In our short editorial, titled "Lacking the Common Touch" and reproduced below, we estimate that a robot revolution, with parallels to the computer revolution, will begin soon after "personal" robots arrive.

(C) Reproduced with permission from Robotics World -
Douglas Publications, Inc. Copyright 1995.

/******************
*** TEXT BEGINS ***
******************/

LACKING
THE COMMON TOUCH

 For the thirty-some years following their introduction in the
late 1940's, electronic digital computers were rarely seen,
touched or used by ordinary citizens. Then one day "personal"
versions appeared, and today, about 15 years later, PC's have
covered the earth and the release of a new operating system is
big news at offices, schools and households.

 "What on earth is an operating system?" asks the citizen from 1980.
 "Well, it's hard to explain," replies the one from 1995.

 Likewise, the first industrial robots appeared thirty-some years
ago, and robots today are rarely seen, touched or used by
ordinary citizens. It's my guess that robots will cover the
earth only after "personal" ones are available and a large
number of garage roboticists bring them to life.
  As with computers, I picture that a mix of large and small
companies and university labs will provide the tools, and
startups as yet unknown will provide the break through,
real-world software.
 Here's the situation. Most of today's robots depend on having
enough absolute accuracy for their task. In contrast, animals
have low absolute accuracy but high resolution and lots of
sensors. If you close your eyes and try to touch your index
fingers together off to your left somewhere, you probably won't
succeed because of low absolute accuracy. But if you open your
eyes, you can touch your fingers precisely together,
compensating for low accuracy with high resolution and sensors.
 Now, a low-accuracy, high-resolution system can be made at much
less expense than a high-accuracy system, since the requirements
for physical rigidity and precise, downstream joint encoders are
removed. As far as I know, no such robot arm is mass-produced.
It might look like this:
 List price $1500. Connects to a PC running Windows. Strong
enough to be moderately dangerous.
 Motors inexpensive and easily replaced. Structure and gearing
largely plastic. Two inexpensive black & white pinhole cameras
fixed to the wrist.
 Two jointed fingers and opposable jointed thumb with
pressure-imaging fingerpads. Repeatability of 5 mm; resolution
of 0.5 mm.
 And here's who buys it:

                            Educators

 When they put a hands-on system like this in a classroom, the
students just catch fire. Operating in teams, 10th-graders
select a task and within an hour they've gotten the idea and are
using glue guns and popsicle-sticks to build fixturing to hold
work pieces for the arm.

                         Process Researchers

 While prototyping new chemical, physical, and biological tests
and production sequences, these developers often need a simple
lab tool that can repeat a set of motions a thousand, but not a
hundred thousand, times. A prototype lab sequence for an
experimental chemical test might involve computer-controlled
instruments like an electronic scale, pH meter, spectrometer and
hot plate, and an arm to move samples between them. To let the
arm tune its position to 1 mm at crucial operation sites, target
decals are strategically affixed close by and the wrist cameras
lock to these distinct decals rather than trying to lock to
indistinct lab objects.
                                             
                       Decision Makers in Firms
                             With No Robots

 They want to see what a robot is, in their own shop, operated
by their own techs and without the need for an "integrator". A
manufacturer with 200 workers at sewing machines assembling
brassieres quickly decides it's not yet time to think about
robots.
 A production machine-shop owner, after watching machinists
experiment with their inexpensive new toy, decides to look into
a real shop arm that could feed parts from a carousel to a
machine tool.

                               Crazies

 For the price, they would buy one just to have it, with the
hope they can come up with a marketable add-on to the base
product. These users are akin to the crazies who bought the
first personal computers, and whose story is known.

 Over time, developers of all kinds will draw from the
primordial soup of neural nets, evolutionary programming and
other bio-related notions contributed to the Internet by
researchers. Developers' implementations will then flow back
across the net for use on customers' standard robots.

 "What on earth is an ideation strophe?" asks the citizen from 1995.
 "Well, it's hard to explain," replies the one from 2010.

________________________________
 Evan Rosen is President of Advanced Design, Inc.,
Tucson, AZ which sells six-servo robot building sets
to educators. Rosen made up the phrase "ideation strophe"
and doesn't know what it means either. The company's
website is http://www.robix.com

/****************
*** TEXT ENDS ***
****************/