1Associate Professor, Department of Civil Engineering, Osmania University, Hyderabad. ORCID: 0000-0001-8133-9871; Email: qamarsultana4@gmail.com
*Corresponding Author: Dr. Qamar Sultana, Associate Professor, Department of Civil Engineering, Osmania University, Hyderabad. Email: qamarsultana4@gmail.com
Smart water distribution networks need smart continuous monitoring to minimize the impact of transient disturbances which conventional sampling strategies cannot easily detect. The current study proposes the AI-IoT solution to the Anomaly Detection in real-time in Multi-node drinking water distribution systems using high frequency sensor readings. A homogenous preprocessing pipeline standardizes the heterogeneous streams of nodes which in turn models temporal features which capture short-term trends and inconsistencies. In statistic strength with weak supervision strategy, no known contamination labels, proxy anomaly ground truth with multi sense node specific thresholds is obtained. The accuracy of the RF classifier trained on chronological validation was 92.19, recall was 89.18 and ROC-AUC was 0.951 which is a high anomaly detector. The variability of pH and ORP contributed to the prediction of turbidity-related characteristics which became the most important. Prioritization- being based on streaming simulation and severity was used to convert the detections into actionable alerts. The framework illustrates how intelligent monitoring (scaled and deployment oriented) may be applied for improving smart water distribution systems in terms of their safety and resilience.
Keywords: Smart Water Management, Internet of Things (IoT), Anomaly Detection, Machine Learning, Water Distribution Networks
How to cite this article: Sultana Q. Smart Water Management: Integrating AI and IoT for Real-Time Monitoring of Water Quality and Supply Networks. Int J Drug Deliv Technol. 2026;16(12s): 265-274. DOI: 10.25258/ijddt.16.12s.28
Source of support: Nil.
Conflict of interest: None