Drone and aerial asset inspection
Computer vision reviews drone and helicopter footage of lines, towers, track and structures.
The problem
Inspection footage piles up faster than engineers can watch it.
Data we need
- Inspection imagery and video
- Defect taxonomy
- Asset locations
How it works
- Agree defect classes with engineers
- Label historical footage
- Train detection and severity models
- Link findings to asset records
- Engineers confirm before work is raised
What it changes
Targets are agreed with you before we build, and measured on your data before release.
- All footage reviewed, not a sample
- Defects located on the map
- Work raised directly from findings
Computer vision
Detection, segmentation and measurement from cameras, drones, satellites and scanners.
- Object detection
- Segmentation
- OCR