Semi-autonomous teleoperation control system with layered architecture-An application to space robots

Abstract

A layered architecture was proposed for the command and control system of semi-autonomous space robots, and the proposal was validated by the demonstration of replacing an orbital replacement unit on a testbed for space robots. The command and control system consists of four layers: a task scheduling module, a macro-task module, a micro-task module, and an intelligent control module. The macro- and micro-task modules were implemented on top of Lisp and have a symbolic world model which is integrated with an online robot simulator developed upon a commercial robot simulator for industrial robots. The task decomposition and expansion from macro commands to micro commands was carried out dynamically by consulting the world model, and the impending collision between a manipulator and obstacles was detected by the online robot simulator. A small visual marker which can measure 6-DOF was designed, and a visual feedback system was developed with the visual marker.