1 Moving Beyond the Application: Design Challenges For Ubiquitous Computing Mark Stringer, Geraldine Fitzpatrick, John Halloran, Eva Hornecker INTERACT Lab, Dept. of Informatics, University of Sussex Falmer, Brighton BN1 9QH, UK M.Stringer@ sussex.ac.uk, geraldine@sussex.ac.uk, johnhall@sussex.ac.uk, eva@ehornecker.de ABSTRACT Weiser predicted a world in which computing moves ‘beyond the desktop’ and is increasingly integrated into the everyday environment and embedded in everyday practices and activities. We think that, as this becomes a reality, the concept of ‘application’ needs to be re-thought and broadened, motivating this by a look at key successes in ubiquitous computing. We discuss some of the challenges to design and evaluation methodologies that a broadening of the definition of ‘application’ might produce. INTRODUCTION Although ubiquitous computing by its nature moves beyond the desktop [6], research in ubiquitous computing still tends to be restricted to a desktop-centric view of what constitutes an application. We believe that this view (a) prevents the community from recognising how successful ubiquitous computing comes about; and (b) means that key issues for design, development and evaluation are not being sufficiently acknowledged. To make this argument we draw on examples of successful ‘ambient’ ubicomp, as well as our own experiences of engaging with ubiquitous computing applications through projects currently being conducted in the Interact Lab at the University of Sussex. The Interact Lab is primarily concerned with the user- centred iterative design, deployment and evaluation of innovative technologies. Our aim is to design technologies that fit into and enhance quality of life for the people they are designed with and for. In particular we are concerned with embedding applications (’beyond the desktop’) in everyday contexts of use, be they home or office or outdoors, leisure or work. We are not just interested in what is technically feasible but, more importantly, what is useful or desirable, enhancing people’s lives and interactions. We consider it critical to conduct research in real-world, ‘in the wild’ settings with existing homes and public spaces etc, rather than purpose built settings or labs, to explore the very practical issues of situating ubicomp technologies into everyday environments. We also believe that involving people in helping to define what they want to live with is critical. This implies designing applications for the messy real world in the messy real world with the collaboration of its inhabitants. Underpinning all of the projects that we discuss here are the common and fundamental research questions: What is ubiquitous computing for? What are key examples of success? How do we go about finding a genuinely welcome place for ambient, pervasive and ubiquitous computing applications? What methodologies should we use to discover them? How can we engage all sectors of society in the design of applications fundamentally different from currently available applications? LEARNING FROM AMBIENT COMPUTING Currently existing ‘ambient’ computing successes in the real world tell us a lot about what ‘applications’ are - once we move beyond the desktop. There are already many examples of computing disappearing into the environment in the way imagined by Weiser in his seminal paper that kick-started the field of ubiquitous computing [6]. One example is supermarket shopping, where an enormous amount of technology is arrayed in order to support the consumer’s desire to select her own vegetables from an attractively displayed shelf. There is evidence that this interaction is in fact a very primal – the desire to pick fruit from a tree (or shelf) is optimised in the visual system of primates [3]. Another example is the interaction involved in driving a car. Here technology is employed not only in the car to ensure smooth running of the engine and safe braking, even on a loose surface, but also in the traffic lights which control the movement of the car down the road and in the command and control systems which ensure that there is fuel in the pumps at the petrol station. This example highlights another very important feature of many successful disappearing computing applications. The computers that work in them do not form a ‘joined-up’ system. Data does not flow through them end-to-end but is mediated by mothers on the school run, delivery trucks, shop assistants etc. In both cases, although they may have changed in many small ways, driving and shopping are still recognizable as the same activities that they were 50 years ago when no computing technology was involved.