Is your city easy to navigate? Street network properties and their impact on navigation success

Abstract

Wayfinding is an everyday human task when navigating through a city and the routeswhich are taken are always embedded into a street network. Hence, whenever wayfinding studies are conducted, they are influenced by the underlying network. This thesis investigates what street network properties impact the navigation success when using Free Choice Navigation as navigation paradigm.The street network properties characterise the neighbourhood, which surrounds a route from a starting point to a destination. Six different approaches are discussed, on how to define this neighbourhood. The chosen approach contrasts with the prevailing choice of geometric or administrative study areas and the focus is rather on the street network itself. It ensures that all nodes, that could be passed when navigating from A to B under the condition that the final path length is below a certain allowed maximum length, are within the area.After the definition of the neighbourhood area, 34 street network properties are discussed and computed. By using Principal Component Regression, the relation between 30 of those properties and the navigation success rate is investigated. The analysis is conducted for three cities: Vienna, Mexico City and Djibouti City. It is found that the ratio between the allowed detour and the mean segment length within the neighbourhood area impacts the success rate and is positively correlated with it. In addition, the city type of the cities where the routes are located seems to have an impact. The results suggest, that Vienna appears to be more navigable than Mexico City, because the investigated neighbourhoods have a lower mean segment length. Furthermore, Vienna seems more navigable than Djibouti City, because the investigated routes are longer on average.