Smart IoT and XGBoost-Based Model for Early Detection and Multi-Class Leak Risk Classification in the Drinking Water Supply and Sewerage Sector

Water distribution systems face significant challenges due to aging infrastructure, increasing water demand, and limited real-time monitoring capabilities, resulting in leaks, water losses, service interruptions, and higher operational costs. To address these issues, this study proposes a mobile application based on IoT technologies and Machine Learning for the detection and prediction of failures in domestic water distribution systems. The system uses an ESP32 microcontroller connected to a YFS201 flow sensor and a SEN0257 pressure sensor to collect and transmit real-time hydraulic data. The acquired information is displayed through a mobile application and analyzed using an XGBoost classification model to identify and classify pipeline failure risk levels, supporting early anomaly detection and preventive maintenance. Additionally, the application includes a technician assignment feature that records maintenance responsibilities, improving task traceability and management. Experimental results show that the proposed model achieved an overall accuracy of 79% and a weighted F1-score of 0.77. Its highest performance was obtained in detecting high-risk scenarios, reaching an F1-score of 0.89, demonstrating its effectiveness in supporting timely fault detection and maintenance planning in water distribution networks.

Authors

Publication Details

Journal
Preprints.org
Published
2026-11-18
DOI
https://doi.org/10.20944/preprints202611.1913.v1
Primary Topic
Water Systems and Optimization
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Smart IoT and XGBoost-Based Model for Early Detection and Multi-Class Leak Risk Classification in the Drinking Water Supply and Sewerage Sector

Edgar Jordan Onofre Ruiz, Carlos Alberto Alvarado Quintana, Renzo Martin Lavado Garayar
Preprints.org
Water Systems and Optimization
preprint

Smart IoT and XGBoost-Based Model for Early Detection and Multi-Class Leak Risk Classification in the Drinking Water Supply and Sewerage Sector

Edgar Jordan Onofre Ruiz, Carlos Alberto Alvarado Quintana, Renzo Martin Lavado Garayar
preprint en

Abstract

Water distribution systems face significant challenges due to aging infrastructure, increasing water demand, and limited real-time monitoring capabilities, resulting in leaks, water losses, service interruptions, and higher operational costs. To address these issues, this study proposes a mobile application based on IoT technologies and Machine Learning for the detection and prediction of failures in domestic water distribution systems. The system uses an ESP32 microcontroller connected to a YFS201 flow sensor and a SEN0257 pressure sensor to collect and transmit real-time hydraulic data. The acquired information is displayed through a mobile application and analyzed using an XGBoost classification model to identify and classify pipeline failure risk levels, supporting early anomaly detection and preventive maintenance. Additionally, the application includes a technician assignment feature that records maintenance responsibilities, improving task traceability and management. Experimental results show that the proposed model achieved an overall accuracy of 79% and a weighted F1-score of 0.77. Its highest performance was obtained in detecting high-risk scenarios, reaching an F1-score of 0.89, demonstrating its effectiveness in supporting timely fault detection and maintenance planning in water distribution networks.

Preprints.org
Water Systems and Optimization
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon