Design and Realization of Solar-powered IoT-based Flood Early Warning System with Telegram Messaging, Auto-restart Watchdog, and Power Management

Abstract

A Flood is inevitable in some regions in Indonesia during rainy season. Flood can cause problems ranging from simple problem such as lightly flooded street to a complex one such as water that enters the house. For community living at river side, the flood often comes from excessive water due to rain in the upstream that outflows the river as drainage which can happen unnoticeably. An early warning system is needed to detect flood so that people who lived in a flood-prone area can be warned and be prepared to mitigate the flood. With the help of internet of things (IoT) technology, such a system can be made. In this work, an early warning system was developed using off-grid solar system as power source, ultrasonic sensor to detect water level, NodeMCU as microcontroller, and 4G modem to connect to the internet. The system uses Blynk as its IoT platform and sends alarm messages to locals using Telegram so they are vigilant. The system can detect how much the water level and categorizes it into preset alarm levels. Lab. experiments on the proposed system was performed using power from PV-module installed on the lab.’s rooftop. The system functionality works well for 3 weeks of experiments with battery voltage monitored to vary from 12 V at dawn to 13.7 V at daytime. The experiments show that the system exhibits water level measurement average relative error of 2.22% and accuracy of 97.78%. The system is also furnished with power management and auto restart self-healing functionality which were also observed to work well during the test. Measurements show that the system only consumes energy of 11.52 Wh/day which is equivalent to 8% of battery depth of discharge, hence the system can operate for 3.75 days without charging during successive rainy and cloudy days and has sufficient solar energy harvesting margin to cope with site imperfections like shading and soiling of the PV-module.