Intelligent Audit System

This system automates the audit of public infrastructure such as sidewalks and parks by analyzing images or video streams using YOLOv11. Detected elements are scored using a Mamdani Fuzzy Logic engine to produce a compliance index aligned with Indonesia's PUPR standards.

Built on a Python/Flask backend with YOLOv11 for real-time object detection and a custom Mamdani Fuzzy Logic module for quality scoring.

I led the end-to-end development — training the YOLOv11 model on a custom dataset, implementing the fuzzy inference system, building the Flask API, and integrating everything into a web dashboard for auditors.

• Real-time object detection using YOLOv11
• Mamdani Fuzzy Logic quality scoring engine
• Automated compliance report generation
• Interactive web dashboard for audit review
• Geolocation tagging for audit records