Nodus Fortis · Sovereign Black Box
SECTION 01 / 06
01 / The Problem

Machines that believe whatever they are told

Every autonomous machine that moves — a delivery drone, a self-driving car, an airliner, a ship — trusts one thing completely: that the position it is handed is true. It gets that position from GPS, and it believes it without question.

GPS was never built to be secure. It is a faint signal broadcast from twenty thousand kilometres away — by the time it reaches the ground it is weaker than the background noise of a city. Anyone with a few hundred dollars of equipment can shout over it, or worse, imitate it. That used to be a curiosity. It is now a routine tool of modern conflict, spilling far beyond the battlefield into civilian aviation, shipping, and border regions on three continents.

In plain termsImagine your car's satnav quietly, confidently telling you you're one street over from where you really are — and you believing it the whole way. That is the problem. Not a machine that knows it's lost, but one that is sure it's fine while being completely wrong.
Jamming — drowning it outDenial
SIGNAL— healthy① signal healthy
What you're looking atThe attacker floods the airwaves with noise until the machine can no longer hear the satellites. It goes blind — but it knows it has gone blind. Watch the step label: healthy signal → noise floods in → receiver blind. Because the machine can tell it's lost, it can react honestly.
Spoofing — counterfeiting itDeception
REPORTEDon course ✓TRUE POSITION0 m off
What you're looking atThe attacker feeds the machine a position that looks perfectly real but is a lie. Nothing looks wrong, so no alarm sounds. The green marker (what the machine believes) stays calm on the line while the red marker (where it truly is) slides away. Confident, and completely wrong.

A jammed machine knows it is lost. A spoofed machine does not.

That is why spoofing is the harder, more sinister problem — and it is the problem Nodus Fortis exists to solve.

The principle: a machine should treat the position it is handed as a claim to be proven, not a fact to be trusted — and the instant that claim fails to hold up, reject it and protect the flight, before the lie can steer the machine anywhere.

An attacker can broadcast a false position — but the moment that false position is inconsistent with what the machine can otherwise establish for itself, the device catches the discrepancy and refuses the lie. The attack is defeated not by blocking the signal, but by catching it out. For the great majority of attacks — including the slow, stealthy ones that walk straight past conventional alarms — creating that inconsistency is unavoidable, so they are caught in seconds.

Our honesty is reserved for the hard edge, and it is a narrow one. A richly resourced attacker can, at great expense, craft a lie with no inconsistency to find. We do not pretend to make that impossible. What we do is force any serious adversary into exactly that corner — costly, conspicuous, nation-state-grade equipment — while the practical attacks, the ones an adversary can actually mount on a budget, we catch outright.

That is the product: catch the lie, and price the rest out of reach.

02 / The Threat Landscape

Every way to lie to a machine on the move

This is the full map of attacks on satellite navigation — plain enough for anyone to follow, and anchored to published research and real events that prove each one is not hypothetical. Each attack is shown plain-language first: the one-liner anyone gets, then an everyday analogy, then the detail for a technical reader.

The attacks fall into three families: drowning out the signal, counterfeiting it, and tampering from the inside.

Family 1 Drowning it out — denial

Broadband jamming

Loud

The bluntest attack: transmit raw noise on the satellite frequencies until the machine can no longer pick out the real signal. It goes blind — but it knows it.

blasting an air horn next to someone straining to hear a whispered direction.

The detail
ResearchPsiaki & Humphreys, GNSS Spoofing and Detection (Proceedings of the IEEE, 2016) — the foundational split between denial and deception.
What defines itThe ratio of jamming power to signal power — a contest of brute strength.
Real worldReported: sustained jamming around the Baltic and Kaliningrad disrupted roughly 123,000 flights between January and April 2025 — ordinary passenger aircraft.
Family 2 Counterfeiting it — deception · the dangerous family

Record-and-replay

Quiet

The simplest lie: record the genuine signal and play it back, delayed and amplified. The machine believes it is somewhere it was.

replaying yesterday's directions a few seconds late.

The detail
Also calledmeaconing.
ResearchPsiaki & Humphreys, 2016.
What defines itThe replay delay. Any delay creates a discrepancy; only a perfectly timed, perfectly placed replay avoids one.

The sudden jump

Quiet

A cruder counterfeit knocks the receiver onto a fake signal, so the position appears to teleport — here one moment, a block away the next.

a map pin that instantly leaps across town — impossible movement, and a giveaway if anything is watching.

The detail
ResearchHumphreys et al., Assessing the Spoofing Threat (ION GNSS 2008).
What defines itThe size of the jump, in metres.
Real worldReported: Europe's aviation regulator logged around 6,000 spoofing events in 2024 across the Eastern Mediterranean, some appearing as exactly these sudden errors.

The slow sideways drag

The attack we're built to beat
STEP① smooth takeoverOUTCOMEwatching…

Instead of a violent jump, the attacker takes over the signal smoothly and eases the machine's apparent position sideways — a few centimetres, then a few more — slowly enough that no crude alarm ever trips. Over a couple of minutes a machine can be walked hundreds of metres off course while staying serenely confident.

a satnav nudging you one lane over, then another, so gently you never notice you've been driven to the wrong place.

