Robotics - Sensory Systems

The component of modern robots that was most commonly missing from their early predecessors was the ability to collect data from the outside world. Humans accomplish this task, of course, by means of our hands, eyes, ears, noses, and tongues. With some important exceptions, robots usually do not need to have the ability to hear, smell, or taste things in the world around them, but they are often required to be able to "see" an object or to "feel" it.

The simplest optical system used in robots is a photoelectric cell. A photoelectric cell converts light energy into electrical energy. It allows a robot to determine "yes/no" situations in its field of vision, such as whether a particular piece of equipment is present or not. Suppose, for example, that a robot looks at a place on the table in front of it where a tool is supposed to be. If the tool is present, light will be reflected off it and sent to the robot's photoelectric cell. There, the light waves will be converted to an electrical current that is transmitted to the robot's computer-brain.

More complex robot video systems make use of television cameras. The images collected by the cameras are sent to the robot's "brain," where they are processed for understanding. One means of processing is to compare the image received by the television camera with other images stored in the robot's computer-brain.

The human sense of touch can be replicated in a robot by means of tactile sensors. One kind of tactile sensor is nothing more than a simple switch that goes from one position to another when the robot's fingers come into contact with a solid object. When a finger comes into contact with an object, the switch may close, allowing an electrical current to flow to the brain. A more sophisticated sense of touch can be provided by combining a group of tactile sensors at various positions on the robot's hand. This arrangement allows the robot to estimate the shape, size, and contours of an object being examined.