AI satellite identified cruise ship and booked its own second look

An Irish-built satellite identified a cruise ship in the Pacific and then scheduled its own follow-up observation two days later, without a single instruction from ground controllers. Ubotica, the Dublin space technology company behind the demonstration, claims the feat is a world first for commercial Earth observation.

The demonstration took place in October. Ubotica's CogniSAT-6 satellite captured an image of a cruise ship in the Pacific Ocean and used onboard artificial intelligence to establish the vessel's location. The satellite then calculated when it could revisit that spot and automatically scheduled a second image for two days later. That second observation confirmed the ship had departed, and the entire sequence ran without any human intervention from the ground.

For the maritime sector, the commercial logic is straightforward: faster, cheaper surveillance of vessels that prefer not to be seen. Ubotica points directly at shadow fleets, dark vessels operating with transponders switched off, and threats to subsea infrastructure such as cables and pipelines as the use cases that matter.

What does the demonstration actually change?

The October test does not amount to continuous vessel tracking, and Ubotica is careful not to claim it does. What it establishes is narrower but still significant: a satellite can now interpret what it sees, decide what to do next, and act on that decision independently. That removes the scheduling bottleneck that normally ties space-based observation to ground teams, tasking desks and turnaround times measured in hours or days.

"Autonomous scheduling is about giving satellites the ability to interpret what they see and decide what to do next," said David Rijlaarsdam, Ubotica's director of space system engineering.

The hardware behind the demo is modest. CogniSAT-6 is a 6U CubeSat, a satellite roughly the size of a large shoebox, built and operated by Open Cosmos and launched aboard a SpaceX rocket in March 2024. The intelligence sits onboard rather than on the ground, which is what allows the spacecraft to react in orbit instead of waiting for instructions.

Ubotica envisages the next step as a constellation-level capability. Multiple connected satellites would let one spacecraft identify suspicious maritime activity and automatically direct another to investigate, handing off the follow-up without any human in the loop.

Who is backing the technology?

The demonstration builds on a funding round SplashTech reported in June, when Ubotica raised $11 million to accelerate deployment of its Live Maritime Intelligence platform. The company is targeting governments and maritime authorities that need faster surveillance of vast ocean areas, a demand driven largely by sanctions enforcement and infrastructure protection.

For carriers, shipowners and insurers, the implications cut both ways. Autonomous observation could tighten the net around sanctioned tonnage and darken vessels, raising the compliance and reputational risks of chartering unknown counterparties. At the same time, legitimate operators stand to benefit from quicker verification of vessel movements and faster incident detection near critical infrastructure.

The technology also matters for underwriters and regulators watching subsea cables, where attribution of damage often depends on how quickly a suspect vessel can be re-imaged near the incident site. A satellite that repositions its own observation window within days, rather than waiting on a tasking queue, shortens that evidentiary gap considerably.

Ubotica's next milestones will be scaling from a single CubeSat to coordinated multi-satellite tasking, and converting an October proof of concept into a revenue-generating surveillance service for the maritime authorities now funding it.