Image Courtesy of http://www.societyofrobots.com/sensors_sharpirrange.shtml
How A Robot Detects Its Own Motion
One of the most important types of movement for a robot to detect is its own. Robots can do this using what is called an accelerometer. An accelerometer measures acceleration with tiny levers. When the robot accelerates in one direction, it moves the lever, which is detected by the circuitry.
This technology can be used by robots that walk upright and have to balance themselves. It can also be used to detect angle and G-force in autonomous aircraft, and many other sorts of robots.
Simple Motion Detection Techniques
The easiest way for a robot to detect motion is by bouncing a signal off anything repeatedly and measuring the difference between two consecutive signals. In theory, the signal can be pretty much anything--a sound, a radio wave, a beam of light. In practice, it should be something easy to produce and measure. It should also not be something like sound that the environment around t produces a lot of, since this will throw off the readings.
One of the most common ways for robots to detect motion is by bouncing an infrared signal off nearby objects. It detects the angle at which the object bounces off the object to determine the distance. A close object will bounce the signal off at a sharp angle, whereas a faraway object will bounce the signal more directly. By measuring the difference between two signals, a robot can see if any objects around it have moved.
Advanced Motion Detectors
One of the most advanced techniques in motion detection is computer vision. Robots can look at images from a video camera and detect real-world objects around them. These techniques require sophisticated computer processing. The robot learns to detect the edges of objects for 3-dimensional representations of its environment and interact in a complex, lifelike way.
Some autonomous robots use all of these tricks and more. A target drone, for example, will compare a global positioning signal with maps in its memory to navigate to its target location. It will also bounce radar signals off the ground and surrounding objects to help it navigate and avoid collisions. During its flight, It will use accelerometers to monitor and control its own movement. When it gets to its target, it may use computer vision to recognize the building it is supposed to attack.
Tags: bounce signal, computer vision, detect motion bouncing, difference between, measuring difference, measuring difference between