Evaluation of Preference- and Constraint-Sensitive Path Planning for Assisted Navigation in Indoor Building Environments

Abstract

Independent mobility in people with physical disabilities notably enhances their quality of life by allowing social inclusion and promoting self-esteem while reducing caregiver burden. Despite significant technological advances in the last few decades, people with physical disabilities who rely on wheeled mobility devices continue to face significant challenges when independently navigating inside complex indoor environments. Several technological solutions have been explored to address this issue in previous work. Specifically, vision-based techniques that utilize fiducial markers as visual cues for assisted navigation have received particular attention because of their cost-effectiveness, reconfigurability, and ease of installation. Although such previous work addressed indoor localization for assisted navigation, it did not adequately address the fundamental aspects of indoor path planning, particularly in the context of considering human preferences or physical constraints present in the built environment. This paper addresses these key research gaps by developing fundamental rules for creating an indoor attribute-loaded graph network and a marker network map, developing a generic algorithm for determining optimal paths (instead of shortest paths only) from attribute-loaded networks, and integrating the findings within an interactive user interface (UI). As a precursor to eventual studies with people with mobility impairments, a scenario analysis followed by a preliminary usability study in a real physical environment involving 10 participants without disabilities was conducted to evaluate the navigation interface. Results demonstrate the feasibility of the proposed navigation system and provide design insights for improving the usability and performance of the current user interface.