The challenge of sharing and communicating information is crucial in complex human-robot interaction (HRI) scenarios. Ontologies and symbolic reasoning are the state-of-the-art approaches for a natural representation of knowledge, especially within the Semantic Web domain. In such a context, scripted paradigms have been adopted to achieve high expressiveness. Nevertheless, since symbolic reasoning is a high complexity problem, optimizing its performance requires a careful design of the knowledge. Specifically, a robot architecture requires the integration of several components implementing different behaviors and generating a series of beliefs. Most of the components are expected to access, manipulate, and reason upon a run-time generated semantic representation of knowledge grounding robot behaviors and perceptions through formal axioms, with soft real-time requirements.
View on arXiv