Ontology-Based Cognitive System for Contextual Reasoning in Robot Architectures

Abstract

We present a hybrid system for reasoning over contextual and spatial elements of environments. The system, designed for deployment in robot architectures, leverages a knowledge base framework for common-sense reasoning through the integration with the symbolic and sub-symbolic structures of a cognitive architecture. Methodological and functional aspects are illustrated, together with an overview of the results obtained in a synthetic simulation experiment.