The increasing reach of the Internet of Things (IoT) is leading to a world rich in sensors [3] that can be used to support physical analytics -- analogous to web analytics but targeted at user interactions with physical devices in the real-world (e.g. [2]). In contrast to web analytics, physical analytics systems typically only provide data relating to sensors and objects without consideration of individual users. This is mainly a consequence of an inability to track individual mobile user interactions across multiple physical objects (or across sessions of interaction with a single object) using, for example, an analogue of a web cookie. Indeed, such a "physical analytics cookie" could raise significant privacy concerns.

However, in many cases a more "human-centric" approach to analytics would enable us to provide new and interesting insights into interactions between mobile users and the physical world [1]. In our work we endeavour to leverage synthetic user traces of human mobility, and data from real IoT systems, to provide such insights.