The visual sensing systems are more sophisticated and are usually based on television cameras or laser beam scanners. The signals of the cameras are preprocessed in hardware and can be fed at rate of 30 or 60 frames per second into a computer. The computer analyzes the data and extracts the required information, such as the presence, identity, position, and orientation of an object to be manipulated, or part integrity and completeness of products under inspection.
An acoustic sensor senses acoustic waves and interprets them. The level of sophistication of acoustic sensors varies from a primitive detection of the presence of acoustic waves to recognition of isolated words in continuous human speech. In addition to human-robot voice communication, acoustic sensing can be utilized by robots to assist in controlling arc welding, to stop the motion of a robot when aloud crash is sensed , to predict a mechanical breakage about to happen, and to inspect objects for internal defects.
Finally there is a class of noncontact systems that uses a projector plus an imaging device to obtain surface shape information or range information.
There are two basic methods of using sensors: static sensing and closed-loop sensing. Frequently, sensors are used in robot systems in a mode wherein sensing and manipulation alternate .That is, the manipulator is stationary while sensing is being done, then motion is done without further reference to the sensors. This method is referred to as static sensing. With this method, the vision is used to establish the position and orientation of an object to be grasped, then the robot moves blindly to the spot.