🔔 Subscribe to ShipUniverse Weekly →
SHIPUNIVERSE / TECH • NOVEMBER 3, 2026 • MARKET REPORT
Windward's New AI Cameras Could Change How the World Tracks Ships
A ship doesn't vanish when its AIS goes silent. It only vanishes from one source of information. Windward's acquisition of Azimut.ai brings cameras into the same system as vessel tracks, ownership records, cargo intelligence and sanctions screening. That could prove valuable wherever a port, insurer or government needs to know which ship actually passed, not merely which ship reported a position.
The camera isn't the product. The verified event is.
Intelligence workflowAzimut says it can detect, classify and follow vessels through video, including estimating course and speed. Windward plans to pair those observations with its existing maritime data. The commercial value appears at the last two steps.
See the vessel
A shore or ship-mounted camera observes a hull, including one without a usable AIS signal.
Test the match
Compare appearance, time, location and motion with AIS, imagery, RF and other records.
Add context
Connect a credible identity to ownership, voyage history, cargo and compliance exposure.
Support action
Present a traceable finding for a human to clear, investigate, monitor or escalate.
What the pilot proves, and what it doesn't: Azimut reports 464 targets detected at Ashdod Port and observations as far as 7 km using daylight cameras. Those are vendor-reported pilot figures, not independently demonstrated global coverage or 464 confirmed vessels concealing their identity.
Who has a reason to pay?
Buyer comparisonThe strongest demand is likely to follow expensive decisions already made with incomplete vessel information.
| Buyer | Decision | Product that sells | What the contract depends on |
|---|---|---|---|
| Ports and terminals | Is an arrival who it claims to be? Was an unannounced vessel inside the approach? | Gate and anchorage monitoring, arrival alerts, camera integrations. | All-weather performance, integration, low operator workload. |
| Insurers and P&I clubs | Does a claimed voyage match independent observation? | Traceable sighting history, anomaly review, risk-screening tools. | Accurate identity, auditable evidence, retention and appeals. |
| Government and coast guards | Which uncooperative or unidentified vessel crossed a monitored area? | Secure sensor feeds, multi-source alerts, investigation systems. | Site coverage, cybersecurity, resilience and lawful collection. |
| Sanctions and trade compliance | Can a cargo or vessel declaration be independently checked? | Screening API, ownership links, review queues, documented findings. | Low false positives, explainability, clear escalation rules. |
Three places the money could accumulate
Revenue architectureAccess to the waterfront
Camera rights at ports, coastal facilities and chokepoints are scarce. Installation, connectivity, maintenance and permissions decide how quickly a network scales.
The history behind a sighting
A single observation may be worth little. A continuous, searchable record linked to vessel identity and risk context can support subscriptions and data feeds.
Alerts people can act on
Ports, insurers and compliance desks already have systems. The product sticks when evidence appears in the decision queue rather than another dashboard.
The bottleneck: a camera network is not just a software rollout. A site host must agree, power and connectivity must work, and somebody must maintain the sensor. A defensible business model also needs clear rules for data access, corrections and commercial reuse.
Data has deal value. The $1B claim still needs a test.
Financial perspectiveSpire sold its maritime data business to Kpler for approximately this amount before adjustments, plus a separate services arrangement. It demonstrates transaction appetite for maritime data, not the annual size of the camera-surveillance market.
First proprietary visual-sensing step
Windward announced the Azimut.ai acquisition and said it intends to build a visual intelligence layer at strategic passages. The purchase consideration and a funded global site rollout were not publicly specified in the announcement.
Change how ships operate
Hardware, onboard integration, safety assurance and operating permissions can make commercialization demanding. Customers pay to change vessel operations.
Improve how voyages are verified
Existing vessel activity becomes more observable. Customers pay for sightings, context and faster decisions without replacing a ship's navigation system.
Market thesis, not market forecast: a billion dollars means a hypothetical industry-wide annual spending pool across sensor operators, intelligence platforms and enterprise contracts. It is not Windward revenue, a disclosed addressable-market estimate or a prediction that surveillance will outperform autonomy.
Four limits that matter in procurement
Reality checksRange is not coverage
The reported 7 km daylight detection maximum is not guaranteed identity at 7 km. Fog, darkness, sea state, viewing angle and obstructions create gaps.
A match is not a verdict
Similar hulls, name changes and poor imagery can mislead. AIS may be switched off for permitted safety reasons; a gap does not prove wrongdoing.
The site is part of the deal
A reported Bornholm camera-hosting approach linked to Azimut's Varuna project illustrates why site consent and transparency matter. Windward says customers receive analytics, not raw video.
How much paid coverage would a billion-dollar market require?
Change the number of paying camera sites and separate enterprise contracts. These are illustrative industry-wide annual spending assumptions, not published vendor prices or customer counts.
of the test threshold
Camera-site contracts:
Separate enterprise deals:
Formula: (paying sites × annual site spend) + (non-overlapping enterprise contracts × annual fee). Does not include one-off equipment sales or prove customer demand, profit, market size or Windward revenue. Do not count a software license twice when it is bundled with a site contract.
We welcome your feedback, suggestions, corrections, and ideas for enhancements.
Please click here to get in touch