Power system operators are increasingly adopting digital inspection methods to improve visibility of infrastructure conditions. However, manual review of large volumes of images and 3D data remains costly, time-consuming, and difficult to scale.

‍

This paper evaluates whether AI-powered asset analytics combined with human verification can reliably detect powerline deficiencies at scale while reducing human workload and cost. Empirical results are presented from a 15,000 km inspection of Sweden’s medium and low voltage distribution grid, comparing three approaches: conventional helicopter patrols, manual review of digital imagery, and AI-assisted digital inspection, where humans review only AI-flagged structures. Digital inspections were performed using Arkion’s AI-powered defect-detection platform.