Development of an IoT-Based Real-Time Water Level Monitoring System for Flood Observation

Abstract

Flooding remains a significant urban hazard because rising river levels can disrupt mobility, public activities, infrastructure, and environmental conditions. Continuous water-level observation is therefore needed to provide timely operational information in flood-prone urban tributaries. This study aims to develop and field-evaluate an IoT-based water-level monitoring system that supports continuous measurement, MQTT-based data transmission, threshold-based status classification, and real-time dashboard visualization, rather than validated flood early warning. The system was implemented at the Sekanak tributary in Palembang, Indonesia, by integrating a water level sensor, WELPRO 8026ADAM data acquisition module, TGW-725 gateway, RUT200 industrial router, MQTT publish–subscribe communication, and a Node-RED web dashboard. Incoming measurements were processed in Node-RED and classified into Safe, Alert, and Danger categories using predefined water-level thresholds. Evaluation was conducted as an initial 24-hour field observation with a 30-second sampling interval under non-danger monitoring conditions. The system recorded 2,880 measurements and displayed current water level values, historical trends, and classification labels on the dashboard. The observed water levels ranged from 1100 mm to 2100 mm, with an average value of 1200 mm. Most measurements were classified as Safe, several were classified as Alert, and no Danger-level measurements were recorded. Because latency, packet loss, data completeness, dashboard response time, and actual flood-event performance were not quantitatively assessed, the findings should be interpreted as evidence of functional continuous monitoring and visualization rather than comprehensive real-time flood-warning validation.