Tackling Vegetation and Asset Failures with Technology

Power outages are an unfortunate reality for utilities. Two of the most prevalent causes? Vegetation and asset failures. In fact, in a study done by Hoosier Energy, together these two factors caused the majority (54%) of power outages. Unsurprisingly, power sector players are looking for ways to tackle vegetation and asset-related failures. This article explores how applying artificial intelligence and machine learning can address the issue.

In an IDC study, downtimes and outages cost over a third of respondents $100,000 or more per outage in additional costs. That’s spending that goes beyond lost revenue, damaged customer service scores, and regulatory issues. So, of course, this is a major area of concern for utilities.

Upgrading processes around vegetation management and asset inspection can have a real impact by:

• Replacing expensive legacy survey methods (land-based line patrol)
• Addressing variation and integrity issues in asset-related data
• Optimizing asset risk management programs to move from time-based to condition-based
• Adding predictive analytics to benefit maintenance planning and field crew scheduling
• Improving the quality of the system of record

Transforming Vegetation Management and Asset Inspection at Scale, Digitally

Over the past decade, vegetation management and asset inspection capabilities have grown dramatically. Enabling technology capturing visual data and processing that data at scale is transforming decision making. Drones, helicopters, LiDAR, and various new file formats are helping to streamline end-to-end visual inspection data management.

With more ways to inspect the asset, comes more data. For humans, the amount of input is overwhelming.

Taking field, aerial and other visual inspection data, GE Digital Grid Analytics can ingest, review, and quickly assess risk. The data management platform’s AI-engine also supports data mining, image recognition, and predictive modeling. A human clicking through each data source one by one faces a challenging task. But, with technology processing the many data points, looking for potential risks, the utility can efficiently generate insights that can be turned into action. Further, the AI-enabled platform can detect what it is looking at and recognize defects in asset component parts. Or, running a change detection program, the utility can see what  is missing or needs repair. This saves on maintenance costs as the crew goes out to the field armed with awareness of what to expect and how to address it.

Step One: Leverage Data Already There

Consider the legacy approach: the utility might have a jpeg image of a broken insulator on a hard drive or server. That file number 6,321 shows there is a problem. But the maintenance planner does not know where the asset is located and what is needed to fix the problem. Without contextualization, prioritization and action are challenging.

Instead of being daunted by the rapid growth of data and the need for faster consumption, consider what data assets already exist. The utility likely already has operational data from sources including:

• Industrial Control Systems
• Asset Management
• Distribution Management
• Industrial Internet of Things
• SCADA
• Process Data Historians
• Field Devices
• Laboratory/Quality Systems

The AI and ML-enabled model doesn’t require utilities to reinvent the wheel. Instead, a cloud-based platform provides a framework for inputting, integrating, processing, optimizing, contextualizing, and analyzing data already in the enterprise as well as the data coming in from new technology. With greater transparency into the entire network, the utility can do better hotspot detection, land surveying, emergency response, and network inventory too.

All of this supports improved safety and reliability. The technology can reduce the number of times staff have to risk hazardous conditions and improve the overall inspection process. Plus, with predictive analytics, people are given an improved decision-making framework.

Getting Better and Smarter with Digital

AI can provide value quickly; however, a key starting point is determining the maturity level of your data capture. Then, in partnership with a strategic partner like GE Digital Grid, we’ll need to understand how the utility's’ data is optimized to drive different value streams and what strategies the utility has in place to process the data and baseline the AI.

Machine learning technology is built to always be learning and improving. This is a journey. GE Digital Grid can help. Our team is industry recognized. GE Digital won a 2021 Frost & Sullivan Best Practices Award for Global Transmission & Distribution Grid Analytics Product Leadership. This award recognized our investment in “supporting customers with industry leading solutions to accelerate the Energy Transition while making the electric grid more secure and resilient.”

For more information about the value of a visual intelligence platform, view this on-demand webinar with GE Digital’s Jett Winter, Digital Transformation Executive, and Sylvain Mandrau, Visual Intelligence Product Manager along with Kevin Prouty Group VP of Energy and Manufacturing Insights at IDC. Or if you’re ready to find how a visual intelligence platform can help you utility, please contact us now.