Smart Automated Solar Panel Cleaning System
Graduation project: a YOLOv8-nano model detects dust on solar panels and the system cleans them in response. Dataset pipeline, embedded hardware, a mobile app and backend services integrated into one system.
- AI & Computer Vision
- 2026
- Completed
- AI, backend and integration (graduation project)
Overview
Dust on solar panels quietly cuts their output. This graduation project at King Faisal University (2025–2026) detects it automatically: a YOLOv8-nano model, trained on a purpose-built dataset, recognises dust on the panel surface and the system cleans the panel in response.
My part spanned the dataset pipeline, training and deploying the model, and integrating the AI with the embedded hardware, the mobile app and the backend services, so that detection, cleaning and monitoring work as one system.
Key features
- YOLOv8-nano dust-detection model
- Dataset pipeline and model training
- Embedded hardware integration
- Mobile app
- Backend services