Robots and Automatic Factories
| Date | 01 November 1981 |
| Published date | 01 November 1981 |
| DOI | https://doi.org/10.1108/eb057217 |
| Pages | 2-7 |
| Author | W.B. Heginbotham |
| Subject Matter | Economics,Information & knowledge management,Management science & operations |
Robots and Automatic
Factories
W. B. Heginbotham, OBE, DSc
Director General, PERA, Melton Mowbray
Industrial robots bear little resemblance to the science
fiction concept of futuristic machines - they are little
more than machine tools except that their prime func-
tion is handling not making components. Being designed
to replace a human operator, many of the machines cur-
rently available are somewhat similar in scale to a
human being although their continuous weight lifting
abilities can be much in excess of a human. They gener-
ally have one manipulating arm and there are three clas-
sical types of robot architecture:
(1) Cylindrical
(2) Polar
(3) Arm and Elbow
Their control and positional systems can vary to pro-
vide:
(i) Point to point motion; and
(ii) Continuous path control.
The point to point systems can be controlled by simple
stops,
peg drums, plug boards, paper or magnetic tape,
although the latter facilities are more common for con-
tinuous path control. The continuous path systems are
inherently more sophisticated and costly and are neces-
sary only for operations where contours are to be traced
or a known path of arm motion is necessary in order to
avoid obstructions.
Robot designers originally conceived industrial robots
to fulfil three main requirements:
(a) To provide an "off-the-shelf solution to the gen-
eral problem of handling in industry;
(b) To provide a machine that was capable of auto-
mating variable batch production situations; and
(c) To provide a virtually direct replacement for man-
ual operation of production machines which were
originally designed for manual operation.
The accent has been on the ease of arm motion
reprogrammability in terms of positionally determined
situations.
On seeing an industrial robot for the first time it
appears to be almost humanoid in its capabilities and
one can be deluded into thinking that here at last is the
direct replacement for a man. No other delusion can be
more dangerous if it is taken too literally. The facts fall
better into place if one looks at the device more as an
"automated crane" controlled by an "electronically
determined" cam system. The most versatile feature of
the device is contained only in its ability to have its arm
reprogrammed to a set of space/time determined co-
ordinates. Once the programming is carried out the
machine reverts to a mechanisitic piece of automation
which will carry out a sequence of events more or less
completely unconscious of its environment.
Therefore, if the operation demands that all that is
required is to record the space/time related skill of a
manual operator, then success should be achieved. For
instance, one could record the movements of a conduc-
tor for an orchestra, the robot arm would on future
occasions faithfully conduct the orchestra in precisely
the same way as the conductor had done. It would,
however, be completely unaware of the fact that the
first violin was absent, or indeed, even whether or not
the orchestra was there at all! One could equip each seat
with a sensor to detect the presence of players and
inhibit the robot motion till all were present but even
then it could be fooled by a "non player" sitting in any
chair. This simple concept completely describes the
extent of the "robot intelligence" at this time and pre-
vents one from expecting too much of such devices.
To summarise therefore, robots are ideal where:
(a) Low accuracy curve tracing is required;
(b) No gripping is necessary;
(c) Where a standard non-variable gripping surface
exists naturally on the component (i.e. die cast-
ing);
(d) Where the skill of the manual operator is restricted
to judgements which result in a clearly defined
space/time pattern; and
(e) On relatively dedicated precision applications
when sufficient peripheral "tooling" can "clean
up"
the final component position to the required
accuracies. This, of course, restricts the general
intercomponent versatility.
Hence the full potential for robot versatility within a
particular field of activity depends on how that activity
fits the robot architecture and technology. Typical oper-
ations can be arranged into different categories as fol-
lows:
Category A
Operations such as spray painting, shot peening, spot
welding, stud welding, flame cutting, applying gasket
sealing compound, coating, marking, heat sealing, glass
cutting, water jet cleaning, drilling and any routines
where no gripping is required.
An industrial robot seen for the
first time appears to be almost
humanoid in its capabilities
2 INDUSTRIAL MANAGEMENT + DATA SYSTEMS
To continue reading
Request your trial