Advanced Distribution Inspection Project

SPN: 3002019622

This is a distribution inspection research project. When it comes to inspections, we know that O&M spending is prioritized based on value to the customer and utility. While Distribution inspections could improve reliability, there are cases where the return does not justify the investment. However, we may be able to improve the ROI by using automated drones for data capture and then artificial intelligence for data review. This approach has promise, but we need to understand how it works and how it performs before we can confidently deploy a hands-off automated solution. Read more…

Dexter Lewis

Project Manager

Overview

Introduction to the advanced distribution inspection (ADI) supplemental project.


Supplemental Projects • UAS • Distribution

Research Updates

Ongoing research in advanced distribution inspection.


ADI • Advanced • Distribution • Inspection • Automation

ADI Results

ADI Results


Supplemental projects • Testing • Results

UAS Workshop

2021 Unmanned aircraft systems for electric utilities virtual workshop content


ADI • Advanced • Distribution • Inspection • Automation

Distribution Taxonomy

Advanced distribution inspection curated taxonomy

Propeller Collision Tests

Controlled collisions between a drone propeller and overhead assets


UAS • Distribution

“This is a distribution inspection research project. When it comes to inspections, we know that O&M spending is prioritized based on value to the customer and utility. While Distribution inspections could improve reliability, there are cases where the return does not justify the investment. However, we may be able to improve the ROI by using automated drones for data capture and then artificial intelligence for data review. This approach has promise, but we need to understand how it works and how it performs before we can confidently deploy a hands-off automated solution.

In the Advanced Distribution Inspection project, we plan to evaluate this new method. We’ll start in our lab where we can develop, test, and refine overhead distribution inspection requirements. Then, we will evaluate how well these specifications translate to the field. This field validation is important since the distribution environment poses some unique challenges. Then, we will label the field collected data to train and evaluate AI models.

The results from this work may help us answer questions around advanced inspections, including: How do speeds and costs compare to traditional inspections? Are there scenarios where drones and AI struggled? How accurate is AI compared to manual review?

You may have experiences in automated inspections already. If so, you may recognize the need for confidence in your data collection method, and the need for automated image review with AI. At the end of the day, there is likely not a solution that justifies an inspection for every mile of distribution. However, with the results of this research, you can make a data-driven decision, with confidence, that you’re applying the technology where it makes the most sense. And as a side benefit, whenever you’re asked about “drones, automation, or AI” you can reference this project.

This research leverages EPRI’s previous experiences and expertise. Many of you have participated in projects that used our laboratories as a controlled testing environment and have seen that we can validate hypotheses quickly, even up to full-scale structure testing. EPRI has a history of safe and compliant drone operations in the field. Our experts have produced visual inspection guidance in other research programs such as transmission that we can leverage. EPRI has experience evaluating multiple AI solutions and engaging that industry. Based on all those previous successes, we are confident that we have the capabilities to execute this project, and are excited to get started!”Dexter Lewis