Granularity of origins and clustering destinations in indoor wayfinding

A key prerequisite for indoor wayfinding through verbal turn-by-turn route instructions is correctly identifying the origin of the requested route. Here, we show that the origin only needs to be specified down to a granularity that is dependent on the chosen grammar of the instructions when considering the geometric structure of the environment. This research will show that expensive infrastructures enabling high positioning accuracy in indoor environments are unnecessary for wayfinding with route instructions. The relationship between granularity and grammar is identified by a similarity index that reflects the similarity of route instructions between the origins according to their spatial relationship. Moreover, with a chosen grammar some destinations are described through identical route instructions from a given origin. We undertake a clustering analysis of such destinations to demonstrate their similarity in route instructions. This clustering will diagnose a situation in which the wayfinder follows the route instruction to a destination, but they arrive at a different destination within the same cluster.