A Comprehensive Review of AI and IoT-Based Water Pipeline Leakage Detection, Prediction and Monitoring Systems

Water pipeline leakage is a major challenge in water distribution and supply systems because undetected leaks can result in water loss, infrastructure damage, economic losses, and inefficient water management [5], [13]. Conventional leakage inspection methods often depend on manual inspection, periodic maintenance, or specialized equipment, which may not provide continuous real-time monitoring [5], [13]. Recent developments in the Internet of Things (IoT), sensor technology, artificial intelligence (AI), and machine learning (ML) have created opportunities for automated and intelligent pipeline monitoring [9], [11]. This review presents a study of recent techniques used for water pipeline leakage detection, localization, prediction, and monitoring. Different sensing approaches, including flow, pressure, acoustic, and vibration measurements, are reviewed along with machine-learning and deep-learning methods [1], [3], [4], [8], [12], [13]. Existing research demonstrates that pressure and flow measurements are widely used for hydraulic anomaly detection, while vibration and acoustic measurements can provide additional information about leakage events [5], [6], [7]. AI models such as Artificial Neural Networks, Random Forest, Support Vector Machines, K-Nearest Neighbors, Convolutional Neural Networks, and ensemble techniques have been investigated for leakage classification and localization [1], [3], [4], [8], [12], [13]. IoT-based systems further enable continuous data acquisition, remote monitoring, and real-time alerts [9], [11]. However, challenges remain in obtaining representative real-world datasets, reducing false alarms, handling sensor noise, and deploying models on resource-constrained edge devices [13], [14], [15]. Based on the reviewed literature, a multi-sensor IoT and AI architecture using flow, pressure, and vibration measurements with Raspberry Pi-based processing is proposed as a suitable direction for a low-cost academic prototype [5], [9], [11], [12]..