The aim of this paper is to integrate perception, action and symbolic knowledge to allow an autonomous robot to operate in unstructured environments and to interact with non‐expert users.
To achieve such goals, a cognitive robot architecture is proposed based on the integration between subsymbolic and linguistic computations through the introduction of an intermediate level of representation based on conceptual spaces.
The architecture has been tested in the CiceRobot project on tasks related to guided tours in the Archaeological Museum of Agrigento. Experimental results show that robot cognitive behaviors allow one to achieve a full functional robotic museum guide. In particular, through the interaction with visitors the robot is able to customize the tour depending on their preferences.
The paper presents a significant case study because it involves perception, planning and human‐robot interaction. The proposed architecture addresses the capacities which are generally addressed by an intelligent agent: the capability of representing itself and the external world, of imagining possible evolutions of the world, of paying attention to the relevant events, of planning and evaluating situations and actions.
