We are pleased to share our founder’s presentation, “Advanced Reconstruction: Creating the As-Built Model,” originally delivered during the December online session of the Metrology & Reverse Engineering Technology Development Programme and now available for download. This release coincides with our participation in the on-site workshop for Fusion Innovation in Barcelona, where the discussion around high-fidelity digital models and qualification workflows continues to gain momentum.
The presentation focuses on the concept of the As-Built model, defined as the digital representation of a component’s actual manufactured geometry, rather than its nominal design intent. By leveraging surveyed data such as point clouds and acquired profiles, As-Built models provide a critical bridge between metrology, design, and simulation—particularly in non-conformal or high-precision manufacturing environments.
A central theme of the work is the advanced reconstruction workflow, beginning with the preprocessing of surveyed data. This includes noise filtering through wavelength reduction and cloud decimation, followed by least-squares fitting, unit consistency checks, and robust alignment strategies based on bounding boxes, least-squares methods, and Hausdorff distance metrics. These steps ensure that measured data can be reliably compared and integrated with nominal CAD geometry.
The core reconstruction process relies on space-warping techniques based on Radial Basis Functions (RBF). Originally developed for surface reconstruction, RBFs are employed here as a meshless morphing field capable of handling very large datasets efficiently. Once trained, the warping field can be reused to morph CAE meshes, CAD brep models, and surface representations, enabling consistent updates across multiple digital domains. This approach supports fast evaluation on both CPU and GPU architectures, making it suitable for industrial-scale applications.
The presentation further demonstrates how reconstructed As-Built models can be used to update CAE, CAD, and surface models, creating what is referred to as a Digital Shadow of the component. These updated models play a key role in qualification, structural analysis, assembly verification, and maintenance planning, and can also support immersive applications such as VR and XR-based shape interrogation.
Several industrial and fusion-related case studies are presented, including automotive body clearance control, PCB deformation and thermal fatigue prediction, TF coil housing reconstruction, and vacuum vessel thermal shield validation. Together, these examples illustrate how advanced reconstruction techniques enable accurate prediction of real-world behavior, improve design refinement, and support complex assembly and lifecycle workflows.
This presentation offers a comprehensive technical overview for engineers and researchers working in metrology, reverse engineering, CAD/CAE integration, and digital twin technologies, particularly in high-precision and fusion-related applications.