Geoscience Reference
In-Depth Information
specificities of indoor spaces. Therefore, several adjustments and improvements
to the original algorithm are proposed which will be implemented in future work,
in an effort to improve the overall user experience during navigation in indoor
environments.
Keywords
Indoor • Navigation • Algorithms • Wayinding • Cognition
1 Introduction and Problem Statement
Over the last decade, indoor spaces have become more and more prevalent as
research topic within geospatial research environments (Worboys
2011
). Past
developments in the modelling and analysis of three-dimensional environments
have already given us a better structural understanding of the use and possibilities
of indoor environments (Becker et al.
2013
; Boguslawski et al.
2011
). These evo-
lutions combined with the rapid progress in spatial information services and com-
puting technology (Gartner et al.
2009
) have put three-dimensional modelling and
analyses more and more in the spotlight. Also, given the fact that as human beings
we spend most of our time indoors (Jenkins et al.
1992
), indoor environments have
become an indispensable part of current geospatial research.
Within indoor research, applications that support navigation and wayfinding are
of major interest. A recent boost in technological advancements for tracking peo-
ple in indoor environments has led to increasing possibilities for the development
of indoor navigational models (Mautz et al.
2010
). Alternatively, several research-
ers have developed a wide variety of indoor navigational models ranging from
abstract space models (Becker et al.
2009
) and 3D models (Coors
2003
; Li and He
2008
) to pure network models (Jensen et al.
2009
; Karas et al.
2006
; Lee
2004
).
While these models might be useful in specific situations, a general framework
for indoor navigation modelling has still to reach full maturity (Nagel et al.
2010
).
Far more recent is the commercial interest with public data gathering for naviga-
tion support in several indoor buildings (e.g. Google Maps Indoor), which demon-
strates the current importance of this application field.
While a considerate amount of work is oriented to the abstract modelling
and technological aspect of navigation, the algorithmic development to support
navigation in indoor built environments has so far been left mostly untouched.
Appropriate and accurate algorithmic support is nonetheless a necessary compo-
nent for a successful wayfinding experience. In outdoor research, a wide variety
of different algorithms exist, initially originating from shortest path algorithms,
studied for over 50 years in mathematical sciences (Cherkassky et al.
1996
). Many
of them are based on the famous Dijkstra shortest path algorithm (Dijkstra
1959
)
with gradually more and more adaptations and extensions for better performance
in terms of speed, storage and calculation flexibility (Zhan and Noon
1998
).
Over time, alternative algorithms were proposed adding a more cognitive notion
to the calculated paths and as such adhering to the natural wayfinding behaviour
