Error Assessment of Point Cloud and BIM Models to Actual Works

Abstract

3D Laser Scanning technology is being widely applied in the construction industry 4.0 for promoting the digitalizing and BIM modeling procedure, also called Scan-to-BIM. Scan-to-BIM technology helps to capture the continuously existing condition of the project and construct a BIM model in just a short period of time, also, update and store the data of the project throughout construction stages. Scan-to-BIM technology could, however, entail many risks if the users cannot control the errors of the point cloud model and BIM model. Although point cloud collected by 3D laser scanners has high accuracy, there are mistakes during Scan-to-BIM process, which cause errors. Therefore, the 3D BIM model based on the point cloud contains errors of dimension, coordinate, position, and these errors could cause difficulties for construction project management engineers and make the BIM model meaningless. This paper presents the causes of the errors, also analyzes and evaluates the errors of Scan-to-BIM process by the methodology of comparing point cloud and BIM models with physical objects to control the errors and support the next step of using the VR, AR, and MR. The data used in the paper is obtained from a real project which applies Scan-to-BIM technology to verify the theory.