Wind Turbine Inspection

Wind turbine defect detection, drone imaging, few-shot detection

🏆 Best Paper Award — ISPACS 2025

Our work on wind turbine defect detection was recognized with the Best Paper Award at ISPACS 2025.

Drone-based Inspection

Wind turbine blade inspection is critical for maintaining operational efficiency and preventing catastrophic failures. Traditional inspection methods are time-consuming and hazardous. We develop automated defect detection systems using drone-captured imagery, leveraging deep learning and computer vision techniques for efficient and accurate inspection.

Drone-based wind turbine inspection
Drone-based wind turbine blade inspection system and captured imagery.

Defect Detection

We develop deep learning models for detecting various types of blade defects including cracks, erosion, and lightning damage. Our approaches address the challenges of limited training data and class imbalance through few-shot learning and data augmentation strategies.

Defect detection results
Automated defect detection results on wind turbine blade imagery showing detected anomalies.

Few-shot Detection

Collecting large annotated datasets for wind turbine defects is expensive and impractical. We develop few-shot detection methods that can learn to detect new defect types from only a handful of examples, enabling rapid deployment of inspection systems for diverse turbine configurations.

Few-shot detection
Few-shot defect detection framework and results with limited training samples.

Related Publications

  • I. Gohar, A. Halimi, and S. John, "A comprehensive review of wind turbine blade defect detection using deep learning," Renewable and Sustainable Energy Reviews, 2025.
  • I. Gohar, A. Halimi, and S. John, "Graph attention network for wind turbine blade defect detection," IEEE Trans. Instrumentation and Measurement, 2025.
  • I. Gohar, A. Halimi, and S. John, "Few-shot wind turbine blade defect detection," in Proc. IEEE ISPACS, 2025. Best Paper