In this report we analyze state-of-the-art algorithms for human monitoring in generic environments and we focus on video monitoring of human activity. The objective is to define the basis upon which the ACTIBIO dataset will be recorded. The first sections are devoted to generic technical aspects of multi-modal monitoring, including modalities that are not foreseen for authentication within ACTIBIO, but that can provide knowledge about deployment of a multi-modal sensor network. Next, we analyze state-of-the-art algorithms for human monitoring, focusing on methods for activity and behaviour understanding at different levels of abstraction. Finally, we present a recording protocol comprising technical issues derived from the knowledge acquired during the state-of-the-art review and including also a script responding to the specifications of ACTIBIO use cases and scenarios.