Autonomous Site Inspection of Power Transmission Line Insulators with Unmanned Aerial Vehicle System

Abstract

The inspection of overhead power transmission line and assets is an essential aspect to improve the overhead power transmission efficiency and to ensure an uninterrupted power supply. This article has mainly focused on the progression and fabrication of indigenous quadcopter/Unmanned Aerial Vehicle (UAV) for carrying autonomous operations in a coordinated movement along the overhead transmission towers for capturing the images and videos of transmission insulators and assets. A custom based dataset of power line insulators is created by using the quadcopter for overcoming the data scarcity and to perform Deep Learning (DL) assessment for (i) inadequate data for training and (ii) power line insulator detection and faults. The experimental results showcase that, the suggested DL architecture identifies power line insulators and associated faults, such as cracks, broken disk and missing top caps etc. With a detection speed of 56.8 frames/sec and an accuracy of 94.1%, the proposed DL technique has much promise for intelligent examination of power grid insulators. Ecological Footprint assessment of different power line inspection methods are also examined in this study.

Keywords:

• quadcopter/drone
• power line dataset
• deep learning
• fault detection
• you only look once (YOLO)

ACKNOWLEDGEMENT

The corresponding author would like to thank Motilal Nehru National Institute of Technology Allahabad, Prayagraj, India for providing lab facilities to carry out this research work.

Authors Contribution

MD. Faiyaz Ahmed: Visualization, Conceptualization, Writing - original draft, Methodology. J. C. Mohanta: Investigation, Supervision.

Disclosure Statement

The authors declare that there are no conflicts of interest.

Data Availability Statement

The experimental code implemented in this article is available with the corresponding author upon request.

Additional information

Notes on contributors

MD. Faiyaz Ahmed

MD. Faiyaz Ahmed received the M.Tech. degree in Automation from Jawaharlal Nehru Technological University, Hyderabad (JNT UH), India, in 2018 and Ph.D. specialization in Robotics from Motilal Nehru National Institute of Technology Allahabad, Prayagraj, India in 2023. Currently he is working as an Assistant Professor in the Department of Robotics and Automation at Vignan’s Foundation for Science, Technology & Research, Guntur, India. His current research interests include Unmanned Aerial Vehicles (UAV), Planning and Navigation, Mobile Robots, SLAM and Additive Manufacturing. Email: [email protected].

J. C. Mohanta

J. C. Mohanta is presently working as Associate Professor in Department of Mechanical Engineering, Motilal Nehru National Institute of Technology Allahabad, Prayagraj, India. Initially, he served as Engineering Officer in Mechanical Engineering Division of Central Power Research Institute, Bangalore. He has 15 years of research and teaching experience. He is a member of The Institution of Engineers (India) and many national and international societies. Presently he is engaged in research work focusing on navigational behaviour of Mobile Robots and Unmanned Aerial Vehicles (UAV) using various AI techniques. Corresponding author. Email: [email protected].