Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{200204,
        author = {AmirthaVarshini BA and Saravanabalaji M and Karthika R and Baranidharan P},
        title = {A Machine Learning Framework For Leak Detection In Water Pipelines Using Multi-Sensor Data},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {12},
        pages = {550-557},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=200204},
        abstract = {Ensuring efficient monitoring of water pipeline systems requires accurate, real-time, and cost-effective leakage detection techniques that can be deployed in practical environments. This work presents an integrated leakage detection framework that combines multi-sensor data acquisition, frequency-domain signal analysis, and machine learning-based classification for real-time monitoring. A Raspberry Pi-based system is used as the central processing unit, enabling continuous acquisition and processing of signals from a hydrophone, differential pressure transducer (DPT), and pressure sensor.
The hydrophone captures acoustic variations caused by leakage, while the DPT and pressure sensor monitor pressure fluctuations within the pipeline. The acquired time-domain signals are transformed into the frequency domain using Fast Fourier Transform (FFT), enabling enhanced identification of leak-related patterns. Key spectral features such as peak frequency, power spectral density (PSD), mean amplitude, and bandwidth are extracted to form a structured dataset.
Machine learning models, including Logistic Regression and Random Forest, are employed to classify leak and no-leak conditions. Experimental evaluation demonstrates that the proposed multi-sensor approach improves detection accuracy compared to single-sensor systems. The system is implemented in real time, enabling continuous monitoring and immediate detection of leakage events. The results highlight the effectiveness of integrating signal processing and machine learning for intelligent pipeline monitoring. Future work includes extending the system with IoT-based remote monitoring and advanced deep learning models for improved performance.},
        keywords = {Water Leakage Detection; FFT; Machine Learning; Random Forest; Logistic Regression; Multi-Sensor Systems; Real-Time Monitoring},
        month = {May},
        }