Lightweight Mobile AR for Crane Safety Education: Dual-Mode 3D Visualization to Enhance Load-Chart Interpretation and Spatial Awareness

Mobile crane operations are among the most hazardous activities in construction, and effective safety planning often hinges on a practitioner’s ability to infer complex three-dimensional relationships from static twodimensional load charts and schematic diagrams. This study reports the development and preliminary evaluation of AR Crane Safety Viewer, a low-resource mobile augmented reality application designed to reduce this interpretive burden in crane safety education. Implemented in Unity 2022.3 LTS using AR Foundation and ARCore, the system adopts a dualmode workflow consisting of a stable 3D viewer for controlled conceptual inspection and a markerless AR mode that enables contextual placement of the crane model on real-world surfaces. The application embeds on-model annotations to externalize key geometric parameters boom angle, boom length, load radius, and lifting height thereby supporting learner understanding of the crane working envelope through interactive exploration rather than memorization. In contrast to high-fidelity simulators that demand specialized hardware and high computational resources, the proposed tool targets mid-range Android devices to facilitate deployment in classrooms and pre-task briefings. The architecture prioritizes consistency of interface and interaction across modes, incorporating drag-based rotation, pinch-to-zoom scaling, and optional auto-rotation for hands-free demonstration, alongside an integrated screenshot workflow to support documentation. Functional testing on a Samsung M33 5G indicates reliable operation of both viewing and AR placement features under typical indoor conditions. Collectively, these results suggest that lightweight interactive 3D visualization combined with contextual AR placement can strengthen prelift awareness and improve comprehension of safety-critical crane terminology. The study contributes an implementable design pattern for accessible AR-enabled learning tools in construction safety training and provides a basis for subsequent controlled studies quantifying learning gains and error reduction in load-chart interpretation.