Quick and clean: constraint-based vision for situated robots

Knowledge-based vision for robots needs a radically new approach. The traditional approach has not made substantial progress for various reasons including the engineering problems of building systems based on a hybrid on-line/off-line paradigm. A new situated agent approach is presented. The constraint net model of Y. Zhang and A.K. Mackworth allows the designer to specify the robot´s vision, control and motor systems uniformly as on-line systems. If the perceptual and control systems are designed as constraint-satisfying devices then the total robotic system, consisting of the robot symmetrically coupled to the environment, can be proven correct. Examples of this approach are given