Knowledge Representation and Reasoning for Ubiquitous Robotics
Robotic and intelligent systems intending to interact at a symbolic level, require a connection between the abstract and the physical levels of representation. The term anchoring in robotics has been used to refer the creation of such symbol-percept connections that specifically refer t o the same physical objects. To date, anchoring systems have proven useful for systems using planners, multi-modal sensing modalities. Most of these applications have focussed on the autonomous robot with little interaction with human users, resulting in arbitrary labels for symbols.
A recent trend in robotics is moving towards the notion of symbiotic robotic systems which consists of a robot, human and (smart) environment cooperating together in performing different tasks. In such systems a richer symbolic representation both to facilitate the interaction with humans and to store properties about objects that can come from different sensing devices. In this project we connect a real knowledge representation system to an anchoring module which is able to provide extended reasoning about natural objects. Using the tools of the anchoring module and the KB a deeper interaction between human-robot or robot-robot can be obtained.