RoboPV: A modular system for enhancing the efficiency of autonomous aerial monitoring of photovoltaic plants

Document Type : Original Article

Authors

1 Department of Electrical and Computer Engineering, Concordia University, Montreal, Canada

2 Department of Ocean Operations and Civil Engineering, Norwegian University of Science and Technology (NTNU), Ålesund, Norway; Department of Sustainable Systems Engineering (INATECH), Albert Ludwigs University of Freiburg, Freiburg, Germany

3 Department of Aerospace Engineering, Amirkabir University of Technology, Tehran, Iran

10.22060/aest.2025.5849

Abstract

This paper presents RoboPV, an innovative embedded software for autonomous aerial monitoring of photovoltaic (PV) plants. RoboPV automates monitoring with features like optimal trajectory planning, image processing, and real-time fault detection through four integrated components: boundary area detection, path planning, dynamic processing, and fault analysis. A specialized encoder-decoder deep learning model processes aerial images to identify plant boundaries, while a unique path planning algorithm ensures complete area coverage. During flights, a neural network analyzes images for automatic fault detection. RoboPV also includes decision-making algorithms for various flight conditions, is compatible with low-power micro-computers, and supports the MAVLink protocol for multi-rotor operations. A six-degrees-of-freedom dynamic model was tested in a SIMULINK environment, achieving 93% accuracy in autonomous inspections of large-scale PV installations.

Graphical Abstract

RoboPV: A modular system for enhancing the efficiency of autonomous aerial monitoring of photovoltaic plants

Keywords

Advances in Energy Sciences and Technologies
Volume 1, Issue 2
September 2025
Pages 216-227

Files

History

  • Receive Date: 23 July 2025
  • Revise Date: 14 August 2025
  • Accept Date: 18 August 2025
  • Publish Date: 01 September 2025

Share

How to cite

Statistics