I built an AI system that detects solar panel faults from a single image โ and deployed it with a live demo.
Here's what it does:
๐ฒ ๐ณ๐ฎ๐๐น๐ ๐ฐ๐น๐ฎ๐๐๐ฒ๐: Bird-drop, Dusty, Snow-Covered, Electrical Damage, Physical Damage, Clean
ย โข ResNet101 fine-tuned on solar panel imagery
ย โข 96.3% training accuracy with strong per-class F1
ย โข REST API deployed on Hugging Face Spaces (Docker + Flask)
ย โข React frontend with real-time confidence scores
๐ช๐ต๐ ๐๐ต๐ถ๐ ๐บ๐ฎ๐๐๐ฒ๐ฟ๐:
According to Raptor Maps' 2025 Global Solar Report, equipment-driven underperformance has increased 214% since 2019 โ resulting in $10 billion in lost energy value in 2024 alone. The average solar site loses 5.77% of its power output annually, costing ~$5,720 per MWdc.
The use cases that excite me most:
ย โข Drone inspection over utility-scale farms
ย โข Ground robotics for close-up panel analysis
ย โข Direct integration into SCADA systems via REST API
๐๐๐ถ๐น๐ ๐๐ถ๐๐ต: PyTorch ยท ResNet101 ยท Flask ยท Docker ยท HuggingFace Spaces ยท React + Vite