Skip to content
The Platform

From murky footage to actionable inspection data

The full pipeline: footage ingest, frame-by-frame detection, anomaly classification, and compliance-formatted report assembly. Automated end to end, with surveyor review built in at the sign-off stage.

8.2 hrs footage ingested
14 anomalies detected
3 hrs 47 min to draft report
DNV-ST-0126 classified
Pipeline overview

Four stages, one continuous workflow

Each stage feeds directly into the next. No manual data transfer, no re-entry of survey metadata.

STAGE 1 Footage Upload MP4 / AVI survey files + GPS + metadata STAGE 2 CV Analysis Frame-by-frame detection Confidence scoring STAGE 3 Surveyor Review Anomaly log confirmation Professional sign-off STAGE 4 Report Export PDF / structured data Compliance-formatted
Stage 2 in depth

How the detection models work

Detection models are trained on annotated ROV footage from offshore wind monopile, J-tube, and cable inspection programmes. Training data covers variable lighting conditions, turbidity levels, and fouling states seen in North Sea surveys.

Each frame is classified across six anomaly categories. A confidence score is assigned per detection, allowing the inspection engineer to prioritise review by severity and certainty.

GPS position, depth, timestamp, and frame number are embedded in every anomaly record, so findings map directly to asset location without manual cross-referencing.

See compliance output
Computer vision analysis running on ROV underwater survey footage
Integration

Fits the workflow you already have

No new survey equipment or specialist software required on the vessel.

Browser-based upload

Upload survey files directly from the vessel or from shore. Standard video formats accepted. Batch upload for multi-asset surveys.

API for pipeline integration

REST API for operators who want automated footage transfer from existing data management systems. Webhook notifications on completion.

Multi-user project workspace

Operators, surveyors, and asset managers each have role-based access. Review tasks assigned, comment threads retained for audit trail.

Structured data export

Reports export to PDF and JSON. Anomaly log exports to CSV for ingestion into maintenance management and condition monitoring systems.

Cross-survey trend analysis

Anomaly records from multiple surveys build a structural condition timeline. Detection models flag progression on known findings automatically.

UK data residency

All footage and reports stored in UK data centres. SOC 2 Type II assessment is planned for 2026. UK GDPR-compliant data handling and configurable retention periods per project.

See the platform on your own footage

The early-access programme is open to offshore wind and marine infrastructure operators. We process a pilot survey at no cost so your team can evaluate detection quality on your own assets.