ROBUST AUTONOMOUS ASSEMBLY IN ENVIRONMENT WITH RELATIVELY HIGH LEVEL OF UNCERTAINTY

Abstract

This paper deals with issues of robustness and adaptivity in a real-world automated assembly system. The focus of research is set on robustness achieving for constantly changing environmental conditions, especially light conditions and changing orientation, position and distribution of the parts for assembly. An increase in the level of information/knowledge used by the robotic system means a relative increase in the level of control. Combination of different types of sensors (vision ‐ mobile camera, laser, optical, etc.) is used as a main perception for exploration of the robots’ work space. Developed application shows that proper combination of sensors ensures consistent information flow. In this manner stochastic information variation is minimized and robustness of the system increased.