A distributed system for robotic applications

Abstract

A distributed control system for robotic applications is presented. This system allows a real-time performance increase, by parallelism of computation, and offers great modularity and versatility in robot control algorithm implementation. The robot control theory based on task-functions is outlined, and a model for robotic applications is proposed. This philosophy is more general than the classical hierarchical structure and allows smart sensor feedback at the servo. Software and hardware architectures are proposed to implement such control algorithms. All the system activity depends on communication between tasks and so the design of a real-time communication system is considered.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>