Helsing and its international partners producing an autonomous maritime underwater surveillance package – incorporating the Lura artificial intelligence (AI)-enabled advanced software platform, installed onboard the SG-1 Fathom uncrewed underwater vehicle (UUV) – are preparing to scale production of the Fathom gliders.
Such plans reflect early interest in the capability and offer capacity to support customer trials or initial orders, Naval News understands.
Scaling will now commence, with trials and testing of the capability package already underway.
The partners see the package as designed to be more affordable than traditional ways of conducting underwater tasks like anti-submarine warfare (ASW), critical undersea infrastructure security, or wide-area maritime surveillance, and underlining lessons being learned from the Russo-Ukraine conflict including delivering more scalable, affordable mass.
“We’re planning to manufacture SG-1 right here in the UK, initially in the 100s and eventually in the 1,000s,” Amelia Gould, Helsing’s General Manager Maritime, told the Combined Naval Event (CNE) conference in Farnborough, UK on 21 May.
“The threat is real, and the threat is now,” Gould added. Thinking differently about how to build mass quickly is central to the Lura/Fathom capability package. “Finding that needle in the haystack is becoming more challenging for humans. This is also coupled with increasing threats, not just to our infrastructure but to freedom of navigation for our goods,” Gould told the CNE audience. The Fathom/Lura package provides a spread of sensing capability for finding the ‘needle’, allowing higher-end, crewed platforms like a towed-array sonar-capable ASW frigate to conduct the fix, Gould continued.
Noting that Western navies’ operational demands stretch from the Baltic Sea to the Atlantic and Pacific oceans, Gould said “We needed to design something that could operate right across the world”: the Fathom/Lura combination “[provides] sensing at scale across any body of water”.
Fathom units can be operated by the partnership as a contracted service, or by navies directly.
The package – being developed by Helsing, through its partnership with Blue Ocean Marine Tech Systems, Ocean Infinity, and QinetiQ – was unveiled at His Majesty’s Naval Base Portsmouth, UK in early May, with a shore-side presentation of Fathom’s capability and Lura’s interface.
In a statement on 13 May, Helsing said it plans for the package to be available for deployment within the year. This encompasses any prospective procurement timeframes.
“We’re pretty much ready to go …. We’re now ready to deploy to sea,” Gould told a media briefing in London on 13 May.
“We’re also establishing the UK industrial base to manufacture Fathom and support this capability at scale,” Gould added. “We understand from the Russo-Ukraine conflict how important it is to have a resilient, simple, and scalable manufacturing solution to ‘flex’ with operational demand.”
The capability has been coming together for around 18 months. Initial AI benchmarking trials were conducted with the UK Royal Navy (RN), followed by further benchmarking activities with two other navies. In-water trials have been conducted off the UK and Australia. Partnering with several naval customers has helped shape the package’s development as a diverse, operationally relevant construct covering different oceanographic conditions and operational requirements.
The underwater battlespace remains complex and contested, with sub-surface platform detection being extremely challenging. NATO adversaries are operating more underwater platforms more regularly, while the alliance and its partners have limited numbers and coverage in response, particularly as current ASW systems remain complex and costly to build and operate.
“The need for a new response is urgent …. The problem needs new thinking, new solutions, and new technology approaches,” Gould told the media briefing.
Operational vision
“Our vision is to be able to search, deter, and defend [across] the world’s oceans with AI and autonomy at an affordable scale,” said Gould. “The idea is this becomes another feed into the recognised maritime picture.”
“The war in Ukraine has taught us that to truly understand the battlespace, you need sensors – lots of them,” Gould added.
The war has also underlined the need for simple-to-produce, affordable, portable, and attritable systems, Gould continued. “The idea is getting as many sensors in the water in the easiest way possible to get that sense of mass.”
The package’s hardware and software components are designed to be vertically integrated, combining to offer enhanced AI sensing and processing onboard ‘mass-deployable’ gliders.
Fathom is a 2m, 60 kg glider that, as a containerised system, can be rail-launched from shore or at sea, including autonomously. With design adjustments, it can be deployed from torpedo tubes or extra-large autonomous underwater vehicles. Travelling at 2-3 kt, Fathom adjusts its buoyancy to glide through the water column, carrying an in-built passive sonar for sensing. Endurance can reach three months, depending on mission requirements and configuration. Designed originally for oil-and-gas sector environmental monitoring, Fathom has been modified for use as a defence system. “We’ve adapted it for ASW to create a persistent, flexible constellation of underwater sensors,” said Gould.
