جامعة البيان

Water pollution represents one of the most pressing environmental challenges facing developing nations, particularly in regions where industrial development outpaces regulatory oversight. The Tigris River in Iraq exemplifies this challenge, where rapid industrialization and inadequate monitoring infrastructure have converged to create significant public health risks. This work introduces Dijlah Eye, a comprehensive IoT monitoring system that combines multiple water quality sensors with machine learning techniques to detect contamination events in real time. Our approach departs from conventional laboratory-based analysis by deploying distributed sensors capable of immediate anomaly detection and automated response mechanisms. The system integrates four distinct sensing modalities—pH measurement, dissolved solids quantification, turbidity analysis, and temperature monitoring—with an Isolation Forest algorithm that learns normal water chemistry patterns without requiring labeled anomalous data. Upon detection of anomalous conditions, the system activates automated purification hardware and dispatches alerts to stakeholder email addresses. Field validation confirms 95% detection accuracy while maintaining a hardware cost well below $150 per installation, making the technology economically viable for deployment across resource-limited contexts.

This report presents Safety Cam, an intelligent computer vision system designed to enhance workplace safety and minimize asset loss in industrial environments. The system utilizes real-time deep learning to detect the compliance of personal protective equipment (PPE) and track tools/assets using workplace surveillance cameras. In case of violations or unreturned equipment, immediate notifications are transmitted via Telegram API to administrative groups and individual worker devices.