Page 146 - Concise Encyclopedia of Robotics

Groundskeeping Robot
sequence of movements that a robot arm undergoes in an industrial
robotic system.
Gross motion planning can be done using a computer map of the envi-
ronment. This tells it where tables, chairs, furniture, and other obstructions
are located, and how they are oriented.Another method is to use proximity
sensing or a vision system. These devices can work in environments un-
familiar to a robot, and for which it has no computer map. Still another
method is the use of beacons.
Suppose a personal robot is told to go to the kitchen and get an apple
from a basket on a table. The robot can employ gross motion planning to
scan its computer map and locate the kitchen. Within the kitchen, it
needs some way to determine where the table is located. Finding the basket,
and picking an apple from it (especially if there are other types of fruit in
the basket, too), requires fine motion planning. Compare FINE MOTION PLAN-
NING and GRASPING PLANNING.
GROUNDSKEEPING ROBOT
There are plenty of jobs for personal robots in the yard around the house,
as well as inside the house. Two obvious applications for a groundskeeping
robot includes mowing the lawn and removing snow. In addition, such a
machine might water and weed a garden.
Riding mowers and riding snow blowers are easy for sophisticated
mobile robots to use. The robot need not be a biped; it needs only to have
a form suitable for riding the machine and operating the controls. Alter-
natively, lawn mowers or snow blowers can be robotic devices, designed
with that one task in mind.
The main challenge, once a lawn-mowing or snow-blowing robot has
begun its work, is to do its work everywhere it is supposed to, but
nowhere else.A robot owner does not want the lawn mower in the garden,
and there is no point in blowing snow from the lawn (usually). Such a
robot should therefore be an automated guided vehicle (AGV). Current-
carrying wires can be buried around the perimeter of your yard, and
along the edges of the driveway and walkways,establishing the boundaries
within which the robot must work.
Inside the work area, edge detection can be used to follow the line
between mown and unmown grass, or between cleared and uncleared
pavement. This line is easily discernible because of differences in bright-
ness and/or color. Alternatively, a computer map can be used, and the
robot can sweep along controlled and programmed strips with mathe-
matical precision.
The hardware already exists for groundskeeping robots to withstand all
temperatures commonly encountered in both summer and winter, from
Alaska to Death Valley. Software is more than sophisticated enough for

