Loading...
Autonomous image analysis for drone-based power line inspection
Magodi, Patricia Lulu Tran, Hannah Rubio García, Fernando Das, Lokesh Manoharan, Arun
Magodi, Patricia Lulu
Tran, Hannah
Rubio García, Fernando
Das, Lokesh
Manoharan, Arun
An error occurred retrieving the object's statistics
Citations
Altmetric:
Authors
Magodi, Patricia Lulu
Tran, Hannah
Rubio García, Fernando
Das, Lokesh
Manoharan, Arun
Tran, Hannah
Rubio García, Fernando
Das, Lokesh
Manoharan, Arun
Other Names
Location
Time Period
Advisors
Dao, Mai
Original Date
Digitization Date
Issue Date
2026
Type
Abstract
Poster
Poster
Genre
Keywords
Subjects (LCSH)
Citation
Magodi, P. L., Tran, H., Rubio García, F., Das, L., Manoharan, A., & Dao, M. Autonomous image analysis for drone-based power line inspection. -- FYRE in STEM Showcase, 2026.
Abstract
Power line inspection is important for keeping the electrical grid safe and working properly. Traditional methods using human workers are expensive, time-consuming, and can be dangerous, especially when inspecting high or hard-to-reach areas. This project explores the use of drones equipped with sensors and cameras to measure power line sag and improve monitoring of power lines. Power line sag depends on factors such as tension in the wire, the distance between poles, and different conditions, which all affects how safely the lines operate. Our project uses drones to collect real-time aerial images of transmission towers, conductors, and insulators, which are analyzed using a YOLOv8 object detection model to identify defects such as damaged insulators and structural issues. The model was trained using the publicly available TTPLA dataset and evaluated using the propriety real-world inspection data provided by Evergy. Model performance on the training data, which constitute of a small fraction of the available data, was assessed using quantitative metrics including precision, recall, and mean Average Precision (mAP), demonstrating the effectiveness of the approach in detecting power line components and defects. From the initial findings, we conjecture that this also helps improve dynamic line rating by allowing better use of power lines based on real-time conditions. Overall, this project demonstrates the potential of how drones and machine learning can be used to make power line inspection easier and improve the performance of the electrical grid.
Table of Contents
Description
Poster and abstract presented at the FYRE in STEM Showcase, 2026.
Research project completed at the School of Computing, the Department of Electrical and Computer Engineering, and the Department of Mathematics, Statistics, and Physics, Wichita State University.
Research project completed at the School of Computing, the Department of Electrical and Computer Engineering, and the Department of Mathematics, Statistics, and Physics, Wichita State University.
Publisher
Wichita State University
Journal
Book Title
Series
FYRE in STEM 2026