Streamlining Asset Inspections with GridOS® Visual Intelligence

In past blogs about GridOS Visual Intelligence, we’ve covered how utilities can use it to assess and address damages caused by severe storms. Today I wanted to discuss a resiliency challenge that is always a threat, rain or shine. That is asset health.

When I say “health,” I mean the physical condition and performance of key grid assets, like transformers, lines, poles, and the like. That is a major concern for utilities around the world.

And rightfully so. It’s no secret that grid assets are aging well past their prime. Some are already past their estimated lifespans and are operating on borrowed time. Amplifying the problem is the fact that once an asset fails, getting a replacement is a major undertaking. For example, an article by Power Magazine indicated that lead times for large transformers can be as long as four years from ordering, and the prices are skyrocketing in parallel due to ballooning costs of raw material like copper and steel.

To get the most out of their existing assets, utilities must unlock proactive maintenance, or the ability to identify and eliminate asset issues before they can cause a full-on power grid disruption. And the best way to do so is with the right technology.

With that, let’s cover the fourth Visual Intelligence use case: asset inspection.

As you may recall from past blogs, Visual Intelligence works by overlaying vegetation scans and other imagery against network maps. It then uses a simple color-coding system to flag any threats to resilience, while also automatically generating and dispatching detailed work orders for resolution to the relevant crews. This workflow is perhaps best known for its vegetation management use case, which identifies overgrown or unstable vegetation that poses a risk of damage.

In a testament to the versatility of Visual Intelligence, this workflow can also be leveraged to identify and rectify assets in need of maintenance. We engineered Visual Intelligence with AI-enabled anomaly detecting capabilities which recognize physical defects on key assets that might impact performance. Examples include rust and corrosion on metal components, sagging lines, frayed wires, and rotting utility poles, among others. Any detected defects are flagged and sent to the appropriate stakeholders for prioritization and tasking.

AI-enabled asset inspection with Visual Intelligence unlocks significant gains in efficiency and accuracy over traditional field inspections. As an added bonus, utilities can optimize their maintenance and capital expenditures by identifying physical defects long before they can snowball and cause a full asset failure – in other words, unlocking risk-based maintenance schedules.

Take a look at the below image, which shows a corroded transformer in one of a utility’s key network locations.

As part of the customer’s new risk-based asset maintenance schedule, Visual Intelligence identified the presence of corrosion on these transformers (note that they are outlined in red) and quickly created a work order to ensure they were promptly replaced. Had the utility still been following its traditional cycle-based maintenance plan, it likely wouldn’t have noticed the compromised transformer until it failed completely and caused an outage.

In one memorable success with Visual Intelligence’s asset inspection capabilities, a leading European utility faced the critical challenge of managing and maintaining over a quarter of a million kilometers of aging overhead lines. It hoped to avoid having to bury its entire network – a solution estimated to cost tens of billions of Euros. Line burying also risked making its overhead-line technicians’ knowledge obsolete, and it wanted to preserve and enhance that invaluable expertise. And in addition to reducing incident rates to improve grid reliability, the utility wanted to minimize its environmental impact by extending the lifespan of its existing overhead networks. This, it calculated, would help it avoid more than 100,000 tons of CO₂ equivalent compared to full renewal or burial. The utility found that GE Vernova’s GridOS Visual Intelligence platform could satisfy all its goals and requirements in managing and maintaining its existing line network.

The utility has implemented Visual Intelligence as a single, standardized technology tool for detecting anomalies that indicate equipment issues. This consolidation ensures consistent data interpretation and streamlined decision-making throughout all regions.

A key part of addressing the utility’s line management and maintenance needs is the continuous updating and accuracy of the MV overhead-line equipment database. To improve control and monitoring of the lines, Visual Intelligence leverages high-resolution imagery and AI-powered analytics to automate the utility’s equipment inspections and prioritize any necessary maintenance activities. This digital transformation has unlocked faster and more precise asset audits, contributing to safer, less costly, and more environmentally sustainable network operations.

Thanks to its advanced AI capabilities, Visual Intelligence has also streamlined the utility’s inspection and diagnosis processes and helped it realize significant cost savings by preventing unnecessary infrastructure replacements. Overall, Visual Intelligence has positioned the utility at the forefront of efficient, resilient, and cost-effective grid maintenance.

To learn more about asset inspections and other Visual Intelligence use cases, check out our solution paper on the subject, “Orchestrate the Grid with Visual Precision: GridOS Visual Intelligence.”