A STUDY ON THE APPLICABILITY OF DEEP LEARNING-BASED DETERIORATION DETECTION TECHNOLOGY TO SEWER PIPELINE SCREENING SURVEYS

Abstract

 In the maintenance and management of social infrastructure, streamlining and improving the efficiency of inspections and surveys for the purpose of understanding the current conditions has become a pressing issue in recent years. In response to this challenge, technologies utilizing AI, such as deep learning, have gained attention. However, deep learning heavily depends on the quality and quantity of data, and achieving practical accuracy remains a bottleneck. Therefore, in this study, we aimed to improve the accuracy of deterioration detection in sewer pipes using deep learning, and subsequently examined practical applications for its implementation in actual operations. As a result of the study, it was confirmed that deterioration detection utilizing deep learning has potential for application in screening surveys of sewer pipes.