10 September 2026

^{Image: Oshen}

 

In September 2025, a boat roughly the length of a child sailed into one of the most violent places on Earth.

Hurricane Humberto had intensified into a Category 5 storm north of the Caribbean. Waves rose around the small vessel as wind gusts exceeded 150 miles per hour. Its wind sensor was eventually destroyed, but the boat continued transmitting.

The vessel was not crewed. It was a 1.2-metre C-Star, an autonomous ocean robot about four feet long, built by British startup Oshen. As it passed through Humberto’s eyewall and reached the edge of the eye, it sent measurements from the surface of the ocean to scientists every two minutes.

According to the US National Oceanic and Atmospheric Administration (NOAA), it became the first uncrewed surface vessel to capture and transmit data from inside a Category 5 hurricane. Two other C-Stars entered Humberto around 12 hours apart, encountering it while the storm was at Category 4 strength.

The C-Star is a compact, wind-propelled and solar-powered vessel that can navigate autonomously, communicate by satellite and operate as part of a swarm. If one robot or sensor fails, others can continue the mission.

Oshen must now prove that it can manufacture such robots in much larger numbers. It has the financial means to achieve this, having recently raised $5 million in a funding round led by Lunar Ventures, with participation from AlbionVC, Twin Track and Concept Ventures. The company says it has orders for another 100 C-Stars and plans to expand their use in weather forecasting, ocean science, infrastructure monitoring and maritime security.

 

Inside Oshen’s wind-powered C-Star

 

The C-Star looks like a bright yellow model yacht carrying a compact weather station.

It is only 1.2 metres long, making it small enough for one person to launch from a boat; in other words, it doesn't need the expensive cranes and equipment that larger autonomous vessels require.

The robot has a rigid wingsail that uses the wind for propulsion, while solar panels generate electricity for its sensors, onboard systems and communications equipment. A small electric thruster provides additional control rather than continuously driving the vessel across the ocean.

This combination allows a C-Star to remain at sea for months without carrying the battery capacity that a fully propeller-driven vessel would require. Oshen says its longest mission has lasted eight months, although endurance depends on the sensors, route and environmental conditions.

Its small size is central to the proposition. Conventional ocean-observation campaigns may require a crewed research vessel, specialist personnel and extensive planning. Such ships can carry sophisticated laboratories and many types of instrument, but they are expensive to operate and cannot remain across large areas of ocean indefinitely.

A small robot does not need to replace a research ship to change the economics. It can remain in one region after the ship has gone home, while multiple vessels spread out to measure conditions across hundreds of kilometres. Instead of concentrating every instrument on one highly capable platform, dozens or eventually hundreds of C-Stars could form a distributed swarm across the sea.

The principle resembles a sensor network on land, except that every station moves, the surface beneath it is never still and the weather can tear individual instruments apart.

The loss of the wind sensor inside Humberto illustrates the resilience of that approach. The failure prevented the C-Star from continuing to measure every variable, but the vessel remained operational and continued transmitting other observations. Two more C-Stars followed it into the storm, providing data at different times and locations.

 

Surviving storms without a crew

 

An autonomous ocean robot cannot depend on a single clever feature to survive a storm. Its hull, sail, electronics, sensors and software must continue functioning as one system while being thrown around by waves.

Oshen has disclosed the main components of the C-Star, including its wingsail, solar panels and thruster, but not the full details of its storm-survival design. Its performance inside Humberto demonstrates that the vessel can endure extreme conditions. It does not reveal precisely how the hull behaves when rolled by a breaking wave or how the control system responds when individual components fail.

The C-Star’s low-power architecture helps limit its dependence on energy-hungry mechanical systems. Its rigid wingsail provides propulsion without a conventional fabric sail, while solar panels replenish the electrical supply. Batteries keep essential systems operating when sunlight is limited.

Every component must also withstand a substance that steadily attacks almost everything around it. Salt water corrodes metals, penetrates imperfect seals and leaves deposits on sensors and solar panels. Repeated impacts impose mechanical stress, while marine growth can gradually alter the performance of a hull or measuring instrument.

Then there are hazards no navigation algorithm can neatly eliminate: drifting debris, ships and fishing gear. Encounters with marine life are also possible, although Oshen has not published detailed information about how often its robots experience these problems.

Small robots cannot carry the heavy redundancy of a large ship. Oshen’s answer is partly based on the resilience of the swarm. When multiple C-Stars are distributed across an area, damage to one vessel need not end the mission or eliminate every source of data.

