Join

Artificial intelligence for Iran

Azam Karami
Azam Karami
Automatic Fault Detection in Power Transmission Lines
About:

We have created a new and unique user-friendly web-based software called “AFTL” using deep learning techniques to detect over 80 types of defects in power transmission lines, including electrical, mechanical, and foundation issues such as damaged or broken insulators, shortage of bolts and nuts, corrosion, and rust. This is a national project with a Technology Readiness Level (TRL) of 8 and was chosen as the top project of Tavanir in 2020, 2022, and 2023. Tavanir is the main power generation, transmission, and distribution company in Iran. High-resolution RGB images were captured by UAVs (Drones) from power transmission lines for this project. The training and test datasets for the software were created by annotating more than six million images from faulty captures over a span of seven years using a specialized software called “Vira Label.” This software helps experts create labels around faults in UAV images and check them quickly. We also developed an end-to-end and user-friendly software, where users only need to upload the UAV images, and the report of faults can be easily downloaded in different formats such as PDF, Excel, and GIS. Technicians can use the results in the report to perform other analyses, such as estimating the expected average lifetime of components, and it also shows the exact geographic locations of faults in GIS format.

Target audience:
Electric Companies
Status: Ongoing project
Interested in cooperation with:
Worldwide
Presentation
Download
One second, we are looking for what you need