Page 202 - Concise Encyclopedia of Robotics
P. 202

Manipulator
                              In a system with a linear-dropoff profile, the sensor output is zero until
                            the robot comes within a specific distance of the object (in this case
                            approximately 6.5 units as shown on the graph).When the robot is closer
                            than the critical radius, the sensor output varies according to a straight-
                            line function with negative slope, as shown, reaching a maximum when
                            the robot is about to strike the object.
                              In a system with an exponential-dropoff profile, the sensor output varies
                            inversely with distance from the radius. There are no abrupt transitions
                            or bends in the curve as with the other two profile schemes. The sensor
                            output drops to zero at a considerable distance from the object; when the
                            robot is about to strike the object, the sensor output is maximum.
                              See also DISTANCE MEASUREMENT and PROXIMITY SENSING.
                         MANIPULATOR
                            A manipulator consists of a robot arm, and the gripper or end effector at the
                            end of the arm. The term can also refer to a remotely controlled robot. See
                            END EFFECTOR, ROBOT ARM and ROBOT GRIPPER.
                         MASTER–SLAVE MANIPULATOR
                            See TELEPRESENCE.
                         MEAN TIME BEFORE FAILURE/MEAN TIME BETWEEN FAILURES (MTBF)
                            The performance of a robot, computer, or other machine can be specified
                            in various ways. Two of the most common are the mean time before failure
                            and the mean time between failures, both abbreviated MTBF.
                            Component
                            For a single component, such as an integrated circuit, the MTBF (mean
                            time before failure) is the length of time you can expect the device to
                            work before it fails. This is found by testing a number of components and
                            averaging how long they keep working.
                              A simplified example of MTBF, calculated in hours on the basis of
                            the performance of five identical, hypothetical light bulbs, is shown in the
                            drawing. The lifetimes are averaged to get the result. For the results to be
                            meaningful,the number of samples must be much greater than five.Testing
                            a large number of components, such as 1000 or even 10,000, eliminates
                            coincidentally skewed results.
                            System
                            In the case of a system such as a robot or computer,the mean time between
                            failures is determined according to how often the machine breaks down.
                            As with the component-testing method, it is best to use many identical




                                                   
   197   198   199   200   201   202   203   204   205   206   207