> ## Content Index > Fetch the complete content index at: https://unhyd.com/llms.txt > Use this file to discover other available public pages before exploring further. # Autonomous Maritime Teaming Moves From One Boat to Two - URL: https://unhyd.com/article/autonomous-maritime-teaming-shield-ai-kraken-k3-scout/ - Published: 2026-09-28T13:07:43.000Z - Updated: 2026-10-03T16:10:13.000Z - Description: Shield AI and Kraken tested a two-vessel autonomy workflow in Portsmouth. The real question is whether fleet software can make humans more effective without obscuring their control. - Author: Unhyd Editorial Staff - Tags: Technology, AI **Shield AI and Kraken Technology Group said on September 28 that they had demonstrated two K3 SCOUT uncrewed surface vessels working together under Shield AI's Hivemind mission-autonomy software.** The Portsmouth, U.K., exercise matters less as a test of one boat than as a test of whether software can coordinate a small fleet without turning each vessel into a separate workload for an operator. According to the companies' announcement, the two vessels carried out an intelligence, surveillance and reconnaissance mission. They searched a designated area using onboard radar, cameras and Automatic Identification System data; after identifying a vessel of interest, the software directed the two craft to track and escort it before replanning the search. Shield AI described this as its first maritime autonomy demonstration in the U.K. The central fact is a company-led demonstration, not an independently reported operational deployment. That distinction is important. The release establishes what Shield AI and Kraken say they tested in Portsmouth; it does not publish the test's reliability data, its conditions or an independent assessment of the result. Still, it offers a useful marker of where maritime autonomy is moving: from navigation and remote control of a single vessel toward coordination, task allocation and shared sensing across more than one craft. ## Why autonomous maritime teaming is different A vessel that can hold a route or avoid an obstacle is one problem. A group of vessels that can divide a search area, share sensor information, change their plan when a contact appears and remain within human authority is a larger systems-integration problem. The value proposition is not simply fewer people on board. It is whether operators can set an objective and supervise a coordinated mission without having to manually choreograph each movement. That is the role Shield AI assigns to Hivemind, which it calls an AI pilot. In the release, the company says the software can work within operator-defined parameters and established command authority. Kraken's own K3 SCOUT documentation likewise describes supervised autonomy and human-on-the-loop oversight. Those are design claims rather than a public account of how the Portsmouth exercise was governed, but they identify the operational question that will determine whether fleet autonomy is useful: where the system may act independently and where a person must decide. ## The platform is already being positioned for integration Kraken lists the K3 SCOUT as an 8.4-metre uncrewed surface vessel with a 600-kilogram payload capacity and a top speed of 55 knots. Its modular architecture and open interfaces are relevant here because multi-vessel autonomy has to sit above the boat's sensors, communications and mission payloads rather than replace them. A software layer that works only with one sensor package or one communications arrangement will struggle to become a fleet capability. There is also a live procurement context. In March, the Royal Navy announced a £12.3 million contract for 20 uncrewed boats from Kraken under Project Beehive. The Navy said the fleet would support operations, training and development as it builds experience combining crewed and uncrewed forces. In July, it reported an airdrop trial involving a Beehive-specification K3 SCOUT. Those developments do not validate Hivemind or make the Portsmouth exercise a Royal Navy deployment; they do show that the vehicle family is moving through real training and experimentation channels. ## What readers should watch next The companies say they plan further trials with U.K. and NATO partners over the coming year, including in contested environments. The important follow-up will not be another polished demonstration alone. It will be evidence about how the team behaves when communications degrade, weather changes, navigation data conflict or an operator intervenes. Public detail on those questions would make it easier to separate an autonomy demonstration from a repeatable operational capability. For now, the September 28 test is a narrowly evidenced step: two K3 SCOUTs and Hivemind were demonstrated together in an ISR scenario, according to the companies involved. The broader significance is the direction of travel. In maritime systems, the competitive challenge is increasingly not only building an uncrewed vessel, but proving that its software, sensors, people and partners can operate as one coordinated system. ## Sources - [Shield AI and Kraken Technology Group demonstration announcement](https://shield.ai/shield-ai-and-kraken-technology-group-demonstrate-hivemind-enabled-autonomous-maritime-teaming/?ref=unhyd.com) (September 28, 2026) - [Kraken Technology Group: K3 SCOUT platform](https://krakentechnology.com/platforms/k3-scout?ref=unhyd.com) - [Royal Navy: Project Beehive procurement announcement](https://www.royalnavy.mod.uk/news/2026/march/11/20260311-kraken-usv-procurement-announcement?ref=unhyd.com) (March 11, 2026) - [Royal Navy: K3 SCOUT airdrop trial](https://www.royalnavy.mod.uk/news/2026/july/08/20260708-uncrewed-vessels-airdropped-in-pioneering-trial-supported-by-royal-navy?ref=unhyd.com) (July 8, 2026)