Context
Utilities run inspection and maintenance programs across grid assets and related field operations, generating large volumes of visual inspection material that must be reviewed and translated into maintenance and operational priorities.
Challenge
Traditional asset performance management (APM) provides necessary information but lacks real-world context, slowing down risk detection.
AI Solution
Cloud-based platform for visual data, with ready-to-use deep learning models supporting multiple utility use cases, including asset inspection, vegetation management, storm response, and related network analytics. Visual Intelligence helps utilities turn large volumes of inspection data into actionable insight to detect risks earlier and better prioritize maintenance and operations.
Impact & evidence
Multiple benefits have been identified such as : vegetation budget decrease, outages decrease, extension of lifespan, increased length of inspected asset per year.
Implementation / Adoption
Visual Intelligence is a SaaS (Software as a Service) solution and convenient to deploy. Accurate GIS and high quality datasets either Lidar or Photos are required to train AI for asset inspection use cases and configure the platform. POC (Proof of Concept) can be developed within a few weeks while the full deployment takes ~3 months.
Data
The solution relies on satellite, drone, VIS/IR camera and LIDAR datasets.
Future enhancements
Planned enhancements: Improve predictive models using historical data; enhance the planning module to enable customers to combine multiple data sources and optimize their operations; improve the integration of 3D representation for asset inspection.
Additional information
https://www.gevernova.com/software/products/gridos/visual-intelligence