Wayfinding is a fundamental aspect of human spatial behavior, which involves the coordination of perception, cognition, and movement to navigate through environments. While considerable research has advanced our understanding of navigation behavior, especially in controlled or virtual environments, there are still open questions about how people navigate in complex, real-world outdoor settings. Existing wayfinding models, such as the four-step framework by Downs and Stea [1977], provide a useful foundation, yet they are largely theoretical and rarely validated with empirical behavioral data. This dissertation addresses this gap by aligning theoretical models with empirical observations of wayfinding behavior in outdoor real-world environments. Using a multimodal approach, it integrates mobile eye-tracking, inertial motion sensing, and self-reported user data across a series of real-world experiments, including a large-scale study with over 300 navigation trials. These data were analyzed using a combination of machine learning and deep learning methods. The research employs Human Activity Recognition (HAR) concepts and methodologies to understand wayfinding behaviors and investigates how spatial familiarity, environmental complexity, and individual traits influence navigation. It combines theory-driven analysis, grounded in existing cognitive frameworks, with data-driven exploration through unsupervised learning, which provides complementary perspectives on the structure of navigation behavior. The findings contribute both conceptually and methodologically. They provide an empirical basis for established wayfinding models in naturalistic contexts and also reveal behavioral dynamics that offer new insights into wayfinding and open new paths for research that could lead to the refinement of existing theories or the development of new ones. From a practical perspective, the findings can inform the design of adaptive, user-aware navigation systems. In a broader sense, the work provides a methodological basis for future research at the intersection of cognitive science, GIScience, and artificial intelligence in the study of human mobility.