Drone Inspection Analytics demo
Weeks of image review turned into a ranked defect queue. 1:15
What the demo shows
A workflow solution for line inspectors that screens drone images and flags asset defects.
Six agents screen every flight overnight: they match each image to its structure, box each candidate defect against the defect catalog, find the same component in last year's flight and measure what changed, then rank the boxes so the growing and urgent ones come first. The inspector confirms or rejects each box in a gallery by structure, the asset engineer agrees every Urgent, and confirmed defects go to work planning as drafted work requests.
Transcript
- Drone Inspection Analytics screens drone images of power lines and flags defects for line inspectors.
- One flight of one line brings back more than 6,000 images.
- By hand, review takes weeks.
- Overnight, agents matched every image to its tower and boxed 118 possible defects.
- Ines works tower by tower.
- She confirms a corroded clamp and rejects rust that is only fresh paint.
- Every box is compared with the same component from last year's flight.
- On Tower 87, the insulator crack has doubled from 11 to 23 millimeters.
- The growth rule makes it urgent.
- Ines promotes it, the engineer agrees, and a work request is drafted.
- Reviewed in two days, not six weeks, with the growing defects first.
- Built on Botminds: agents, data, screens, and governance on one platform.
Transcribed from the film’s narration. Captions are on the player.