Abstract

This paper presents a signal processing architecture for a sensory-motor system based on the smart sensor paradigm. The architecture is designed for an obstacle avoidance task by a mobile robot in an unstructured environment. Drawing inspiration from the field of behavior-based robotics, the development of the architecture is guided by an emphasis on the requirements of an obstacle avoidance behavior for a mobile robot. The architecture is simple enough for a smart sensor, but incorporates features which enable it to deal with realistic, unstructured environments. It differs from existing systems by using a special foveation scheme to facilitate the detection of real world objects. From this, a motor control signal is produced by using a biologically inspired technique of aligning sensory and motor maps. The effectiveness of the architecture is explored through computer simulation, including an obstacle avoidance simulation in a 3D virtual environment.