Rigved IT

Blade Inspection AI

Wind turbine blade defect detection & anomaly mapping · Rigved Infotech

Models loading… Workforce Dashboard
Blades Inspected
0
unique images this session
Defective
0
FAIL verdicts
Passed
0
no damage found
Defect Rate
0%
High Severity
0
needs urgent repair
Avg Inference
—
per image · CPU
Inspection Feed
Loading inspection models…
First run builds the anomaly memory bank (about a minute).
Labelled defect (ground truth)
Image
—
Detected / Labelled
—
Top Confidence
—
Anomaly Score
—
Latency
—

Current Result

— —
Waiting for first inspection…

Defect Log

Defective blades will appear here.
Detector: YOLOv8 fine-tuned on the ILIAD Wind Turbine Damage Dataset (CC BY 4.0). Feed images are held-out test/validation images not used for training. Anomaly map: PatchCore-style nearest-neighbour distance on ResNet18 features of damage-free blade regions.