it back and it would store it away. And if it could do that, then
> why not have it store common parts, like nuts and bolts and wire? As I
> reflect on that, that idea was a predecessor to the general idea of an
> internet of packets of things but on a small scale.
> The open manufacturing relevance? Well, maybe, like you suggest, such
> devices in partnership with people, may make it easier to make all sorts of
> structures in outdoor locations more quickly.
>
> Actually, on this theme, a major time cost in using machine tools is
> positioning clamps to hold materials worked on. Even back in the 1980s I
> know there was work to automate that. So, devices that hold things easily
> will also lead to breakthroughs in lowered cost and greater flexibility in
> producing things is small batches (or single items) in a machine shop.
Vern, mentioned in the last email, was last spotted working on a
robotic arm for holding third tools and clamping things down while he
does his soldering. I keep picturing Dr. Octopus when he brings this
up-- a Marvel comics character that had multiple arms attached through
his spinal cord. Wearing it might not be as great an idea as something
stationary or with wheels on the ground or on rails from above for
better positioning ability and movement stability. It was either
talked about in person or lost somewhere in the mailing list archives:
http://lists.puremagic.com/pipermail/robotgroup/
Basically it's for those situations where you wonder where your
requisite third arm is :-). This is kind of like expanding your
workspace too, in the sense of adding multiple monitors to a personal
computer, which really substantially helps with information
intensive-tasks (API documentation reading especially).
Doesn't everyone else find it funny how there's a poor sap who goes
through more than a decade of education and many quarters of millions
of dollars into debt, just to go through medical school to be asked
"Scalpel?", only to then go on to hand somebody the scalpel. Just
seems kind of like a downer for that poor assistant. That's something
an overhead arm should be built for.
In previous houses I've lived in, we've had entire walls for our tools
and so on, in one house there was a 'shed' built into the back of the
garage, and there was a systematic filling of the tools and the
hanging of such; SMT components filed into bins and drawers in the
electronics labs; so there's definitely some room for automatic tool
acquisition. As for the "last mile problem", of physically handing or
accepting a tool without dropping it, I'm guessing that's an issue of
getting the tool within the field of vision of the worker, leaving the
"glanceless handoff" for a thing of the (near?) future.
> Anyway, to address this systematically, we would need to make a detailed
> list of activities people really do around the home (even tiny things) or on
> the job site, and think in detail about how each little thing might be
> automated alone, redesigned to be unneeded, be automated as part of a
> package, or be just made really fun to do, so doing it is a non-issue.
Dishes can be replaced with disposable dishes, or in the ideal
situation we can just get rid of the whole eating concept and go
straight to intravenous feeding (ok, maybe not as feasible). As for
clothing, you could stop wearing. Okay, maybe not. There was something
on Slashdot a while back about titanium dioxide or with some silver as
useful with fabrics for making it virtually non-dirtyable.