ScholarWorks@성균관대학교

초록

As the members of temporary structures at construction sites are generally reused, inspection of individual temporary equipment members before use is essential to ensure safety during construction. Visual inspection, which is a common practice for assessing the quality of reused temporary equipment, is labor-intensive and depends on the skill of the inspector; thus, automated and accurate inspection methods are desired. However, little research has been devoted to the development of such methods. Because inspection criteria are mostly relevant to deformation, the shape measurement of temporary equipment members is the most fundamental consideration. This paper proposes an automated inspection method for temporary equipment based on measured 3D (three-dimension) shapes. A hardware system with a laser profiler and motion stage was designed to automatically measure 3D shapes. The 3D shape of each member has damage characteristics, such as hole deformation, thread abrasion, excessive clearance, and dents. Damage-detection algorithms for each inspection criterion were developed using machine learning and the geometric features of 3D shapes. The proposed method was validated using temporary equipment samples, with and without damage.

키워드