Page 297 - Concise Encyclopedia of Robotics
P. 297

Sentry Robot
                            two or more redundant sensors to minimize the number of false negatives
                            and false positives.
                              Whenever a single sensor is used to detect a phenomenon or occur-
                            rence in the environment, there is a chance for error. If the output of the
                            sensor is a simple “yes/no”(logic 1 or logic 0), the output will occasionally
                            be 1 when it should be 0 (false positive) or vice versa (false negative). If a
                            sensor detects a range of values such as visible-light intensity, the measure-
                            ment is always subject to some error.
                              Suppose  two  binary-output  (1  or  0)  sensors  are  used  to  detect  or
                            measure the same phenomenon. The output of the combination can be
                            considered 1 if and only if both sensors output 1; the output of the com-
                            bination can be considered 0 if and only if both sensors output 0. Usually
                            the two sensors will agree, but occasionally they will not. In cases where
                            the two sensors disagree, the robot controller can instruct the sensors to
                            take another sampling. In the case of analog sensors, such as those used
                            to measure visible-light intensity, the outputs can be averaged to get a
                            more accurate reading than either sensor produces alone.
                              Numerous competing sensors can be used to obtain much greater
                            accuracy than is possible with a single sensor. In general, the larger the
                            number of competing sensors, the less frequent will be the errors in a
                            binary digital system, and the smaller will be the error in an analog sys-
                            tem. There are various ways in which the sensor outputs can be com-
                            bined  to  obtain  a  result  of the  desired  accuracy, while  maintaining
                            reasonable system speed.
                         SENSOR FUSION
                            The term sensor fusion refers to the use of two or more different types of
                            sensors simultaneously to analyze an object. Examples of characteristics
                            that can be measured include mass (or weight), volume, shape, light reflec-
                            tivity, light transmittivity, color, temperature, and texture.
                              Sensor fusion is used by smart robots to identify objects. The robot con-
                            troller can store a large database of objects and their unique characteristics.
                            When an object is encountered, the sensors provide input and compare
                            the characteristics of the object with the information in the database.
                            Compare SENSOR COMPETITION.
                              See also BACK PRESSURE SENSOR, COLOR SENSING, EYE-IN-HAND SYSTEM, FEEDBACK, JOINT-
                            FORCE SENSING, OBJECT RECOGNITION, TACTILE SENSING, TEMPERATURE SENSING, TEXTURE SENS-
                            ING, and WRIST-FORCE SENSING.

                         SENTRY ROBOT
                            A sentry robot is a specialized type of security robot that alerts people to
                            abnormal conditions. Such a robot can be designed to detect smoke, fires,




                                                   
   292   293   294   295   296   297   298   299   300   301   302