Detecting Road Pavement Cracks Based on Acoustic Signature Analyses

Transportation infrastructures can benefit from structural health monitoring in terms of pavement management systems and risk management. Pavement cracks, both visible and concealed, impact road agency budget but unfortunately there is lack of nondestructive methods to assess them. Consequently, the objectives were confined into setting up and improving a nondestructive, acoustic- and sensor-based method. An experimental investigation that was carried out on an asphalt concrete road pavement, aiming at deriving the Structural Health Status (SHS) of road pavements based on their acoustic response to a proper mechanical excitation (acoustic signature). The method was applied using as sensor device a microphone-based electronic system, which is able to gather only the ground-born sounds. Sensor data (i.e., the acoustic responses) were analyzed in three domains of analysis, i.e., the time, the frequency, and the time-frequency domain. Consequently, meaningful features (e.g., energy and entropy of the Continuous Wavelet Coefficients, spectral centroid) were extracted and used to derive the SHS of the road pavement under investigation, which represents a valuable information for different stakeholders (e.g., authorities, drivers, etc.). Results show that by using a small number of meaningful features and by applying a hierarchical clustering procedure, it is possible to recognize the variation over time of the acoustic signature of the infrastructure due to the presence and the propagation of internal and external cracks. Hence, the proposed method can be efficiently used to monitor the SHS of road pavements during their lifetime, and, consequently, to improve pavement management systems and risk management processes.