As computer systems become ubiquitously embedded in our environment, computer applications must be increasingly aware of user context. In order for these systems to interact with users in a meaningful and unobtrusive way, such as delivering important reminders at an appropriate time, their interfaces must be contextually-aware. This vision of future computer systems and the insight that the implementation of contextually-aware systems requires contextually-aware analysis and development tools has motivated the two primary contributions of this work. First, a Context-Aware Experience Sampling Tool has been designed, implemented, and tested. Second, this tool has been used to develop an algorithm that can detect transitions between human activities in office-like environments from planar accelerometer and heart rate data. The Context-Aware Experience Sampling Tool (CAES) is a program for Microsoft Pocket PC devices capable of gathering qualitative data, in the form of an electronic questionnaire, and quantitative data, in the form of sensor readings, from subjects.

Abstract—Providing accurate and opportune information on people’s activities and behaviors is one of the most important tasks in pervasive computing. Innumerable applications can be visualized, for instance, in medical, security, entertainment, and tactical scenarios. Despite human activity recognition (HAR) being an active field for more than a decade, there are still key aspects that, if addressed, would constitute a significant turn in the way people interact with mobile devices. This paper surveys the state of the art in HAR based on wearable sensors. A general architecture is first presented along with a description of the main components of any HAR system. We also propose a twolevel taxonomy in accordance to the learning approach (either supervised or semi-supervised) and the response time (either offline or online). Then, the principal issues and challenges are discussed, as well as the main solutions to each one of them. Twenty eight systems are qualitatively evaluated in terms of recognition performance, energy consumption, obtrusiveness, and flexibility, among others. Finally, we present some open problems and ideas that, due to their high relevance, should be addressed in future research. Index Terms—Human-centric sensing; machine learning; mobile applications; context awareness. I.