CASE STUDY 01
DVI
SHIPBUILDING & MARINE ENGINEERING
Identifying blocks and tracking ship assembly from drone imagery
One of the world's largest shipbuilders · two assembly docks · 2023 – present · operating infrastructure
CHALLENGE
- Ships are assembled from very large numbers of individual blocks, scattered across docks at a scale that defeats manual observation.
- Block tracking relied entirely on manual inspection: significant manpower, inconsistent accuracy, and assembly status that was often already out of date when reported.
- Inaccurate status propagated directly into production delays, because the plan was being rebuilt against a picture of the yard that was no longer true.
SOLUTION
- Dynamic Visual Intelligence deployed to identify ship blocks from drone imagery captured several times daily.
- A fully automated pipeline combining proprietary segmentation, mesh-reduction and multi-view feature matching.
- The key step is the use of the client's own CAD data to establish block identity, then compare observed location against the planned workflow — so a deviation raises a notification the same day.
RESULTS
- Errors and delays inherent in manual tracking effectively removed from the progress reporting chain.
- Block location and assembly progress became an observed quantity available to planning, rather than a compiled claim.
- Provides the observation feed that RTO re-plans against.