Page 193 - Concise Encyclopedia of Robotics
P. 193
Local Feature Focus
• How long after their start-up will the robots be closest to each
other?
• Will the robots collide unless they change course or speed?
• What will be the coordinates of robot A at the time of closest
approach?
• What will be the coordinates of robot B at the time of closest
approach?
Computers can be programmed to solve these problems quickly and
easily. Smart robots solve such problems when necessary, without over-
sight by a human operator. These calculations are important if there are
numerous robots in a small work area and the robots lack sophisticated
collision-avoidance sensors.
LOAD/HAUL/DUMP
Load/haul/dump, abbreviated LHD, is a type of mobile robot used in
mining and construction. It does exactly what its name implies. With the
aid of a human operator, LHDs load cargo, haul it from one place to
another, and dump it in a prescribed location.
In mining, LHDs have an easier time than in general construction.
The geometry of a mine is easily programmed into the robot con-
troller; the layout changes slowly. Reprogramming does not need to be
done often. In construction, however, the landscape is more compli-
cated, and it changes rapidly as work progresses. Therefore, the com-
puter maps must be revised often. In mining, all the loads are usually
the same in terms of weight and volume, because each load consists of
a prescribed amount of a single substance such as coal or iron ore. In
construction, the nature, and thus the weight and volume, of the load
can vary.
LHD vehicles use various methods for navigation, including bea-
cons, computer maps, position sensors, and vision systems. LHDs can be
autonomous, although there are advantages to using a single controller
for many robots.
See also AUTONOMOUS ROBOT and INSECT ROBOT.
LOCAL FEATURE FOCUS
In a robotic vision system,it is usually not necessary to use the whole image
to perform a function. Often, only a single feature, or a small region
within the image, is needed. To minimize memory space and to optimize
speed, local feature focus can be used.
Suppose a robot needs to get a pliers from a toolbox. This tool has a
characteristic shape that is stored in memory. Several different images