What is necessary for robots to coexist with human beings? In order to do so, we suppose, robots must be moralagents. To be a moral agent is to bear its own responsibility which others cannot take for it. We will argue that such anirreplaceability consists in its having an inner world — one which others cannot directly experience, just as pleasure and pain.And personality of a moral agent, which is to be irreducible to a mere difference of traits or features of individuals, is firmlyrooted in such an inner world.We will support our theses by referring to our experiment in which humans and robots interact with each other doing acoordination task. This experiment will provide an empirical analysis of the human-robot relationship with regard to learningmechanism, moral judgement, and the ascription of the inner world.