Data Upload & AI Analysis

Upload captured imagery, validate metadata and run AI crack detection

Step 1

Upload dataset

Drag the drone or thermal image folder in, or select an inspection to attach it to.

Step 2

Validate metadata

Date, zone, device, flight height, operator and GPS are checked automatically.

Step 3

Run AI analysis

Crack detection and leakage-risk scoring produce a suspected finding list.

Step 4

Send to review

Findings move to the Crack Register for CAIC engineers to confirm or reject.

New upload

View AI findings
INS-2026-014 · Z-A1 · Drone RGB

Processing queue

Batch 03 — Zone B1/B2

INS-2026-014 · 1842 images · 38.4 GB

Analysis complete

95 suspected findings so far

Thermal set — Zone C1/C2

INS-2026-012 · 640 images · 9.2 GB

AI analysing

41 suspected findings so far

Pre-flight plan — Zone A1/C1

INS-2026-015 · 0 images · —

Awaiting upload

0 suspected findings so far

What the AI returns

Crack detection

Crack detection

Location, estimated width and length, confidence score and severity proposal.

Leakage risk

Leakage risk

Thermal anomalies consistent with trapped moisture under the coating.

Repair comparison

Repair comparison

Before / after evidence pairs used for CAIC verification.

Metadata validated on upload

  • Date, time and zone reference
  • Device, camera and flight height
  • Operator and capture method
  • Weather / environmental notes
  • Image reference and GPS coordinates

Thermal sets are best captured after rainfall, when moisture contrast is highest.