Quantcast

Increased Autonomy in Industrial Robotic Systems: A Framework

Research paper by Krister Brink, Magnus Olsson, Gunnar Bolmsjö

Indexed on: 01 Aug '97Published on: 01 Aug '97Published in: Journal of Intelligent & Robotic Systems



Abstract

This paper presents an event based control system structure for the control of a robot workcell and its implementation. The goal for this control system is to autonomously manage the dynamic environment of a robot workcell. The presented control system is event driven and operates from tasks and a World model, defined in a task oriented programming session. During realisation of the tasks, the World model is continuously updated by information from sensors. The system always operates on the latest information which may result in re-planning of sub-tasks or whole tasks. The autonomous functionality of the presented system is established through reactive re-planning and configurable corrections. A high level adaptation of the model in the control system to the workcell is of great value for the performance of the robot system. An important and efficient use of robot motion control for application process controlling purposes are enabled through the control systems internal configurable interaction with the environment.