Tutorial eBooks PDF File Download

For this surveillance robot, the main purpose is only to wander around the corridors, in order to detect intruders, using a light or movement following behavior. It is not desired to have the robot following walls too closely or having all the space covered, since the main purpose is just to “look around” for motion or light sources, keeping as far as possible from all walls. The robot keeps itself faraway from walls by having the capability of adapting its behaviors, in order to be able to enter and explore narrow spaces and to adapt easily to differently reflective environments. Earlier work done with Dr. Gregor Schöner [1], in the field of target acquisition and obstacle avoidance through optic flow of the retina image, inspired us to use some of the basic dynamic equations techniques for the motor speed control and adaptive threshold process. The adaptive threshold was also strongly inspired through another work in neural networks [2], where we studied the mechanisms and dynamics of single neuron learning and habituation. In surveillance robots, we expect to get an overall behavior which permits an efficient coverage and exploration of the entire space where it should wander about, without the problems of getting indefinitely stuck in particular confusing situations that may occur seldomly.