Page 23 - Concise Encyclopedia of Robotics
P. 23
Analogical Motion
Neural networks
Neural network technology uses a design philosophy that differs radically
from that of conventional digital computers. Neural networks are good at
spotting patterns, which is important for forecasting. Rather than working
with discrete binary digits, neural networks work with the relationships
among events.
Unless there is a malfunction, a digital machine does precise things
with data. This takes time, but the outcome is always the same if the input
stays constant. This is not the case with a neural network. A neural net-
work can work more quickly than a digital machine. To achieve speed,
precision is sacrificed. Neural networks can learn from their mistakes.
According to some scientists, this technology is a diversion and distrac-
tion from the proven mainstream; according to other scientists, it holds
great promise.
AMUSEMENT ROBOT
An amusement robot is a hobby robot intended for entertainment or
gaming. Companies sometimes use them to show off new products and
to attract customers. They are common at trade fairs, especially in Japan.
Although they are usually small in size, they often have sophisticated
controllers.
An example of an amusement robot is a mechanical mouse (not to be
confused with the pointing device for a computer) that navigates a maze.
The simplest such device bumps around randomly until it finds its way
out by accident. A more sophisticated robot mouse moves along one wall
of the maze until it emerges. This technique will work with most, but not
all, mazes.
The most advanced amusement robots include androids, or machines
with a human appearance. Robots of this type can greet customers in
stores, operate elevators, or demonstrate products at conventions. Some
amusement robots can accommodate human riders.
See also ANDROID and PERSONAL ROBOT.
ANALOGICAL MOTION
The term analogical motion refers to a variable or quantity that can have
an infinite number of values within a certain range. This is in contrast to
digital variables or quantities, which can have only a finite number of dis-
crete values within a given range. Thus, analogical control is representative
of so-called smooth or continuous motion.
A person moving freely around a room, varying position to any point
within a specific region, has the capability of analogical motion. The
human arm can move to an infinite number of positions in a fluid and
continuous way, within a certain region of space. This, too, is analog