Edinburgh Napier University

  This paper is concerned with understanding the needs of Companion owners (the people formerly known as ‘users’). The problem with developing technologies such as companions is in knowing what the requirements are. People cannot really be expected to express their needs for companions before the technology that will drive the idea has been invented. Yet we know that the technology that will provide the sort of personalised, persistent interactions that characterise companions is coming, the question we are interested in is how requirements for companions can be generated. We are concerned with the whole interaction design, not just the speech recognition and language understanding, the gestures, or the inferences the companion can make. We are interested in how the companion will learn, or be instructed, so that the interaction can evolve and develop with the individual. We are concerned with the whole interaction experience and with how the different components fit together. In this chapter we present a conceptualisation of the companion idea and then illustrate how the development of personas and scenarios in the context of companion technologies can help us generate the requirements for these technologies. Finally we comment on some of today’s technologies that already starting to demonstrate the characteristics of companions