QUALITY CONTROL IN CIVIL CONSTRUCTION:

STATISTICALANALYSIS OF A POINT CLOUD FOR STRUCTURAL AS-BUILT BIM

Authors

  • Rubens Lage Lopes
  • Brismark Goes Rocha
  • Braulio Soares UERN

Abstract

Traditional quality control techniques on construction sites are time-consuming and challenging, involving multiple inspections and communication among stakeholders. Early detection of flaws is essential to avoid waste. Building Information Modelling (BIM) is revolutionizing civil construction, especially with point cloud technology, which enables real-time and more efficient quality control. In this context, the present study analysed the construction accuracy of different building elements, including columns, walls, floors, and beams, using detailed measurements and statistical tests. For the columns, most measurements showed narrow confidence intervals, except for one face, suggesting a possible systematic construction error. Student’s t-tests revealed significant differences in several face mean comparisons. For the walls, the confidence intervals were narrow, with a balanced distribution between positive values (deformations protruding outward from the planned surfaces) and negative values (deformations intruding inward from the planned surfaces), although the tests indicated significant differences in most mean comparisons. On the floors, a trend toward preferentially positive construction was observed, with some narrow confidence intervals, aligned faces, and significant differences among many means. For the beams, the limited sample size hindered definitive conclusions, but the tests indicated significant differences in the means of some faces. The study therefore identified trends of preferentially positive construction in the columns and floors, with significant variations in measurements, while the walls showed greater balance. The statistical analyses highlighted the importance of rigorous monitoring to ensure structural construction accuracy and to correct possible systematic erros.

Downloads

Download data is not yet available.

Published

2025-09-02

Issue

Section

Artigos