PARAMETRIC GENERATION IN HBIM WORKFLOWS FOR SLAM-BASED DATA: DISCUSSING EXPECTATIONS ON SUITABILITY AND ACCURACY

In the wide framework of Scan-to-BIM 3D modelling procedures, the complexity of the architectural heritage and its components is evidencing a gap of best practices and specification in the HBIM-modelling and thus it is requiring important considerations about the modelling strategies and protocols between the requested level of detail (LOD), the expected accuracy and above all the actual user-oriented requirements. Several works are largely focusing on developing modelling workflows for traditional static LiDAR scanning sensors. The amount of starting data is as much an asset when a bottleneck. In fact, the chance to benefit from on-site faster data acquisition is needful at times, and data collection procedures are directing toward rapid mapping 3D approaches, evolving from traditional static scanning toward MMS (Mobile Mapping Systems) based on SLAM technology (Simultaneous Localization and Mapping) implemented in portable devices. The potential derived from these solutions can contribute to increasing massive cost-effective documentation, and also in view of BIM-HBIM models generation, and this needs further researches. At the same time, the descriptive capabilities of this class of portable scanners do not reach the precision of static solutions. Many considerations towards the analysis of geometry, grade of generation (GOG) and details can be thus conducted, especially under the point of view of time-cost balance. This paper aims to presents the entire modelling workflow (from raw data acquisition up to parametric modelling) based on a first comparison between TLS (Faro Focus 3D) and hand-held scanner Zeb Revo (by GeoSLAM), focusing on the Bramante’s Canonica Court in the Basilica of Sant’Ambrogio. First, the two raw data are compared, considering geometric features (data density, precision, possibility to detect edges, details and accurate curvature). Then, some wellestablished modelling procedures developed for TLS data, as triangulation mesh and NURBS generation, are applied to MMS point cloud to assess their suitability, identifying limits and potential. Different elements belonging to the architectural structure hierarchy are considered in a multi-scale perspective: the vaulted system of the porch, the columns and the arches of the porch with their different architectural elements.