XR (extended reality) environments have driven professionals in the construction industry to adopt advanced digital surveying and 3D modelling techniques, improving project quality and site management through enhanced visualization, accuracy, and precision. Photogrammetry and laser scanning have been crucial in the scan-to-XR process, enabling the development of digital twins (DT) throughout the construction lifecycle. However, converting survey data into usable models for immersive experiences requires expertise in digital representation and software development. Digital representation transforms raw data into functional models, yet challenges remain. Survey outputs typically feature high polygon counts for detail, which can overload XR applications, causing slowdowns and reducing fluidity, especially on devices with limited resources. Additionally, high-resolution textures further strain computational power and memory. Optimising these textures is key to balancing visual quality and performance. XR platforms like Unity and Unreal Engine demand specific rendering standards, and non-optimised models can fail to meet these, requiring further adjustments. This study aims to develop a pipeline for creating high-fidelity DTs of an ongoing construction site. Data is collected through photogrammetry and laser scanning, followed by model optimisation for XR applications. The optimised model is integrated into real-time platforms for interactive use, producing both VR and web-XR models. This pipeline will aid in construction-site management, safety inspections, and communication between stakeholders, contributing to DT technology and to the efficiency of the construction industry.
Representation Advances And CHallenges association