Traditional ocean observation often depends on a small number of valuable platforms that must be protected and recovered. Oshen is pursuing a different model: make each robot compact and affordable enough to manufacture in large numbers, then use the swarm to provide coverage and redundancy.

 

Satellite communication and autonomous navigation
 

The C-Star uses satellite communications to send measurements from beyond the range of coastal mobile networks. During Hurricane Humberto, the vessels transmitted near-real-time observations that were made available to NOAA’s National Hurricane Center and the international Global Telecommunications System used by weather agencies.

Their ability to continue reporting every two minutes, even after one wind sensor was damaged, showed that the remaining onboard instruments and communications system could keep operating in extreme conditions.

Satellite connectivity does not turn an ocean robot into a continuously remote-controlled boat. Bandwidth and onboard energy are limited, and communication may not always be available.

The robot must therefore retain enough autonomy to continue its mission without constant instructions from operators on land. Onboard control systems manage its navigation and power use. When a satellite connection is available, the vessel can report its position and deliver compact packages of sensor data.

Images and video present a different problem. A weather reading consists of a relatively small amount of data; continuous high-resolution footage does not. NOAA says the C-Stars can transmit photographs during their missions, while video is generally downloaded after the vessels have been recovered.

This is autonomy in a practical rather than cinematic sense. The robot does not need to understand the ocean as a human sailor would. It needs to know where it is, remain within its operational limits and continue collecting useful measurements when direct supervision is unavailable.

 

Filling gaps in hurricane and ocean data

 

Satellites provide an extraordinary view of storms, but they cannot measure everything happening at the boundary between ocean and atmosphere. This is where heat, moisture and momentum pass between warm water and the air above it, influencing how hurricanes develop.

Aircraft can fly through a storm and release instruments into it. Buoys, gliders and autonomous floats provide longer-term ocean observations. Each platform captures a different part of the system, and large areas of the ocean remain sparsely sampled.

Autonomous surface vessels such as the C-Star can help fill the spaces between them, measuring wind speed and direction, sea-surface temperature, air temperature, atmospheric pressure and relative humidity.

One vessel recorded a minimum pressure of 955 millibars inside Humberto. Together with an increase in measured sunlight, the low pressure helped confirm that the C-Star had passed through the eyewall and reached the edge of the eye. The National Hurricane Center referred to the C-Star observations in an official forecast discussion.

Over time, the data can also be used to evaluate how well computer models represent the exchange of energy between the sea and atmosphere. Measurements taken at different times and locations may reveal how quickly a storm is changing, including periods of rapid intensification.

It's important to point out, however, that better measurements do not automatically produce a perfect forecast. Scientists must assess sensor accuracy, account for the vessel’s movement and determine how observations should be incorporated into forecasting models.

The Humberto deployment, specifically, involved seven C-Stars across NOAA’s 2025 hurricane programme, making it an operational demonstration rather than a new global observation network.

It nevertheless showed that a small platform could reach a part of a hurricane where direct surface observations are scarce—and continue reporting after the weather had begun damaging its instruments.

 

Scaling production from 15 robots to 15 every six weeks

 

Oshen was founded in 2022 by aeronautical engineer Anahita Laverack and electrical engineer Ciaran Dowds. The company produced 15 C-Stars during its first three years. It now says it can build the same number every six weeks/

The new funding will support not only this but also the development of passive acoustic systems for detecting ships and monitoring subsea infrastructure. This is relevant to Oshen's customers and partners, which include NOAA, the UK Met Office, Météo-France, Sweden’s Voice of the Ocean Foundation and the US and Royal navies.

One particularly ambitious scientific deployment is expected to send 50 C-Stars into the North Atlantic Subpolar Gyre. Oshen says the deployment forms part of a £2 million project backed by the UK’s Advanced Research and Invention Agency. It is intended to contribute observations to efforts to detect early warning signals of potentially disruptive changes in the North Atlantic climate system.

It will also help Oshen to prove that hundreds of robots can produce consistent, trustworthy ocean data over such extensive missions and that those observations offer customers something they cannot obtain as efficiently from buoys, drifting floats, larger autonomous vessels, or occasional research expeditions.

 

 

Liked this article? You can support our independent journalism via our page on Buy Me a Coffee. It helps keep MoveTheNeedle.news focused on depth, not clicks.

👉 https://buymeacoffee.com/movetheneedle.news