The case for
infrastructure-independent autonomy.
Autonomous platforms today inherit a structural dependency on external signals and connectivity that cannot be assumed in contested environments. Clausal exists to remove that dependency.
Modern platforms are
built on fragile assumptions.
Nearly every autonomous platform operating today — aerial, ground, maritime — depends on GNSS for position, timing, and orientation. GNSS was designed as a peacetime infrastructure. It was not designed to survive deliberate denial, jamming, or spoofing at scale.
When that signal disappears, most platforms do not degrade gracefully. They stop navigating accurately. They lose the position integrity that every downstream function depends on: routing, object classification, coordination, return. The platform is operational. The mission is not.
Connectivity dependency compounds this. Systems architected around cloud inference, remote command, and network-based coordination fail in the same environments where GNSS fails. The attack surface is not a single signal. It is the entire infrastructure layer these platforms assume will be available.
This dependency is not unique to defense. Energy infrastructure, emergency response systems, and critical logistics networks carry the same structural exposure: they are built on GNSS and connectivity assumptions that were made in a more stable world. The consequences of failure are different, but the underlying problem is the same. Clausal's focus is where the consequences are sharpest, but the engineering applies wherever the infrastructure cannot be guaranteed.
External infrastructure — GNSS, cellular, satellite link, cloud compute — will be available when the platform needs it.
In contested environments, that assumption is precisely what an adversary will invalidate. Degraded and denied conditions are not edge cases. They are the operational baseline.
What changed in the
European security environment.
For decades, European defense operated within a framework of deep security dependency on the United States and a procurement model built on the assumption that advanced capability — sensors, software, AI systems, and the infrastructure behind them — would be available from allied markets when needed. The embedded compute and AI that powers those systems was largely developed outside Europe, trained on infrastructure outside Europe, and supported by supply chains outside Europe.
That model no longer holds. European nations are rebuilding organic defense capability at a pace not seen in a generation. The demand is not simply for more platforms. It is for platforms that can be produced at scale, integrated rapidly, and that operate without reliance on external supply chains or infrastructure that could be disrupted, restricted, or denied.
The requirement for infrastructure-independent autonomy — navigation, perception, and decision capability that functions without GNSS, without connectivity, without external compute — has moved from a long-term aspiration to an immediate operational need.
Model development, training infrastructure, and supply chains must remain within European jurisdiction and control.
Mass production economics require systems that run on microcontroller hardware, not high-power GPU platforms.
Platforms must navigate, perceive, and act in conditions where every external signal is actively contested or absent.
What we are
uniquely positioned to build.
Clausal was founded by Tatu Ylönen, who invented the SSH protocol in 1995 — the security standard that now protects the majority of the world's server infrastructure. SSH was built to solve a specific, high-stakes problem: how do you establish trust and protect communication when the channel cannot be assumed to be secure?
Embedded autonomy in denied environments is structurally analogous. The problem is not capability in ideal conditions. The problem is reliable function when the environment actively works against you. That framing — security engineering for adversarial conditions — is the foundation Clausal brings to autonomous systems.
The team combines that security lineage with direct operational experience. Our Head of Business Development served as a Duty Officer in a NATO peacekeeping Battle Group HQ Tactical Operations Center. We understand the environments our systems are designed to serve — not as an abstraction, but as a firsthand operational reality.
Finland gives us a practical advantage that cannot be replicated: we develop and validate in conditions that most programs treat as extreme edge cases. Snow-covered terrain, low sun angles, dense canopy, extended low-light — these are not test conditions we simulate. They are the environment outside our door.
Operationally, this matters because terrain-referenced navigation must work in winter. A system that performs in temperate conditions and degrades in snow is not a navigation system for northern European defense. It is a liability.
Our dedicated European compute infrastructure — 32× B200, 16× A100, 2,500+ CPU cores, 15+ PB storage — means model development, training, and iteration never leave European jurisdiction. Data sovereignty is not a compliance claim. It is a structural property of how we build.
Platforms that navigate, perceive,
and operate without reliance on contested infrastructure.
Infrastructure-independent autonomy means a platform that does not need to ask for its position. That does not require a signal to know where it is, what it is looking at, or what action to take. That runs the full navigation-perception-decision stack on hardware that costs tens of dollars and fits in the palm of a hand.
This is not a research objective. It is an engineering target with a clear commercial path: microcontroller hardware, mass production economics, and integration capability for the platforms that European defense programs are fielding now and over the next decade.
View our products Get in touchCapability within
a clear governance framework.
Clausal builds navigation and perception systems. What those systems do with that information — and who makes that decision — is a question we take seriously. Infrastructure-independent capability is not the same as unconstrained autonomy. The two are different engineering problems, and we treat them as such.
Clausal systems deliver navigation data and object classification to human operators. Decisions that carry consequence are made by people, not by the platform. Our architecture is designed to support that boundary, not obscure it.
All development operates under EU jurisdiction and is designed to be compatible with the EU AI Act's high-risk AI provisions. Our human-in-the-loop architecture aligns with NATO's own policy position on autonomous systems in lethal applications.
Ready to work
with us?
We work with platform manufacturers, system integrators, and defense innovation programs. If our approach matches your operational requirements, we'd like to hear about them.
Start a conversation