Navigating through physical environments has evolved over time from using stars and maps to support the wayfinding, to employing Global Positioning Systems and navigation services. Turn-by-turn guidance of navigation services is an effective way to support wayfinding, but it may not align with the way humans naturally navigate. Over-reliance on navigation services can lead to confusion, frustration, and even dangerous situations. Humans use environmental cues to support their navigation decisions and understand their position, orientation, and surroundings. Navigation services prioritize efficient route planning and may not consider factors such as complexity that can impact travel. This discrepancy between navigation services and human navigation highlights the importance of incorporating principles of human wayfinding into navigation systems to enhance the overall wayfinding experience.

This thesis aims to improve navigation services by exploring their adaptive capabilities and addressing the discrepancies between navigation services and human wayfinding. The research focuses on identifying difficult-to-navigate intersections and prominent locations along a route that are important for successful navigation, and developing automated ways to identify them. The thesis also explores adapting instruction giving to the route and its surrounding.

The research included in this thesis analyzed geographic data, developed models and measures that extended existing research, and conducted empirical human subject studies. This work developed models that optimize route search for specific criteria, including traffic and social costs. It also proposes approaches to identifying and simplifying prominent locations along a route that define the relationship between the route and the environment. Results show that people tend to prefer less complex routes with fewer prominent locations. Results also indicate that incorporating route-defining locations in route directions can aid wayfinders in forming useful spatial memory of the environment. Additionally, the studies identified the language used and spatial reasoning mechanisms over direction change as sources of mismatches between navigation instructions and human understanding of a given wayfinding situation, which may provide insights into improving the generation of instructions.