How data quality affects model quality in scan-to-BIM: A case study of MEP scenes

Abstract

In scan-to-BIM process, the quality of scan data (i.e. data quality) greatly affects the quality of the generated BIM model (i.e. model quality). However, there has been little research that investigates the relationship between data quality and model quality in a quantitative manner. Hence, this study aims to understand the relationship between data quality and model quality using a case study on mechanical, electrical and plumbing (MEP) scenes. Two MEP scenes were scanned with different scanning settings (angular resolutions and scanning locations), and a total of 20 point cloud datasets were obtained. Then, each point cloud dataset was converted into a BIM model by a BIM modeler. Afterwards, the data quality and model quality were measured for each MEP member. The data quality was measured using Degree of Completeness (DOC) and Point Density (PD), and model quality was measured using recognizability, radius error (for pipe), width error (for duct), and length error. Using these measures, the relationship between data quality and model quality was examined regarding significance and correlation. The analysis results show that DOC and PD play different roles in influencing recognizability and various modeling errors. Given a certain model quality requirement, the required data quality can be inferred from the analysis results, which can help achieve a better planned scan-to-BIM process.