What became of the personal robot, and comments on design

[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. "

I more or less agree with the overall theme of the article as far as it
goes, suggesting that walking, dexterity, sensitivity, seeing, thought, and
multitasking are hard (much harder that we thought), and further that there
are ethical issues involved with creating sentient creatures (whether we
then enslave them or not). In those ways, it is an excellent article. And it
even concludes with the point that "But all is not lost. Though the golden
man is destined to stay on the screen and in our imaginations, his trusty
sidekick R2D2 is on the point of bursting through the celluloid and into our
houses." Which is also true, and a good prophecy.

What the article misses though is how much "invisible" automation is in our
lives now. One hundred years ago, it was predicted that every home would
have *one* electric motor someday. (Belts would route the power to a variety
of appliances.) And fifty years ago, it was thought by some of the leaders
in the industry that there was a market for at most, about seven big
computers worldwide, owned and operated by big research institutes and big
businesses. Now, the typical home in the USA has at least one personoal
computer (with computers embedded in many appliances), and the typical
personal computer has several motors (fans have a motor, CD drives have two
or more motors). So predictions like those can miss out on the big picture.
US American life is very different in many ways than fifty years ago because
of electric motors and computers (the essence of robots). Compared to the
1950s, and ignoring the Roomba, the average US American home has multiple
robots, that don't even look like R2D2. First, there is the refrigerator
robot that does not need ice deliveries, because it makes its own cold, and
which defrosts itself too. Then there is the microwave robot that heats up
food with little attention. There is the dish washing robot which cleans
dishes with minimal hot water. There is the clothes washing robot that
cleans and drys clothes (two robots or one, depending on how sophisticated,
we have one of the combo ones and like it -- a LG WM3677 Washer and Dryer
Combination). *And* there are a vast number of humanoid entertainment robots
in the typical US home, but for reasons of cost and convenience, they only
show up on the television or computer screen in electronic games. :-) And in
the garage, there is a car that has about a dozen computers running motors
and solenoids and other electronics in electric windows, electronic
ignitions, electronic brakes, electronic accelerators, electronic safety
systems like air bags, electronic navigation and guidance system (GPS),
communication systems (ONSTAR), and so on.

I can only wonder if thinking about the "singularity" might be better
informed by considering this process as well, as a sort of evolution of
human-machine symbiosis directed by a social network of incremental small
decisions, upholding some core human values (or aesthetics) while still
looking at change.

One reason we have so many non-obvious robots involves issues of design.
When you look at a problem, like washing dishes, rather than do it the
humanoid way, one dish at a time in soapy water, there is often a more
efficient way to do it that entails redesigning the solution (put all the
dishes in a box together at once and spray the same water at them over and
over). People used to use washboards and wring out each item of clothes,
putting a bunch of cloth items in a tumbler is also more efficient. (In the
Star Trek universe, even in the 1960s and 1970s, it was envisioned that it
would be easiest to take apart clothing into the constituent fibers, clean
the fibers, and reassemble them back into a uniform; I thought that
tremendously wasteful back then when I read of it back then on the Star Trek
Enterprise blueprints -- but now I can imagine how it might be more energy
efficient and also produce higher quality cleaning.)

Outside the home, there is also automation that makes many prepared items
previously made in the home, whether frozen dinners, pre-hemmed pre-washed
jeans, pre-grown seedlings, or pre-sorted information. Again, these all
reduce the desire to have a humanoid robot to do our cooking, sewing,
gardening, or searching.

So, while one might say in complete truth, like in that article, look, there
is no obvious robotics in our lives, what has instead happened is a redesign
of the tasks of daily living to make them easier and to hide the automation.
(Of course, now we deal with replacing or repairing appliances when they
fail, and earning the money to buy them, all other sources of stress.) The
same in general has been said of computers, that it is the "invisible"
computer that is reshaping our lives more than the visible one. Computers
are going from one per home to literally disappearing into the walls and
appliances. Including toys for a kid, I don't have any idea how many circuit
boards and motors I have at home -- possibly hundreds of each.

A big argument for humanoid robots has always been that they can easily fit
into and navigate human dwellings, and that they can use human tools. (Isaac
Asimov's robot stories show a lot of this. robot gardening with hand tools
and turning valves by hand.) But if we redesign tasks so they are not done
by humanoid forms, like dishwashing, then there is less need for this. Also,
humans can do the things humans do, and many humans do like to do some of
them to some degree. Many humans appreciate the privacy of not having
servants with cameras, microphones, and wireless connectivity in their home
(which any modern robot would have). So, the emphasis instead has been on
getting machines to do what humans can't do easily or don't want to do
easily, or in redesigning those tasks so they are more fun or unneeded.

For example, what about a tool to make cleaning bathrooms fun?
http://www.whywork.org/rethinking/whywork/abolition.html
"Is it not odd that farm workers toil painfully in the fields while their
air-conditioned masters go home every weekend and putter about in their
gardens? ... The secret of turning work into play, as Charles Fourier
demonstrated, is to arrange useful activities to take advantage of whatever
it is that various people at various times in fact enjoy doing."

