Bacterial prediction using internet of things (IoT) and machine learning

Area of Science:

• Environmental Science
• Computer Science
• Public Health

Background:

• Water quality is critical for public health, increasingly threatened by pollution and dwindling freshwater sources.
• Traditional water quality monitoring methods are often slow, labor-intensive, and less reliable.
• Developing countries face challenges with ineffective, human-dependent water quality management systems.

Purpose of the Study:

• To develop a systematic and automated system for regular water quality monitoring and management.
• To leverage the Internet of Things (IoT) for real-time, remote water quality measurements with minimal human intervention.
• To enhance bacterial prediction accuracy and efficiency compared to conventional laboratory analyses.

Main Methods:

• Deployment of IoT sensor nodes measuring parameters like dissolved oxygen, turbidity, pH, temperature, and total dissolved solids.
• Data transmission via GSM modules to a central server for processing and analysis.
• Water quality classification using indices and bacterial prediction via machine learning algorithms (SVM, Bayesian Regression).
• Development of a web portal for real-time data visualization, including heat maps and infographics.

Main Results:

• Support Vector Machines (SVM) and Bayesian Regression models demonstrated optimal performance in fecal coliform bacterial prediction, achieving low Mean Squared Error (MSE).
• The IoT system successfully collected real-time water quality data and historical data for analysis.
• A web portal was developed for effective data visualization and monitoring.

Conclusions:

• The proposed IoT system offers a convenient, scalable, and cost-effective alternative to manual laboratory water quality testing.
• Real-time monitoring and machine learning-based bacterial prediction can significantly improve water resource management and public health protection.
• The system provides a portable hardware solution for remote and continuous water quality assessment.