In my master thesis, one important attribute is the temporal nearness of concepts in the domain, such as moving persons and equipment. Special concerns for the case is that the geography is indoors in potentially complex buildings, such as large hospitals. The geography is considered to be known to the actors – assuming that the persons know where they are, where they are headed to and how to get there. This may argue that a floor map depicting where you are, your target and fastest route there is unnecessary – and I agree. However what may be of interest is to get information on how far away you are from your interest points and how far your interest points are from each other. For instance if you are 15 minutes away from your office, you have planned to meet your colleague there at 15 minutes, but your colleague is 20 minutes away. Provided by this information is 5 minutes that you can finish what you are doing, walk to your office, and still meet your colleague in time! (pretty lame example I know – working on my exemplifying skills:). Anyhow, the driving information is thus the temporal nearness. This is highly dependant on the geography in the context and it’s navigational abilities. Building generally has rooms with doors and corridors which in turns comprise and constrains the ability to move in that space (i.e. walking). My attempt at finding the temporal nearness comprise a spatial model of such a geography, a data structure for navigational abilities, implementation of this model in a spatial database (PostGIS) and calculation of the temporal nearness based on this. This blog post tries to explain just this. I plan to post a separate post on how to manage the data for this model - so watch out:)

Spatial model and data structure for indoor temporal nearness

Figure1: Spatial model and data structure