|In previous work the authors have proposed and demonstrated a method designed to form and label location clusters from spatially informative data, such as received WiFi signals. Using data streams from an unknown environment which are not traditionally used for navigation (such as user digital footprint), our algorithm assigns labels or places to characteristic sets of signal observations, the placeprints. These labels have an extent within an unobservable physical space. In this paper, we work towards the optimal measurement of movements between rather smaller areas in closed and unobservable indoor spaces. We investigate this relationship motivated by the desire to understand the task management behaviour of clinicians working in a hospital setting. Hence, we form informatively spatial clusters from observed WiFi signals relating to unobservable sub-ward locations. These are highly informative for the purposes of evaluating doctor performance within clinical settings. Note that here, activities take place in very small spaces to which characteristic digital data streams relate and semantic labels could be assigned; for example, the office, nurse station or the store cupboard. In summary, we present a method for labelling a person’s location at a spatial resolution suitable for understanding their activities in a clinical workspace; using observations of received signal strength from signal transmitters and a set of semantic labels relating to workstations in the clinical environment. Additionally, we discuss the effect of physical alterations on the workplace layout, for the purposes of better estimating and differentiating the positioning of doctors or nurses in relation to workstations of relevance; thus contributing to the study of optimal ward arrangement within clinical settings.
Proceedings of the ION 2017 Pacific PNT Meeting
May 1 - 4, 2017
Marriott Waikiki Beach Resort & Spa
|232 - 238
|Cite this article:
Pinchin, James, Perez, Iker, Sharples, Sarah, "Keynote: The Geometry of Place and Activity," Proceedings of the ION 2017 Pacific PNT Meeting, Honolulu, Hawaii, May 2017, pp. 232-238.
ION Members/Non-Members: 1 Download Credit