iPipe: Water Pipeline Monitoring and Leakage Detection

Abstract

Pipelines are considered to be a part of the transportation sector. The pipeline industry moves various substances such as crude oil, refined petroleum products, and natural gas within thousands of miles of pipelines. There may be various reasons that can cause the failure of these pipeline systems, resulting in some serious disasters. So, it is necessary to detect the leakage of water and oil pipelines by which the accidents can be avoided and damage is minimal. This paper proposes iPipe, i.e. an intelligent water pipeline monitoring and leakage detection system that is based on an acoustic signal method to detect and locate leaks in pipelines by using a network of acoustic and GPS sensors to continuously monitor the sound in the vicinity of the pipe. The system uses signal processing to identify the frequency and characteristics of leak sound and machine learning techniques to differentiate between characteristic sound and the normal sounds in the environment near the pipe. The sensors provide information about the leak as soon as possible to the designated endpoint i.e, cloud server. All of them have different locations and IDs so that it is possible to know where the data came from. This work reduces the need for human intervention by automatically notifying about the leakage. The proposed system has the ability to detect leaks with an accuracy of 95.6%.