SAR Remote Collaboration Platform Supporting ATI for Training Tasks

Abstract Global events, including pandemics and geopolitical conflicts, have catalyzed a fundamental reevaluation of remote collaboration paradigms across training, technical assistance, maintenance, and educational domains. While conventional video conferencing technologies have achieved widespread adoption, significant potential remains for advancing augmented reality (AR)-based remote collaboration methodologies. Though hand-drawn AR annotations (AAS) remain prevalent for visual instruction delivery, their inherent informality often proves insufficient for conveying precise operational directives in industrial contexts. Consequently, multimodal verbal communication emerges as a critical compensatory mechanism to resolve ambiguities and mitigate linguistic barriers. This research introduces a novel Spatial AR (SAR) remote collaboration platform designed to bridge this communication gap through an Adaptive Transformation of Instructions (ATI) methodology. The proposed system enables dynamic conversion of freehand annotations into standardized symbolic representations, thereby enhancing instructional clarity and fostering universal comprehension. A rigorous formative user study was conducted to evaluate the platform's efficacy across multiple dimensions: task completion time, annotation complexity, subjective collaborative experience, system usability (via standardized rankings), and user preference metrics. Results demonstrate statistically significant improvements in user experience when ATI-based cues are integrated into remote collaborative training scenarios. Beyond practical implications, this work provides actionable insights for future investigations into AR-mediated remote collaboration within technical training environments.