Project Abstract

Indoor navigation remains a significant challenge due to the signal limitations of Global Navigation Satellite Systems (GNSS) and the high costs of infrastructure-dependent technologies such as Bluetooth beacons and WiFi fingerprinting. Existing approaches typically rely on pre-deployed physical infrastructure or image-based systems that are sensitive to environmental change, creating a need for self-contained, infrastructure-free alternatives. This study develops and evaluates CAMSNavi, a self-contained augmented reality (AR) navigation system that uses point cloud-based Simultaneous Localization and Mapping (SLAM) and consumer-grade LiDAR to enable real-time indoor navigation without external dependencies. The system uses the LiDAR sensor of an M1 iPad Pro to generate a digital twin of the Computational and Mathematical Sciences (CAMS) building; the resulting 3D model is integrated into Unity via Vuforia Area Targets to enable on-device localization and AR navigation. Preliminary testing across five routes spanning three wings demonstrated marker-less localization with an average time-to-localize of approximately 3 seconds at fixed positions, correct venue selection, and stable end-to-end AR navigation with 3D directional arrows anchored to the physical environment. These findings suggest that consumer-grade LiDAR can support reliable, infrastructure-free AR navigation and provide a foundation for further evaluation of localization performance and multi-area navigation.