Intelligenza artificiale, robot e rappresentazione della conoscenza. Il Laboratorio di simulazione del comportamento e robotica educativa "Luciano Gallino"

Studies of artificial intelligence in the 80s’ focused on knowledge representation formalisms which provided the capabilities of the first expert systems. The logic of such formalisms (at the basis of the modern ones) should become part of the “toolbox” of the social scientists, as a fundamental instrument to interact with the new social robots and to effectively participate in the development project of this “embodied” branch of AI. From this point of view, the article illustrates an expert system programmed to recognise the Saints depicted in religious iconographies; next it exposes some experiences of behaviour simulation through agentbased models, and finally reports on interaction protocol tests with robots conducted in Gallino Laboratory of the University of Turin.