A Study on the Application of AI and XR Convergence Models Based on Heritage Materiality

This study investigates how cultural heritage Digital Twins (DT), integrating Artificial Intelligence (AI) and Extended Reality (XR), enable preservation tailored to the materiality and physical constraints of individual sites. By operationalizing the London Charter's preservation ethics into a framework of nine observable variables, this research addresses complex challenges ranging from ongoing material decay to severe physical absence. A comparative analysis of Digital Dunhuang (2D surface), Notre-Dame Cathedral (3D structure), and Rome Reborn (4D urban space) reveals materiality-driven integration patterns across technological architecture, interdisciplinary governance, and heritage ethics.<br/> Findings indicate that technology integration must align with materiality: preventive conservation via optical diagnosis and generative infill for 2D surfaces; structural resilience through automated classification and spatial optimization for 3D structures; and cognitive reconstruction using parametric procedural generation for 4D urban spaces. Successful DT implementation demands materiality-aware calibration over generic adoption, highlighting the need to resolve challenges in economic sustainability, semantic interoperability, and intellectual transparency. This study defines a tiered approach distinguishing essential variables, such as Method Fitness, from aspirational ones such as Economic Sustainability, to address these limitations. This framework establishes an adaptive baseline, enabling resource-constrained institutions to deploy computational preservation while maintaining strict cultural authenticity.