Challenges in the Design and Development of Context-Aware Applications

Abstract

Context-awareness plays an important role in pervasive computing as adaptations of applications to context changes (changes in computing environment and in user activities/tasks) help to achieve the goal of computing services available everywhere and at any time. There is a growing body of research on context-aware applications that are adaptable and capable of acting autonomously on behalf of users. However, there are still many open research issues that challenge the pervasive computing community. In this talk I discuss several of these research challenges. First, I outline the state of the art in context information modelling, management and reasoning as well as possible future research directions in this area. This is followed by a discussion of context information management that allows development of fault-tolerant and autonomic context-aware applications. As one of the challenges inhibiting the development of context-aware applications is their complexity, I discuss software engineering approaches that ease the task of developing such applications. Context-aware applications adapt to context changes using context information. However this context information may be imprecise or erroneous and therefore can lead to incorrect adaptation decisions creating usability problems and affecting acceptance of context-aware applications. This creates a need for some balance between autonomy of context-aware applications and the user control of the applications. I describe some early approaches my team is working on to tackle this problem. Finally, I discuss research issues related to privacy of context information and how context can be used to enhance security mechanisms within pervasive computing environments.