Smart Water Quality Monitoring System

Abstract

Among the largest global health, environmental, and industrial issues is water pollution.The newest technologies, such as machine learning and the Internet of Things, are being studied to meet the growing demand for water monitoring solutions. In order to determine the factors of water quality, including temperature, turbidity, dissolved oxygen (DO), and pH level, and electrical conductivity in real-time and predict them, this study proposes a Smart Water System for Monitoring Quality, which combines of IoT sensors, cloud storage, and machine learning algorithms. To predict the degree of contamination and issue automatic alerts to the concerned parties through which necessary action can be taken, The gathered sensor data isArtificial Neural Network analysis(ANN) to enable prompt action in the urban and industrial water bodies as well as agricultural waters.

Introduction

• Low water quality can cause environmental damage, health risks, and financial losses.

• Ensuring clean, safe water for drinking, agriculture, and industry is a global priority.

• Traditional monitoring methods are often slow, manual, and resource-intensive.

2. Proposed Solution: Intelligent Water Quality Surveillance System

A smart, multi-layered framework combining IoT sensors, fog computing, and machine learning to monitor and predict water quality in real time.

Key Technologies Used:

• IoT Sensors: Continuously collect data on parameters like pH, turbidity, temperature, dissolved oxygen (DO), and electrical conductivity (EC).

• Fog Computing: Processes data closer to the source, reducing latency and enabling real-time alerts.

• Cloud Computing: Stores large-scale data and performs advanced analytics using Artificial Neural Networks (ANNs).

• Machine Learning: Enables predictive analysis and early detection of water contamination.

3. System Architecture

Three-layer structure:

1. Sensing Layer (IoT)

   • Collects real-time water quality data via distributed sensors in rivers, lakes, reservoirs, and industrial sites.

2. Fog Layer (Smart Gateway)

   • Processes data locally for faster decisions.

   • Detects anomalies and sends instant alerts to stakeholders.

   • Performs data preprocessing and edge analytics.

3. Cloud Layer

   • Long-term storage of data.

   • Runs AI-driven models (ANNs) to detect patterns and predict contamination events.

   • Offers visualization dashboards, reporting tools, and decision support.

4. Machine Learning Model: Artificial Neural Network (ANN)

• Steps:

  1. Standardize sensor data.

  2. Train ANN using backpropagation.

  3. Use thresholds and layers to classify water as safe or hazardous.

  4. Continuously update the model using real-world data.

• Goal: Predict contamination events with high accuracy.

5. Experimental Results

• Simulation Tools: Used Microsoft ML.NET for AI model development and testing.

• Performance Metrics:

  • Accuracy: 93.21% for ANN (better than CNN, RNN, and MLP).

  • Precision, Recall, F1 Score: Also highest for ANN.

• Comparison Models:

  • CNN: 72.13% accuracy

  • MLP: 90.04%

  • RNN: 87.21%

Training Dataset Size Optimization:

• Optimal training data size: 75–80% of total data.

  • <75%: Risk of underfitting.

  • >80%: Risk of overfitting.

• ANN model maintained performance consistency across training sizes.

6. Benefits & Applications

• Real-time monitoring and alerts for water contamination.

• Cost-effective: Reduces manual sampling and lab testing.

• Scalable and sustainable for both urban and rural settings.

• Aids policymakers, water management authorities, and environmental agencies with data-driven decision-making.

7. Related Work

• Past studies explored IoT, edge computing, and ML for environmental monitoring.

• Some innovations:

  • Blockchain for secure data transmission.

  • Solar-powered sensors for rural deployment.

  • FPGA-based preprocessing to extend sensor life.

  • LSTM models to predict water trends in smart cities.

Conclusion

One of the biggest risks to world health is safe water quality. As a result, early monitoring of water quality indicators is necessary to prevent waterborne diseases. This essay suggests An intelligent water quality monitoring device that analyzes water parameters in real time using a Wireless Sensor Network (WSN) made possible by the Internet of Things (IoT). Using IoT-based sensors placed in water bodies, the system first gathers important water quality metrics (such as temperature, turbidity, dissolved oxygen, pH, and chemical pollutants). When comparing the proposed system In contrast to existing water quality monitoring techniques, the proposed system may detect pollutants early and anticipate the dangers that could result in contaminated water or other hazardous situations. The Artificial Neural Network (ANN) model used in this study has an incredibly high accuracy of 94.58.

References

[1] The study \"Prediction of water quality biomarkers in aquatic environments,\" by Hussain and S.-J. Park, was published in August 2021 in Sensors, vol. 21, no. 16, pp. 5334. [2] In September 2016, Q. Song, X. Liu, W. Zhou, L. Wang, X. Zheng, X. Wang, and others published \"Impact of prolonged water stagnation on water quality and ecosystem health: The Kailuan case study,\" in Sci. Rep., vol. 6, no. 1, pp. 1-9. [3] A. De Vos, J. De Keyser, and S. De Raedt, \"Autonomic sensing and monitoring of aquatic pollution: A novel therapeutic approach to water purification,\" Journal of Environmental Science, vol. 348, no. 1, pp. 24-34, January 2015. [4] Hussain and S.-J. Park\'s article, \"Quantitative evaluation of water quality fluctuations after environmental changes,\" appeared in Brain Sci. in July 2021. It was published on page 900 of volume 11, issue 7. [5] P. Verma and S. K. Sood, Fog-assisted IoT- enabled water monitoring of quality in smart cities, IEEE Internet Things Journal, vol. 5, no. 3, pp. 1789-3996, June 2018. [6] A portable system for checking the quality of water of remote areas based on IoT services, IEEE Trans. Inf. Technol. Biomed., vol. 14, no. 3, pp. 748-757, May 2010, Z. Yan, J. Shi, P. Kandachar, and A. Frendenthal. [7] In their paper \"MAFC: Multi-agent fog computing model for monitoring critical water quality tasks,\" A. Mutlag, M. K. A. Ghani, M. A. Mohammed, M. S. Maashi, O. Mohd, S. A. Mostafa, et al., Sensors, vol. 20, no. 7, pp. 1853, March 2020. [8] \"Water quality monitoring and detection platform to prevent anomalous water conditions in real-time,\" by G. Villarrubia, J. Bajo, J. De Paz, and J. Corchado, Sensors, vol. 14, no. 6, pp. 9900-9921, June 2014.

Copyright

Copyright © 2025 Imran Ali S, Prakash O. Sarangamath. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.