IoT-based automated water pollution treatment using machine learning classifiers.

Area of Science:

• Environmental Science
• Computer Science
• Engineering

Background:

• Safe water accessibility is a global concern, with unsafe water posing significant health risks.
• Current water quality monitoring methods lack sufficient parametric measures and advanced technological integration.
• There is a critical need for accurate and efficient water quality classification systems.

Purpose of the Study:

• To develop an Internet of Things (IoT)-based automated system for real-time water quality monitoring.
• To leverage cloud computing and machine learning for accurate water quality classification.
• To provide a technologically advanced alternative to traditional water monitoring methods.

Main Methods:

• Integration of various sensors (pH, temperature, turbidity, conductivity) with Node MCU.
• Utilizing low-power wide-area networks (LPWAN) for data transmission to a cloud server.
• Employing a hybrid machine learning model combining K-Nearest Neighbour (KNN) and Support Vector Machine (SVM) for classification.

Main Results:

• The proposed KNN + SVM model achieved a highest accuracy of 0.94, validated by 10-fold cross-validation.
• The system effectively classifies water quality using sensor data transmitted via IoT.
• Demonstrated superior performance compared to existing water quality classification algorithms.

Conclusions:

• The developed IoT system offers an accurate and automated solution for water quality monitoring.
• This approach enhances public health by reducing risks associated with unsafe water consumption.
• The system contributes to environmental protection through reliable water status assessment.