Digital Twin-Ready Representation of Enterprise Spaces through Integrated 3D Reconstruction Pipeline
Authors/Creators
Description
Digital Twins are increasingly adopted in enterprise environments to support intelligent monitoring, asset management, human-robot collaboration, and data-driven decision making. A fundamental prerequisite for these applications is the availability of an accurate and semantically meaningful threedimensional representation of the physical environment. Although existing reconstruction approaches can generate detailed geometric models, they typically focus either on metric reconstruction or semantic understanding, providing fragmented solutions that are difficult to integrate into Digital Twin platforms. This paper presents an integrated and semi-automated pipeline for generating Digital Twin-ready indoor models by combining image-based photogrammetry, handheld SLAM (Simultaneous Localization And Mapping) acquisition, geometric model fusion, semantic enrichment, and textured mesh reconstruction. The proposed methodology exploits the complementary strengths of photogrammetry and SLAM to produce a unified metricscale point cloud. Experimental results demonstrate that the framework successfully reconstructs metrically consistent and semantically enriched three-dimensional models suitable for DTready space integration with robots, avatars, IoT devices, and real-time services.