Our teams have been hard at work over the last few months, releasing the latest generation of our Inspect Edge system, improving our Cloud application, and releasing new and improved AI models. Read more in our latest release notes!
Inspect Edge Gen 4
We're excited to announce the latest generation of our vehicle-mounted camera + AI system, Inspect Edge Gen 4! Read below for its core features.
New: Left-facing high-res camera: Added a second, high-resolution camera to enable image capture on the left side of the road, eliminating the need to drive in both directions to capture all utility poles.
New: Remote web application: A new web application that allows users to control Inspect Edge directly from a web browser on any device has been released.
Improved Over-the-air (OTA) data upload: We enhanced our OTA upload capability so you can see your images sooner in our Cloud platform.
Inspect Cloud
New: Custom tags: Users can now apply custom, searchable tags to assets.
New: Lighting asset fixture type filter: By filtering by fixture type, users can bucket their lighting assets with more granularity.
New: Asset-by-asset navigation: Users can now quickly review details for any number of assets by navigating between them without returning to the map or list views.
New: Asset summary export: Quickly export any number of assets to a CSV file, including a summary of their pole-top components and associated defects.
Improved image enhancement algorithms: Select images are automatically brightened in post-processing for better viewing.
Improved: Additional base-map styles: More options for base-map styles are now available in Cloud's map view.
Improved: State management: The map view keeps your place even after viewing assets off-map.
AI Model Development
New: Component asset matching: Automatically match each component (e.g., transformer, bypass switch, etc.) to its respective asset, even when several utility poles are in the same image.
New: Cutout state detection: Our new model identifies open and closed cutouts, helping recognize idle transformers that can be repurposed elsewhere.
New: Pole lean measurement (beta): This new model measures how much a pole is leaning, helping prioritize ones that need repair. If you’re interested in validating this new model with us, reach out to our team at [email protected]!
Improved: Asset detection: Our improved on-device AI model now detects utility poles and streetlights with 94% accuracy.
Improved: Component detection: The latest version of our pole-top inventory AI now detects 15 components with an average accuracy of 85%, including streetlight and area light fixtures, streetlight and area light poles, electric distribution poles, double poles, cross-arms, cross-arm braces, transformers, pin and strain insulators, arresters, bypass switches, cutouts, conduits, telecom spaces, and foreign objects.