IoT enabled smart water metering using multi sensor data and machine learning techniques
Water Distribution Systems (WDS) are critical infrastructure assets that deliver water from source to consumers. The increasing scarcity of fresh water has heightened the importance of monitoring these systems. Conventional smart metering solutions require intrusive installation in pipelines, increa...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2025-06-01
|
| Series: | Frontiers in Water |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/frwa.2025.1586916/full |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Water Distribution Systems (WDS) are critical infrastructure assets that deliver water from source to consumers. The increasing scarcity of fresh water has heightened the importance of monitoring these systems. Conventional smart metering solutions require intrusive installation in pipelines, increasing costs and complexity. Moreover, in intermittently operated networks, which are common in India and other countries of the global south, the line is not pressurized for considerable amounts of time, resulting in poor performance of conventional water meters. Periodic maintenance of these meters can cause similar disruptions. This study introduces a novel non-intrusive technique for WDS monitoring by measuring water consumption, offering a cost-effective alternative to existing smart meters. The system can be effectively built, including installation, at a fraction (1/10th) of the cost of existing smart meters. The proposed technique utilizes low-cost level sensors in OverHead Tanks (OHTs), sumps or reservoirs, which are used in many cities, towns, and villages in the global south to cope with the intermittent supply. Two estimation approaches are explored: predefined flow rates from baseline experiments and a dynamic method that adapts to variations in tank level. The methodology is validated through controlled experiments and from actual operating systems, demonstrating its effectiveness in handling fluctuations in inflows and intermittent outlet flows. Results show that while predefined flow rates offer accuracy in stable conditions, dynamic estimation is more adaptable to real-world variability. This approach enables scalable and affordable smart water monitoring, contributing to sustainable water management. |
|---|---|
| ISSN: | 2624-9375 |