This paper investigates the impact of placement and orientation variations on the quality of sensed data. Different types of human movements are considered, namely, balancing, skipping, leaping; climbing up and down a staircase, and running. For data collection, tri-axis accelerometer sensors are used. As target placements, arms, thighs, knees, ankle, and waist are considered. Likewise, four different orientation angles were considered during deployment, namely, 0, 30, 45, and 85 degrees. The features employed to investigate placement and orientation variations were zero\mean-value crossing rate, correlation coefficients, cross-correlation, and auto-correlation. A particular focus was given to steady slow movements (climbing up and down a staircase) and steady fast movements (running). Remarkably, the fast movements are less affected by placement variations in comparison to the slow movements. Moreover, it will be shown that the effect of orientation variations for all types of movements are insignificant when absolute acceleration instead of the accelerations of individual axes are independently considered. ACM Classification Keywords

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.