Non-Destructive Geometric Quality Assessment of Robotically Constructed Masonry Using Terrestrial Laser Scanning and Mobile-Device Photogrammetry

The increasing automation of masonry construction creates a growing need for sensor-based methods for geometric quality control. This study investigates the suitability of two point-cloud-based acquisition systems for the non-destructive assessment of robotically constructed calcium silicate masonry. A Trimble X9 terrestrial laser scanner was compared with a photogrammetric workflow using a fourth-generation Apple iPad Pro. The resulting point clouds were characterized globally and locally in terms of statistical outliers, point density, and spatial homogeneity. Relative positional deviations between adjacent masonry units were then evaluated in three spatial directions using median-based and plane-based approaches. Terrestrial laser scanning exhibited higher point density indicators, a lower proportion of statistical outliers, and lower surface-point dispersion. Both systems indicated the same direction of deviation for the vertical height and short-end-face cases, whereas the long masonry-unit faces were practically flush in the plane-based evaluation. Inter-system differences ranged from 0.86 to 2.74 mm for the median-based evaluation and from 0.37 to 5.02 mm for the plane-based evaluation, demonstrating the influence of surface orientation and estimator choice. Without an independent geometric reference, no conclusions can be drawn regarding absolute measurement accuracy or which system provides the more accurate displacement estimate.