Fathom’s hardware harnesses Lura’s AI-enhanced onboard processing software. The software uses a large acoustic model, and is trained on historical catalogues of diverse underwater acoustic data. “[It is] really important to have different data types for different regions to maximise Lura’s applicability across different ASW use cases,” said Gould.
With Fathom designed to be a simple but effective means for getting Lura to sea, the combination aims to improve coverage, volume, and speed in detection, analysis, and classification. “Most importantly for naval applications, unlike a human operator it can do it ‘24/7, 365’,” Gould continued, adding “One user can operate hundreds of gliders from a maritime headquarters.” Amplifying another lesson from Ukraine, massed sensing capability increases detection probability.
Central to Lura’s output is its onboard processing. “Lura can perform sonar processing activities when detecting, classifying, localising, and tracking,” Gould explained. “It can detect and classify contacts in real time, which gives navies a range advantage; that then gives them a decision advantage.”
In trials in April, Helsing ran simulation modelling surveillance across a 4,000 km2 area a simulated adversary was transiting, with the testing measuring different target paths set against different sensing densities. In the trials, Gould told the CNE conference, a single point of presence (like a surface vessel) listening acoustically across a wide underwater area generated a 40 percent chance of intercepting an adversary transiting that area; however, a constellation of autonomous mass involving around 200 smaller platforms, each with a smaller sensing area, achieved a 97 percent intercept probability for that same target.
Scalable capability
Helsing’s concept of operations (CONOPS) is for the gliders to be deployed as a massed constellation. This mass is scalable, from securing a harbour, to providing an ASW barrier across a choke point, to generating wide-area surveillance. The initial CONOPS is that the gliders will not communicate or co-ordinate with each other. This reduces hardware and software complexity. Crucially, however, it maintains the CONOPS’ core deterrent effect by reducing detection risk. The gliders surface to transmit data, with the onboard processing decreasing the data volume to be sent and restricting the exposure window. This process means that, while sensing and classification happen in real-time, data transfer to another platform or to headquarters happens in near-real-time.
The deterrence-based approach supports other CONOPS, like NATO’s ‘Digital Ocean’ vision, which aims to build integrated wide-area maritime surveillance. With massed gliders operating autonomously in a constellation, surveilling a body of water, and reporting in near-real-time, Gould told the media briefing “This is what we consider digitising the ocean to look like.”
Other lessons from Ukraine have shaped Helsing’s CONOPS. “With uncrewed aerial vehicles having communications degradation or jamming, the way we are seeking to overcome that is putting AI ‘on the edge’: what that means is you do the processing onboard the deployed vehicle,” Ned Baker, Helsing UK’s Managing Director, told the briefing. “Process, then understand: the ‘understand’ happens ‘on the edge’, in the ocean.”
Further learning from Ukraine includes the systems themselves adapting from missions. “Behind Lura and Fathom will be an ever-evolving mission cycle and a continuous feedback loop, meaning Lura will learn from every mission: every time it hears something new, it will iterate – meaning it will evolve at the pace of the threat,” Gould explained.
This underlines the importance of partnering with industry, navies, and defence ministries, Baker added. “We develop capability alongside end users …. What we’ve seen in Ukraine is this is the new paradigm of deploying defence capability: no longer will it be static.”
Working with stakeholders – especially operators – may also generate other capability evolution ideas. This is enabled by the capability package being software-defined, allowing the system, capability, and CONOPS to be iterated.
A final lesson, learned again from Ukraine, is Western navies’ need to trust in uncrewed technologies. The Fathom/Lura package feeds into this learning and trusting in two ways. First, its software-based, iterative design – using an established, dual-use glider – means it can be deployed at sea and linked into maritime command headquarters via standard tactical datalinks, sitting alongside existing systems and processes. Second, the intended ease of at-sea operation, plus the scalable deployment model and iterative learning approach, will help build user time, and thus trust: “The more time it spends at sea, the smarter it gets, the better it gets, and the more confidence you build in the things it’s trying to detect,” said Gould. “The sooner navies get it out there, the sooner it learns.”
