A comprehensive review of unmanned aerial vehicle-based thermal imaging and deep learning for PV power plant anomaly detection and performance assessment

Sabry A. H., BIYIKOĞLU A., ÇAMDALI Ü.

Engineering Applications of Artificial Intelligence, cilt.163, 2026 (SCI-Expanded, Scopus) identifier identifier

  • Yayın Türü: Makale / Kısa Makale
  • Cilt numarası: 163
  • Basım Tarihi: 2026
  • Doi Numarası: 10.1016/j.engappai.2025.113070
  • Dergi Adı: Engineering Applications of Artificial Intelligence
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Compendex, INSPEC
  • Anahtar Kelimeler: And performance assessment, Anomaly detection, Deep learning, Inspection, PV power plants, Thermal imaging, Unmanned aerial vehicle
  • Gazi Üniversitesi Adresli: Evet

Özet

In the context of the rapidly expanding global photovoltaic (PV) market and the critical need for efficient operation and maintenance (O&M) strategies, this review analyzes the synergistic integration of Unmanned Aerial Vehicle (UAV)-based thermal imaging (UAV-TI) with deep learning. This comprehensive study highlights the limitations of traditional, manual inspection methods and establishes the significance of a non-invasive, scalable, and automated approach. Utilizing a systematic methodology to select relevant literature, this paper critically examines the integrated workflow, encompassing UAV flight planning, thermal data acquisition, preprocessing techniques, feature extraction, and deep learning-driven anomaly detection and classification. The review consolidates existing knowledge and synthesizes findings from key case studies, demonstrating the practical efficacy of this hybrid approach in identifying a spectrum of critical PV system anomalies, such as hotspots, module defects, and shading effects. By identifying critical research gaps and discussing current challenges and limitations, this work provides a balanced perspective and offers guidance for future research directions aimed at enhancing the robustness, efficiency, and scalability of these integrated inspection systems.