A convolutional neural network for pipe crack and leak detection in smart water network

Abstract

The implementation of a smart water network (SWN) is viewed as a strategic approach to address many challenges faced by water utilities, such as pipe leak detection and main break prevention. This paper develops a convolutional neural network (CNN)–based model to classify acoustic wave files collected by the South Australian Water Corporation’s (SA Water’s) SWN over the city of Adelaide. The VGGish model (VGG refers to the team who developed the model—Visual Geometry Group) is selected as a suitable transfer learning model to extract features from wave files. The CNN model classifies an acoustic wave file as an anomaly or other background or environmental noise. Identification of a wave file as an anomaly triggers a Siamese CNN model to determine whether it is related to a regular/irregular scheduled event (for example, irrigation system near public parks or water consumption by large buildings). A field investigation is initiated if a wave file is classified as an anomaly and it is not related to a scheduled event. The developed models have been validated using data that is recorded by SWN in Adelaide. This validation data set comprises 1098 wave files, which are recorded by 34 accelerometers and are associated with 32 known leaks. The validation results shown that accuracy of alarms generated by the developed models is 92.44%. The validations confirm the developed models as an effective tool for water pipeline leak and crack detection, which, in turn, enables proactive management of the pipeline assets.

Machine Learning Model for Leak Detection Using Water Pipeline Vibration Sensor

“…Yu et al present a study on the effectiveness and practicability of using machine learning models to identify leaks in real pipe networks by classifying vibration signals collected by piezoelectric accelerometers installed in water distribution systems over several cities of China [21]. Zhang et al describe the development of a convolutional neural network (CNN)-based model to classify acoustic wave files collected by the South Australian Water Corporation's (SA Water's) smart water network (SWN) over the city of Adelaide for pipe leak and crack detection with an accuracy of 92.44% [22]. Vanijjirattikhan et al present the development of an AI-based water leak detection system with cloud information management that can systematically collect and manage leakage sounds and generate a model used by a mobile application to provide operators with guidance for pinpointing leaking pipes [23].…”

A Review on Current Technologies and Future Direction of Water Leakage Detection in Water Distribution Network

“…A more detail discussion of the selected articles is presented in the next section. - [93] Accelerometer Acoustic E L Accelerometer data was transformed to spectogram. The spectogram was then classified with a CNN and if there is any anomaly then the data is then fed to Siamese CNN to check -for scheduled events.…”

Using J-integral analysis to predict the failure of HDPE pipes under crack face load and internal pressure

“…3 Furthermore, high system pressure exacerbates this problem by putting extra stress on the already weakened infrastructure, leading to more leaks and breaks. Addressing these issues requires a comprehensive approach, including regular maintenance and monitoring of the distribution system, as well as the use of advanced technologies to detect 4,5 and repair leaks promptly and the use of modern materials taking into consideration the mechanical strength, design, etc. The presence of cracks and other types of damage in water pipes can significantly impact the overall efficiency of the water distribution system.…”