IOT- Enabled Real -Time Water Quality Monitoring Using Electrochemical Sensors and LoRa Communication

Abstract

The dumping of untreated sewage water into the natural water bodies has been one of the most recurring environmental problem in evolving urban ecosystems. Periodic laboratory testing based traditional sensing methods cannot satisfy the time constraints required for proactive environmental management. This paper focuses on a fully integrated multi-layer IoT framework designed for real-time, continuous monitoring of sewage water treatment plants and their discharge points. Eight electrochemical and optical sensors (for PH, dissolved oxygen (DO), electrical conductivity (EC), oxidation–reduction potential (ORP), turbidity, water temperature, ammonia-nitrogen (NH 3-N) and chemical oxygen demand (COD)) interfaced with aESP32 based edge computing node where a real-time Kalman filter noise reduction and preliminary threshold check is performed before the data is transmitted in JSON format through MQTT protocol to a cloud analytics platform. A novel CNN-LSTM network model is trained on 547,920 enhanced training samples in order to assign a specific class to a given contamination condition on one of the following levels-Normal, Warning, Alert and Critical. A multi-channel notification subsystem generates alarms through SMS (via Twilio API), e-mail, local relays controlled sirens and an automatic regulatory reporting interface. The system is validated in 45-day filed operation at Nesapakkam Municipal Sewage Treatment Plant, Chennai with a 98.12% classification accuracy, 4.23 second end to end alarm propagation and a system availability of 99.87% and is planned to be released under an open-source paradigm and can be easily scaled in rapid urbanizing zones without a considerable capital investment on proprietary components.