Ontology-Based Knowledge Representation and Task Implementation in Collaborative Robotic Cells for Drilling and Riveting on Aircraft Fuselages

As collaborative robots are introduced into various industrial scenarios, the issue of human-robot collaboration in two heterogeneous multi-agent systems becomes prominent. This paper presents an ontology-based knowledge representation and a finite state machine-based task planning control scheme, which are used for drilling and riveting tasks on aircraft fuselages. The knowledge representation system includes robot ontology, task ontology, and environment ontology. This modular domain ontology design facilitates the definition of object properties and data properties for tasks and agents. Cost matrix and decomposed sub-tasks can be estimated from the ontology, and the robot's state transitions are controlled by a finite state machine. This knowledge representation system has been integrated into the Robot Operating System.