In recent years, the availability of an ever increasing number of heterogeneous smart devices has changed everyday life as only few other technologies has been able to do in the past. The diffusion of the Internet of Things (IoT), as well as the spreading of user's personal devices, such as smartphones, tablets, and, more recently, smartwatches or other wristband devices, have enabled new people-centric sensing paradigms in which raw data captured by on board sensors can be analyzed to infer higher-level knowledge. Participatory sensing [1] and mobile crowdsensing [2], for instance, are two common strategies in which the devices owned by a community allow to understand, or even predict, relevant phenomena [3].