Traffic Guardian is envisioned as a real-time traffic incident detection and reporting system that enhances road safety and operational efficiency across Gauteng's high-volume highways. Using live video streams from existing traffic cameras, the system will automatically detect and classify incidents such as accidents, congestion, stalled vehicles, and road hazards.
As the implementing team, our approach is structured around building a robust and scalable prototype within the constraints of the AWS Free Tier. We will use computer vision and deep learning models to analyse live camera feeds, identify anomalies, and trigger real-time alerts with geolocation data. This information will be displayed via a web-based dashboard designed for use by traffic control centre operators.
To ensure a modular and maintainable system, our development process will be guided by software engineering best practices, including:
To adapt to changes and iterate quickly.
To ensure all components align with real-world traffic operations.
To meet POPI compliance and enforce data protection.
Continuous analysis of traffic camera feeds for immediate incident detection.
Automated identification of accidents, congestion, and road hazards.
Real-time visualization and management of detected incidents.
Immediate notifications to traffic control operators and emergency services.
Access to incident data for field personnel via mobile-optimized interfaces.
Searchable database of historical incidents for analysis and reporting.
Our solution will emphasise automation, accuracy, and ease of use, transforming passive camera networks into active monitoring tools while giving traffic personnel the insights needed to respond swiftly and effectively.