Know which GNSS alerts are real
Ship tracking services already flag jamming and spoofing of satellite positioning (GNSS, such as GPS). None we found tells you which alerts are real. Ours does, and shows you the evidence.
The gap
Detection is a solved problem. Tracking platforms watch the position reports ships broadcast (AIS) and tell you that a position went missing or jumped. Most stop there and leave you the hard part: was it jamming, or did the ship simply sail out of range of the shore receivers? Both look the same in the data, so operations teams get a stream of alerts they cannot act on, and the real ones get lost in it. We searched for a product that sorts these alerts by cause and shows its reasoning, and found none. So we built one, from our own research on more than 150 million ship position reports.
What the research says
ships in UAE, Qatari, Omani and Iranian waters hit by GPS and AIS interference within one day in February 2026.
GPS interference incidents a day on ships by 2 March 2026, up from 350 when the conflict began.
of area-wide jamming alerts from a published detection method, over eight days of Danish ship data, were ships out of receiver range.
Source: Waselix analysis of Danish Maritime Authority AIS data (8 days, 2024)
What GNSS triage for shipping covers
-
A verdict on every alert
Each alert from your tracking provider or your own detector gets a verdict and a one-line reason.
-
Says when it cannot tell
If the data cannot settle an alert, it says so and sends it to a person with the evidence for and against, instead of guessing.
-
Built to fail safe
It does not close alerts where other data shows interference was active. One in twenty of its closes goes to a person as a spot check, and any close can be reopened with one click.
-
Evidence you can check
Every verdict lists the data behind it, and each sentence of the draft is checked against that data before anyone reads it.
-
Spoofing flagged
When reported positions jump further than any ship could travel, the alert is flagged as spoofing and shows where the positions went.
-
Independent cross-checks
Each alert is tested against data the alert itself never used, so a ship out of range is not mistaken for a jammed one.
-
Draft reports
Reports come filled in with everything the data can tell. The fields only your crew can complete are left clearly open.
-
One screen for the duty officer
A map and a review queue: open an alert, read the verdict, then sign it or strike it.
-
Your AIS data, any source
Works with AIS position data from your provider or your own receivers: as files today, with a live feed connected during installation.
-
Runs on your own servers
The system and its AI model run on one server with an ordinary graphics card inside your network. Your fleet data stays there, and there is no cloud contract.
Deliverables
- A pilot report on your own historical alerts, showing which ones the system closed and which it sent to a person
- The system installed on your servers and connected to your ship position data
- Report templates matched to how you already report incidents
- Training for your operations team
Who it's for
- Tanker, LNG and container operators working through the Gulf and the Strait of Hormuz
- Shipping companies with fleets in the Red Sea, the Black Sea and the Baltic
- Fleet operations and maritime security centres
- Ports and vessel traffic services
How we work
-
Scoping call
We learn which alerts you receive today, where they come from, and what your team does with them.
-
Pilot on your data
We run the system on your own historical alerts and show you what it closed and what it left for a person.
-
Install and connect
The system goes onto your servers, connected to your data, and your team is trained on it.
-
Ongoing support (optional)
Monitoring and tuning as interference patterns change.
Questions from shipping teams
Doesn't our tracking provider already do this?
Providers detect interference and show you where it is. Deciding which alerts are real, and showing the evidence for that decision, is the step we found no one offering. This service sits on top of your provider's alerts and does that step. You keep your provider.
Where is GNSS interference affecting ships?
In July 2026 the Joint Maritime Information Center reported recurring GNSS interference at consistent levels in the Arabian Gulf and the Strait of Hormuz, with more near Yanbu in the Red Sea and near Cyprus. Similar patterns appear in the Baltic and the Black Sea (UKMTO, JMIC Advisory Note update 071 (July 2026)).
Why do ships disappear from AIS?
Most often because they sail out of range of the shore receivers that pick up their signal. Jamming can look exactly the same in the data. This service separates the two for each alert and flags what it cannot settle.
How many alerts can it close?
On eight days of Danish ship data it closed 76% of alerts on its own, and none of those closes was contradicted by independent checks. Your rate depends on your data, which is what the pilot measures.
What if it closes a real event?
It is built to avoid that. It does not close alerts where other data shows interference was active, one in twenty closes goes to a person as a spot check, and any close can be reopened with one click. On 42 live Baltic alerts in October 2026, none of its closes fell where interference was active.
Can it tell spoofing from jamming?
Spoofing moves positions, so the system flags jumps no ship could make. In live Baltic data in October 2026 it flagged tankers and cargo ships near St Petersburg whose reported positions jumped about 300 km inland. Jamming leaves gaps, and so does a ship that is simply out of range. When the data cannot tell those apart, the alert goes to a person with the evidence on both sides.
What hardware does it need?
One server with an ordinary graphics card. The AI model is small enough to run locally, so there is no cloud contract and nothing to send outside your network.
Does our data leave our network?
No. The system and its AI model run on your own servers. Only public context, such as space weather, is looked up online.
Is the AI making the decision?
No. Rules settle only what the data explains. The AI drafts the rest, each sentence is checked, and a person signs.
Ready to see where you stand?
Book a free scoping call. We'll tell you honestly whether this is a fit.