Avangrid

Background

As a major U.S. energy company operating extensive electricity networks across states including New York, Connecticut, and Maine, Avangrid is responsible for maintaining a diverse asset base exposed to varying environmental conditions. Transmission structures, insulators, and associated components are subject to gradual wear and corrosion, which can affect reliability if not identified and addressed early.

Avangrid was looking to inspect and review deficiencies in a much more efficient way, to gain a clearer, data-driven view of infrastructure conditions, Avangrid initiated a project leveraging Arkion’s AI-powered asset analytics platform to convert drone inspection data into actionable insights.

The challenge

Maintaining transmission assets at scale requires both accuracy and efficiency. Traditional inspection approaches are resource-intensive, rely heavily on manual review, and can result in inconsistent documentation across assets and regions.

Avangrid required a solution that could:

• Deliver objective and repeatable condition assessments across a large number of structures.
• Detect early indicators of component degradation.
• Apply severity grading to support maintenance prioritization.
• Produce structured outputs that could support maintenance planning and reporting workflows.

The Solution

Arkion’s platform was used to analyze drone-captured inspection including high-resolution imagery from Avangrid’s transmission network,. The data was processed using AI models trained to identify structures, components, and visible deficiencies, with a focus on condition-related findings. Using this data Arkion compiled reports on deficiencies such as flashover, damaged pole, loose bolts, and many others. Beyond this, Arkion also created a comprehensive inventory of the components of each power grid asset. 

The resulting analysis provided a detailed condition profile for each assessed structure. AI-generated findings were combined with human verification to ensure accuracy and consistency. Each identified deficiency was assigned a severity level, enabling Avangrid to distinguish between immediate maintenance needs and longer-term asset considerations.

Capabilities used

• AI-based deficiency detection
• Severity grading at component level
• Asset inventory outputs supporting condition analysis
• Human-in-the-loop verification for quality assurance

Key outcomes

Condition reporting - Standardized outputs supporting maintenance planning

Preventive maintenance - Earlier identification of degradation

Safety - Reduced reliance on manual inspections in the field

Overview and Oversight - All visual data unified and analyzed

Looking Ahead

Through this project, Avangrid established a clearer understanding of the condition of selected transmission assets, enabling more informed maintenance decisions based on observed asset health rather than assumptions. The work illustrates how AI-powered analytics can support utilities in extracting greater value from inspection data and strengthening preventive maintenance practices across transmission networks.

About Avangrid

Avangrid is a leading sustainable energy company in the United States, serving millions of customers through electric and natural gas utilities across the Northeast. The company operates regulated electricity networks and renewable energy assets, supporting the reliability and resilience of critical energy infrastructure while advancing the transition to a more sustainable energy system.

‍