Ubicomplexity: Applying Useful Concepts from Complex Adaptive Systems to Ubiquitous Computing

The realization of Mark Weiser’s vision of ubiquitous computing, of “machines that fit the human environment” (1991), is predicated on large numbers of massively interconnected devices that can move fluidly between the edges and the focus of our perception. The number of potential interactions between these devices, their environment, and the humans within it makes it difficult to design both applications and systems-level frameworks for their development. These sorts of multi-agent, massively interconnected systems are the subject of an evolving area of interdisciplinary study, known as complex adaptive systems theory. Drawing researchers from mathematics, biology, physics, sociology, and computer science, CAS represents a sprawling and disjointed body of knowledge, ranging from rigorous thermodynamic models to management textbooks on organizational theory. While the stated aims of CAS and ubicomp appear to be congruent, it is difficult at first glance to identify elements of CAS research that might be applicable to the latter. In this paper, we outline our findings from a special topics course in ubiquitous computing that was conducted in the spring of 2005. Our goals for the course were threefold. First, we set out to survey a broad spectrum of literature in ubicomp, discuss its theory and application, and develop a competency in current areas of ubicomp research. Second, we sought to develop a bibliography of study in complex adaptive systems, and relate concepts from our readings to ubiquitous computing. Third, we wished to create a model for ubicomp applications informed by the paradigms of complexity, and create a final project based on this model. Unfortunately, our goal of a final project within a one-semester timeframe proved to be too ambitious, given our uncertain theoretical grounding. Instead, we are recording our findings as a foundation for future work. While our model for an “ubicomplex” system remains rough-shod, we believe our work shows that concepts learned from complex systems theory can suggest valuable design lessons and directions of study for ubiquitous computing researchers.