Until early 2000’s, actigraphy has been used mainly as a method to assess insomnia, circadian rhythm disorders or excessive sleepiness. Since then, progresses made on sensor’s dimension and consumption, as well as data analysis algorithms, increased the number of application fields. Human Activity Recognition (HAR) is an active research area that classifies user activities from inertial data; it finds application in medicine, athletics, lifestyle monitoring and human/computer interaction. All modern consumer electronics, such as smartphone, smartwatch and personal activity trackers represent a good platform for HAR application, as they use most advanced technology and IMUs. Nevertheless, devices available nowadays on the market have some important limitations.