Development of Autonomous Hammering Test Method for Deteriorated Concrete Structures Based on Artificial Intelligence and 3D Positioning System

Abstract

It is very important nowadays to establish economic and efficient management systems for existing concrete infrastructures, in order to fulfill their service designed lives and even to extend them. Hammering test, which is a manual inspection technique detecting defects and damages, is widely used to judge the soundness of tunnels, bridges, roads and railways. However, the reliability of the diagnosis results using hammering test data is not described quantitatively but handled only with inspectors’ professional experience. This study aims to propose an autonomous hammering test and deterioration diagnosis system, developing artificial intelligence (AI) systems which recognize the internal defects of reinforced concrete members. As for the three-dimensional (3D) positioning system installed in the autonomous solenoid hammering device with a microphone, it is also equipped with a gyro sensor and an accelerometer so that it is ongoingly collecting the hammering sound and detecting the 3D location of hammering location in real time of inspection on site. Consequently, in the present study, the frequency spectrum analysis based on the wavelet transform of acoustic hammering sound data gives the recognition for deterioration probability followed by the artificial neural networks with the diagnosis algorithm consisting of image binary processing of the wavelet tomogram and output system.