Page 45 - Robot Builders Source Book - Gordon McComb

34 Introduction: Brief Historical Review and Main Definitions

oil (or other liquid) passages in proportion to the electrosignal on the inputs of the
valves 3. The oil is collected in a reservoir 9. The power needed to feed the system is
supplied by batteries 10.
As far as is known, the first idea for this type of artificial hand was patented in 1957
in the USSR by N. Kobrinsky, M. Breydo, B. Gurfinkel, A. Sysin, and J. Jacobson. Later,
such hands were created in Yugoslavia, Canada and the U.S.A.

1.6 Relationship between the Level of Robot
"Intelligence" and the Product

The question may arise as to what level of robot will be needed for each particular
area of manufacturing or processing. There is a feeling that, in the immediate (or not-
too-distant) future, all goods will be produced by "intelligent' robots, and no man-
power will be involved in manufacturing processes. In such a scenario, a human being
would command and control the processes by means of computers and pushbuttons.
He would tell the computerized machine to produce, say, strawberry ice cream, and
the machine would be instructed by its electronic (or other?) "brain" to mix the correct
amounts of the right ingredients, to heat or cool them to the right temperatures, to
treat the mixture at the right pressures, and so on. On another day, the system might
receive an order to prepare potato chips, and again the computer would find, in its
memory, the right recipes, would choose the optimal one, would tell the right machine
to do the right job, and so on.
In our opinion this is a childish (although attractive) approach to the future. There
will always be industrial needs for robots of low intellectual level and perhaps also for
manually powered or hand-controlled machines. The reason for this is not solely indi-
vidual preference for some mechanical devices. (For instance, there are many people
who do not like cars with an automatic transmission but prefer those with a manual
gear box.) There are also some objective economic reasons for thinking that "manual"
machines will not be phased out. For example, people will always wear socks or some-
thing like socks. Whatever the material from which they are made, socks will always
be needed in amounts of tens of millions or more. Similarly, people will always need
writing tools—be it goose feathers, fountain pens, ballpoint pens, or electronic pens—
and hundreds of millions of these tools will be needed. Screws, nuts, washers, and nails
are manufactured each year in millions and millions of each shape and size. Food con-
tainers—bags, cans, and bottles—are another example of mass production. Ironically,
even electronic chips used for making robot "brains" are often produced in such large
quantities that flexibility is not required during manufacturing.
All the products mentioned above are characterized by being in demand in large
amounts and over relatively long periods of production. The question is thus whether
it is cheaper and more effective to use specialized, relatively inflexible (or completely
inflexible) machines for such purposes rather than sophisticated flexible robots. The
answer, in our opinion, is "yes." The problem is to define whether it is appropriate for
a particular industrial need to design and build a high-level advanced robot or a "bang-
bang" type of robot.
TEAM LRN