Modelling navigational knowledge by route graphs

Abstract

Navigation has always been an interdisciplinary topic of research, because mobile agents of different types are inevitably faced with similar navigational problems. Therefore, human navigation can readily be compared to navigation in other biological organisms or in artificial mobile agents like autonomous robots. One such navigational strategy, route-based navigation, in which an agent moves from one location to another by following a particular route, is the focus of this paper. Drawing on the research from cognitive psychology and linguistics, biology, and robotics, we present a simple, abstract formalism to express the key concepts of route-based navigation in a common scientific language. Starting with the distinction of places and route segments, we develop the notion of a route graph, which can serve as the basis for complex navigational knowledge. Implications and constraints of the model are discussed along the way, together with examples of different instantiations of parts of the model in different mobile agents. By providing this common conceptual framework, we hope to advance the interdisciplinary discussion of spatial navigation.