Why this is the heroLeft unprotected, a slow sideways drag can pull a drone nearly 200 metres off course with no alarm whatsoever. It is quiet, patient, and devastating — and it is precisely the attack the Black Box is designed to catch, and does, in validated simulation testing. See the measured replay of exactly this attack in Section 03.
The detail
ResearchHumphreys et al., 2008 (a counterfeit that "drags the tracking points off"); Tippenhauer et al., ACM CCS 2011 (the conditions for a seamless takeover).
What defines itThe walk rate — how fast the lie moves. Slower is stealthier, and harder to catch.
Real worldReported: smooth, large-scale position deception over the Eastern Mediterranean, and "circle spoofing" affecting ships near the Strait of Hormuz in 2025–26.

The slow forward-and-back nudge

Fast today · slow on roadmap

The same smooth takeover, but easing the machine forward or back along its own line of travel instead of sideways.

being told you're a little further down the road than you are — betraying almost nothing about your path.

The detail
Why it's hardA gentle nudge along the direction you're already going is the single hardest quiet attack in the catalogue.
StatusThe fast version is caught today; the slowest version is on our roadmap (Section 05).

The perfect shadow

The irreducible edge

The most sophisticated attack imaginable: hold the machine at a fixed false offset and slide it along at exactly the machine's true speed, so nothing about the motion ever looks inconsistent.

a shadow that follows you perfectly — always the same distance away, moving exactly as you do.

The detail
Honest boundaryThis is the hard theoretical limit. But it costs the attacker enormously — phased-array-grade coordinated equipment far beyond a hobbyist's radio. Forcing an adversary to that expense is itself a victory; detecting it is the job of a companion device on our roadmap.
ResearchPsiaki, O'Hanlon, Powell et al., Two-Antenna Differential Carrier Phase (ION GNSS+ 2014).

The full orchestra

Nation-state grade

A coordinated attack that counterfeits several of the machine's information sources at once, in perfect agreement, so no single check finds a contradiction.

a con where every witness has been paid to tell the same false story — so cross-checking them turns up nothing.

The detail
Our answerForce the attacker to corrupt everything at once and keep every lie consistent — a cost that compounds with each source they must fake.
ResearchPsiaki & Humphreys, 2016 — no single check covers every mode, which is why coordinated defence matters.
Family 3 Tampering from the inside — infiltration

The wiretap

Inside the machine

The attacker isn't in the airwaves at all, but inside the machine — on the wiring after the sensor, quietly adding a bias to the data as it flows.

someone editing the letters in the postbag before they're delivered.

The detail
ResearchFawzi, Tabuada & Diggavi (IEEE Transactions on Automatic Control, 2014) — a system becomes unrecoverable once more than half its sources are compromised.
Our lineWe handle the case where the attacker is on the navigation line; a deeper supply-chain attacker is handed off to trusted, signed-at-origin hardware — named honestly, not papered over.

The stolen key

Key custody

If an adversary steals the secret cryptographic keys, any scheme that relies on those keys is beaten by definition.

the strongest lock in the world is useless once someone copies the key.

The detail
Our lineThis is a key-management problem, and we say so plainly rather than pretending any navigation device makes it disappear.
ResearchSatchidanandan & Kumar, Dynamic Watermarking (Proceedings of the IEEE, 2017).

The bottom line: the Black Box beats the attacks that move a machine in ways that quietly contradict reality. It imposes crippling cost on the nation-state-grade attacks at the edge. And the handful that belong to a companion device or to key custody, it names clearly. A defender who can draw the exact line between what they stop and what they don't is a defender you can trust.

03 / The Black Box, Today

A drop-in guardian for the navigation path

A single compact circuit board — about the size of a business card — that installs inline on the navigation path, between the machine's GPS receiver and the controller that acts on it. Nothing else about the host machine has to change: it is a drop-in. From that position it continuously checks the navigation feed, and the instant the feed stops being trustworthy, it intervenes — fast enough to act before a lie can steer the machine.

What you're looking atA replay of the hero attack from Section 02, from Nodus Fortis's validated closed-loop simulation. Switch tabs to compare a clean flight, an unprotected drone driven ~196 m off course with no alarm, and the same attack caught and severed in seconds. Every figure shown is a measured simulation result — it is not a claim about the physical board, which is pending bench validation.

Where the technology stands now — Tier 1:

Detection logic
Validated in simulation
Hardware design
Complete & audited
Core invention
Patent-pending
Readiness
TRL 4
What TRL 4 meansTechnology Readiness Level 4 means the core idea is proven to work in a realistic, closed-loop simulation — tested against the actual autopilot logic used on real drones, not a toy model — but nothing has yet been proven on physical hardware. The design is finished and independently audited; the build is the next step.

Here is the honest, and genuinely strong, way to read that: the hard problem is already solved. The intelligence of the device — the difficult, novel, defensible part — works, and has been proven to work against realistic attacks. What remains is not open research; it is building the board and confirming on the bench what the design already shows. The invention is done. The manufacturing is next.

