Designing through exploration: using observational methods in ubiquitous technology research

The evaluation and study of ubiquitous computing is an essential part of learning from design success and failure. Yet understanding technologies which are, by their very nature, embedded in the world presents new challenges for evaluation. In particular, there are limitations in how approaches based on experimental lab studies can reveal about complex contexts outside the laboratory. Observational methods offer an alternative approach which allows broader lessons to be drawn from prototypes’ use insitu. Part of this is moving from the ‘evaluation’ the use of systems to ‘exploring’ their use. We discuss three different studies where we have used observational methods to explore the use of ubicomp. In the ‘tourism study’ we studied a setting before the introduction of technology to learn what sorts of ubiquitous technologies would be appropriate for tourists. In the ‘lighthouse co-visiting study’ we conducted a field experiment of a prototype museum co-visiting system. Lastly, in the ‘hummingbird study' we looked at the use of a location based technology by ski instructors as part of nonstaged activities. Each of these studies show how observational methods and experimental ubicomp technologies can be combined. From these example we discuss the different stages of conducting an observational study, and give pointers to how ubicomp designers themselves could use observational methods in their work.