IoT-Based Multi-Sensor Water Quality Testing_ A Low-Cost and Scalable Approach

<b>Title: </b><b>IoT-Based Multi-Sensor Water Quality Testing: A Low-Cost and Scalable Approach</b><b>Abstract</b>This research presents the development and evaluation of a scalable, low-cost, and real-time water quality monitoring platform using Internet of Things (IoT) technology. The system is based on the ESP32 microcontroller and integrates a suite of sensors, including the DS18B20 waterproof temperature sensor, a Total Dissolved Solids (TDS) sensor, the JV-3001 turbidity sensor, and a BME280 atmospheric sensor. Data collected is transmitted to both Blynk and Adafruit IO cloud platforms using HTTP and MQTT protocols respectively.The system is powered by a 12V 7Ah sealed lead-acid battery regulated through a waterproof buck converter and can function autonomously for up to 96 hours in field simulations. The monitoring unit has been tested in controlled aquatic environments to simulate field deployment. Over 43,000 sensor datapoints were recorded in 30 days, forming a valuable dataset for future AI-driven forecasting. The platform's educational and open-source architecture enables integration into K–12 STEM curriculum and supports community-driven environmental monitoring. Future plans include integration of RS485 sensors, advanced calibration, and predictive analytics using machine learning.<b>1. Introduction</b>Access to clean and safe water is a fundamental human right, yet millions of people globally face daily challenges due to pollution and poor water resource management. Traditional methods of water quality assessment are labor-intensive, costly, and time-delayed due to lab analysis. This creates gaps in real-time environmental response and leaves communities vulnerable to unmonitored contamination events.Recent advances in microcontrollers, cloud infrastructure, and affordable environmental sensors have made it possible to democratize environmental monitoring. This project builds on open-source hardware, low-cost sensors, and modular design principles to deliver an IoT-enabled system for continuous water quality assessment.The project’s goal is threefold:Develop a portable, power-efficient, and cloud-connected water monitoring system.Empower educators and citizen scientists to use the system for environmental education and advocacy.Contribute to a data-driven foundation for environmental governance, especially in underserved communities.<b>2. System Overview</b>The system integrates the following components:<b>ESP32 microcontroller</b> with WiFi/Bluetooth capability.<b>DS18B20 waterproof temperature sensor</b> to monitor water temperature.<b>TDS Sensor</b> to estimate the ionic concentration of dissolved solids.<b>JV-3001 Turbidity Sensor</b> for water clarity (currently used in analog mode).<b>BME280 Sensor</b> for ambient pressure, humidity, and temperature.<b>12V 7Ah SLA Battery</b> + <b>DC-DC buck converter</b> for power regulation.<b>Cloud Connectivity</b> using <b>Blynk (HTTP)</b> and <b>Adafruit IO (MQTT)</b>.A <b>modular architecture</b> supports flexible sensor expansion, calibration, and software upgrades. The sensor values are displayed live via mobile/web dashboards, and data is logged in the cloud for time-series analysis.<b>3. Methodology</b><b>3.1 Calibration and Testing</b><b>DS18B20</b>: Calibrated using ice and boiling water methods; consistent within ±0.5°F.<b>TDS</b>: Tested using standard solutions (500 µS/cm and 1413 µS/cm); ~5% error observed under fluctuating field-like conditions.<b>Turbidity</b>: The JV-3001 sensor currently provides relative data; calibration to NTU standard is pending.<b>Power Consumption</b>: The full system runs ~96 hours on a 12V 7Ah battery, with testing underway to incorporate sleep cycles and solar charging.<b>3.2 Cloud Transmission and Visualization</b><b>Blynk</b> allows mobile-based dashboard control using virtual pins (V0–V5) and is ideal for fast prototyping and simple alerts.<b>Adafruit IO</b> provides long-term data logging and visualization via MQTT, which is optimized for bandwidth and battery consumption.<b>3.3 Real-Time Transmission and Logging</b>Data is published at 1-minute intervals, generating over 1,400 readings per day.MQTT buffering and SD card logging will support offline deployment with delayed synchronization.<b>4. Results &amp; Performance</b><b>4.1 Controlled Environment Testing</b>Initial testing in a fish tank showed:TDS accuracy within 5% of expected.Biofilm formation on DS18B20 after 72 hours, causing 1.2°F drift.Humidity and temperature trends correlated well with reference indoor sensors.<b>4.2 Adafruit IO Visualization</b>Data trends over 5 days were visualized using custom dashboards, showing successful MQTT transmission. Humidity fluctuations and temperature changes were logged and are being used for future AI model training.<b>5. Community Science and Education Impact</b><b>5.1 Citizen Science Toolkit</b>The project includes:Open-source documentation for replication.Calibration protocols and community training materials.Public dashboards for shared access and interpretation.<b>5.2 Curriculum Integration</b>Water quality modules aligned with <b>NGSS</b> standards are being piloted in schools. These include activities like turbidity observation, water testing, data logging, and coding.<b>5.3 Community Engagement</b>Local educators and environmental clubs have been involved in:Data collection and validation through photo logs.Design of user-friendly dashboards based on field feedback.<b>6. Limitations and Technical Challenges</b><b>Turbidity sensor</b> calibration is ongoing; current values are relative.<b>RS485 sensors</b> introduced communication errors (Modbus instability).<b>Cellular modules</b> (SIM800/900) were incompatible with local bands, requiring a hotspot for WiFi.<b>Sensor fouling</b> and temperature drift require maintenance or AI-based correction.<b>Power management</b> needs further optimization for autonomous, multi-week deployments.<b>7. Future Directions</b><b>7.1 Sensor Expansion</b>RS485 sensors for EC and NTU monitoring.Modbus communication with MAX485 transceivers.<b>7.2 AI Integration</b><b>LSTM time-series models</b> for forecasting.<b>Anomaly detection</b> (e.g., One-Class SVM).<b>Explainable AI (XAI)</b> techniques (e.g., SHAP) to validate insights.<b>7.3 Policy and Governance Integration</b>Exportable data for GIS overlays and water planning.Real-time alerts for local municipalities or citizen watch groups.Tools for <b>evidence-based decision making</b> in water resource management.<b>8. Broader Impact and Replication Potential</b>This system was designed not only for technical performance, but to be <b>replicated in schools, municipalities, and environmental NGOs</b>. It provides an affordable, open, and reliable framework for:<b>Water justice and transparency</b> in marginalized communities.<b>STEM learning opportunities</b> using real-world data.<b>Data-driven governance</b> using modern environmental tech.The system can serve as a foundation for smart water infrastructure, climate resilience planning, and environmental accountability. Its integration of community feedback, machine learning, and accessible hardware ensures it is future-ready.<br>