Enhancement of precision in electromagnetic radar exterior wall inspection through model-based analysis of measurement uncertainty

Abstract A model-based approach using tile-based models was developed to estimate the depth and size of voids within exterior walls. However, during field applications with portable devices, antenna misalignment has been identified as a key factor reducing the accuracy of these estimations. To address this issue, an index was proposed to evaluate measurement uncertainty through model-based analysis, enhancing data accuracy. Additionally, data accuracy was improved by incorporating labeling from electromagnetic radar and hammering inspection results. The efficacy of the proposed method was validated through data analysis from building exterior wall inspections at construction sites, followed by comparisons of the estimated results. Notably, when classifying tiles as healthy or defective, linear separations in void depth and size increased by 0.549, while separation in void size increased by 0.522.