I kind of imagine some kind of vaccum cleaner like thing with a wand with a
spray line and a return line that is self-cleaning itself. We had an earlier
Bissell cleaner that was a little like this but hard to clean. I see they
may have improved on that (below), but I wonder how easy even that is to
clean or if it is sanitary or safe or pleasant to use in bathrooms. I'll
quote the blurb as it is an interesting statement about the future of home
appliances that are somewhat robotic (instead of humanoid robots) and which
are designed considering some "cradle-to-cradle" manufacturing issues.
"Bissell 1400B Little Green Multi-Purpose Compact Earth-Friendly Deep Cleaner"
http://www.amazon.com/Bissell-1400B-Multi-Purpose-Compact-Earth-Friendly/dp/B0016HF5GK
"Equipped with a 2-3/4 amp motor and safe to use on virtually all carpets,
this compact-size Little Green multipurpose deep cleaner can handle both
small and large clean-ups, even stubborn spots and stains, as effectively as
a full-size machine. It combines a cleaning spray with powerful suction to
lift spills before they set in--much better than store-bought chemical
cleaners, which get sprayed on or rubbed in, pushing stains deeper into
carpet or upholstery fibers. The deep-cleaning machine provides a 4-inch
cleaning path, a flexible vacuum hose, and a 15-foot power cord for
accessing hard-to-reach areas. In addition, it uses earth-friendly liquid
cleaning formulas that do not contain any heavy metals, optical brighteners,
phosphates, or dyes. Try the Bissell Pet Odor and Soil Removal solution for
getting rid of tough odors, tracked-in dirt, dander, and other soils from
pets. Designed to keep maintenance to a minimum, the deep cleaner can be
stored ready-to-use and requires nothing more than simply rinsing out the
dirty-water tank after each use. A tough-stain brush attachment comes
included. Suitably named Little Green, the earth-friendly deep cleaner has
been engineered with sustainable features to reduce waste and help protect
the environment. The unit's redesign offers a 95-percent reduction in PVC.
It includes a PVC-free hose, PVC-free solution and collection tanks (made of
recyclable polypropylene), as well as parts constructed from 100-percent
post-consumer recycled plastic. Even its carton and inner packaging are made
from a minimum of 75-percent recycled material--all Styrofoam and PVC
inserts have been eliminated. Bissell has made a strong commitment to a
healthy environment, and its innovative closed-loop manufacturing process
allows new vacuums and deep-cleaning machines to be manufactured using
post-consumer recycled content. The Little Green deep-cleaning machine
measures 9-1/10 by 18-4/5 by 13-1/4 inches and carries a one-year limited
warranty. "

Of course, one can imagine building that sort of thing into a self-powered
base like a bigger Roomba. The Scooba robot spits out water with cleaner and
sucks it back in to clean kitchen and tile and sealed wood floors. But it
only does floors; it can't handle cleaning toilets and bathtubs and walls.

The BBC robot article is kind of ironic in terms of timing, because looking
for it again on a different computer where I first saw it, and putting in
"robot where home" into Google News, produced so many articles about robot
innovation I haven't read them yet. But the first one:
"At Home With Robots: The Coming Revolution"
http://www.technewsworld.com/story/At-Home-With-Robots-The-Coming-Revolution-65290.html
"Normally, we think of robots as humanoid and self-aware, like Rosie or
Data. Some day, they might be more like that, but robots are here in our
homes today, cleaning our floors and even making our coffee. What does the
future of human-robot relations hold?"

Or also:
"Pensonic to unveil robotic products"
http://biz.thestar.com.my/news/story.asp?file=/2008/11/25/business/2604842&sec=business
"Pensonic Holdings Bhd is targeting next year to introduce a new range of
robotic consumer electronic products to mitigate the impact of a global
recession. Group managing director Dixon Chew told StarBiz that the group
planned to introduce the robotic products, in the shape of dinosaurs, pets
and humans, that were designed to perform basic household chores, provide
entertainment and look after children."

That's very funny -- a company introducing robots to combat a recession
(their own company's experience of a recession of course). Like what do
robots do, give people more jobs?
http://marshallbrain.com/robotic-nation.htm
"There is no mystery -- the jobless recovery is exactly what you would
expect in a robotic nation. When automation and robots eliminate jobs, they
are gone for good. The economy then has to invent new jobs. But it is much
harder to do that now because robots can quickly fill the new jobs that get
invented. See the FAQ for additional information."

And then there is Willow Garage, who are working on a home robot.
http://www.willowgarage.com/
And lots of Japanese innovation relating to care for the elderly.

As we think about manufacturing design in an existing context, and how to
detail it from the ground up in an open way, the role automation should play
in that is not always clear. If you design manufacturing tasks for people to
do, you may design them in a very different way then if you design them for
automation (including robots) to do. In the past, that has had a lot to do
with whether you make one of something or whether you set up an assembly
line. But with flexible automation (including 3D printers), it is becoming
an issue even when you make one of something. So, is it "open manufacturing
for robot artisans and automation", or "open manufacturing for people", or
"open manufacturing for robots, automation, and people"? :-)

One of the differences between Bryan's SKDB and OSCOMAK is that Bryan
envisions SKDB as "apt-get chair" which pulls in everything you need to
automatically make a chair from your fablab (even making the tools to make
the tools), and I was looking at OSCOMAK as more a human accessible thing
for a person who wants to make a chair themselves (more like several of the
how-to sites on the internet, but with a lot more metadata to do design of
manufacturing webs and redesign products to fit into them easily). I myself
had started out wanting a full automation system of a self-replicating space
habitat (inspired by James P. Hogan's "The Two Faces of Tommorrow" book with
something like that) so I'm not opposed to such systems (or at least I
wasn't then, I don't know as much now :-). I would think we'll have a full
range of designs eventually, of course. And as we get better 3D printers,
more and more will be made using them because it is easy and convenient. And
designs for them will likely be specified more in terms of the end goal
rather than how to build it, since it will be easy enough to write software
that figures out how to construct something if it is specified in enough
detail. So, an open format for specifying physical objects in detail seems
like some common ground. There are a bunch of such formats already of
course, though I don't know how non-proprietary and well documented they all
are. (I've seen mostly the 3D graphics ones, not the CAD manufacturing ones.)

--Paul Fernhout