Method of automated inspection for reinforcement cages of precast concrete elements

Precast concrete elements (PCE), widely used in the construction industry, have advantages over in-situ concreting. Its structural quality depends on the assembly of the reinforcement cages, and it is essential to ensure that they comply with the design before pouring the concrete. However, the existing manual inspection methods depend on the inspector's experience, are subjective and time-consuming, and use measuring devices such as measuring tapes. An inspection method that is efficient, practical, and objective is needed. Previous studies have explored 3D laser scanning technology. However, only some studies focus on integrating low-cost technologies, such as Building Information Modeling (BIM) and Machine Learning (ML), for quality inspection of reinforcement cages. Following a design science approach, this paper develops a method for automated inspection of reinforcement cages called FV-Smart, which integrates BIM and AM, comparing the assembled with the designed using an artificial intelligence algorithm. The computer vision model presented a performance of 72.6% (Precision), 68% (Recall), and 81.1% (Average Precision). The proposed artifact supports managers in decision-making to increase the reliability and robustness of the information generated during inspections.