04 / The Vision

From a single box to a defensive platform

The Black Box is the first node of something larger — a sovereign trust layer for autonomous machines. Each capability below is a deliberate step on the roadmap, described by what it will do, never how. All of it is future tense.

What you're looking atSix capabilities the platform is designed to grow into — clearly labelled roadmap, not things the device does today. Read each as "what it will do, and why that matters."

Catch the attacker, not just the lie

Roadmap

Today the device catches deception. Tomorrow it will find the source. Anything that jams or spoofs has to transmit — and anything that transmits can be located. The shield becomes a hunter, turning the attack into the moment the attacker gives away their position.

Nearly every attack answered onboard

Roadmap

Each new tier will pull more of the threat landscape inside the line — until what remains is a small, named, and enormously expensive residue. A device against which almost every practical attack simply fails.

Kept alive by the attack itself

Roadmap

Designed to draw trace power from the very radio energy an attacker floods the area with. Scoped honestly: this sustains the trust layer and its tamper-proof record, not the machine's propulsion. A captured, powerless device can still keep an unforgeable log and erase its secrets the instant it is opened.

Strength in numbers

Roadmap

Machines carrying the device will share what they detect. One that spots a manipulated position warns the others and flags the compromised member, so a fleet routes around the deception instead of each machine falling for it in turn.

Worthless to steal

Roadmap

Designed to prove its own integrity and bind its most valuable contents to its own specific hardware, so a captured or cloned unit yields nothing usable — resilient even against future quantum computers. A security feature, and the company's own protection.

Every machine, every domain

Roadmap

The core idea — never let an unverifiable outside signal override what a machine can establish for itself — extends far beyond drones: to ground vehicles, crewed aircraft, ships, and ultimately satellites. One principle, protecting autonomy everywhere.

05 / The Roadmap

The honest phased path

What is done, what is being funded right now, and what lies ahead — at a glance. Green is complete; gold is the immediate, funded step; the rest is the road ahead.

  • Done · nowTier 1 · TRL 4

    The intelligence works

    Proven in simulation against real autopilot software; the full board designed and audited; the core method patent-pending. The hard invention is complete.

  • Funding nowImmediate milestone · target TRL 5

    Build & bench-test the board

    Fabricate, assemble, and bring the board to life on real hardware, and confirm on the bench what the design already demonstrates. Deliberately narrow: manufacturing and verification, not new research. Completing it moves every claim from "proven in simulation" to "measured on hardware."

  • AheadTier 2

    Widen the shield

    Extend the device — and a companion board — to bring the hardest attacks inside the line: the slow forward-and-back nudge and the perfect-shadow offset. Begin the work that lets the device locate the source of an attack — the first step from shield to hunter.

  • AheadTier 3

    From device to platform

    Introduce the trust fabric that lets many machines defend as one; the energy-aware persistence that keeps a captured unit's security alive; and hardened, production-scale protection of the device's own contents. Ship the first variants beyond the drone.

  • VisionThe endgame

    A self-defending trust layer, everywhere

    A distributed, self-locating trust layer spanning every autonomous domain: almost every attack answered onboard, the attacker's own transmissions turned into their exposure, and sovereign control that survives jamming, capture, cloning, and compromise. A world where an autonomous machine can no longer be quietly lied to.

06 / Why Now

The threat is now daily — and allies are funding this exact problem

What was a laboratory curiosity is now a standing feature of modern conflict, bleeding into civilian life. The figures below are public reported figures, presented as reported.

~123,000
flights disrupted around the Baltic, Jan–Apr 2025
Reported figure
~6,000
spoofing events logged over the Eastern Mediterranean in one year
Reported figure
Hormuz
commercial ships pulled off course near the Strait, 2025–26
Reported figure
€10M
joint NATO–Ukraine programme for resilient navigation, launched late 2025, with plans to scale
Reported figure

Allies are spending against this exact problem

When the largest defence alliance on earth funds resilient navigation in GPS-denied and electronic-warfare conditions — the precise problem the Black Box addresses — that category has arrived.

Australia needs it sovereign

Resilient PNT is a named priority, and this technology is dual-use, aligned with the AUKUS Pillar II focus on autonomous systems and information warfare. A sovereign capability is worth more to the nation than an imported one.

The timing maximises value

The invention is proven and the remaining step is a well-defined build — the point in a company's life where modest, well-placed funding produces the largest possible increase in value: a working prototype from a validated design.

Honesty wins defence buyers

Nodus Fortis sells on cost-imposition, not hype. In a field crowded with claims of "unhackable" and "un-spoofable," a company that draws an honest boundary — telling buyers exactly what it stops, what it doesn't, and where the line sits — is the one a serious customer can actually believe.

The Black Box does not promise the impossible. It makes an attack expensive, conspicuous, and locatable — and it tells you exactly where the line is drawn. That honesty is the point, and it is the product.

Nodus Fortis · Launceston, Tasmania, Australia
mia@nodusfortis.com · 0422 156 677
01 / 06