Ontology-based approach for arranging new objects by humanoid robots

Abstract

The robots are programmed in order to achieve arduous works, perform missions and interact with their environment. However, to realize these tasks the robots must rely not only on their internal knowledge but also on its representation. In this paper, we describe an ontology-based approach to resolve the problem of “Object Arrangement”. The proposed solution is also useful to reason, adapt and communicate with its environment (robots and human). It uses the local and global arrangement algorithms on the three following ontologies: (1) SUMO as the Foundation Ontology, (2) the environment Ontology and (3) the Robot Ontology. The local arrangement algorithm interrogates, the environment and the robot ontologies, to arrange the object if its semantic exists in the environment. Otherwise, it employs the global arrangement algorithm which interrogates the Foundation Ontology SUMO to capture the semantics of the new object; then the local arrangement algorithm is operated in order to arrange it. The experiment is achieved by implementing our personal humanoid robot (Knowledge-based for Humanoid Advanced robots First Generation K.H.A.1) in LITIO laboratory, Oran, Algeria. Finally, the experimental study shows promising results of the proposed approach.