Skip to content

Home/Alternatives/Commodity trading

The alternative to Kpler for Commodity Trading

Evaluations of maritime data providers reasonably begin with the established names. This page sets out the infrastructure we operate, how each position is produced, and the questions to put to any provider before a feed enters production.

100,000+

Vessels tracked / day

49M+

AIS positions / day

40+

Countries covered

<5s

Antenna to your API

Who we are, and how the data is actually produced.

Every position we supply originates on physical infrastructure we operate. None of it is acquired from, or blended with, a third party feed.

  • Collection runs on infrastructure we own.Worldwide AIS Network ApS is a Danish company based in Copenhagen, CVR 45911748, backed by Den Danske Maritime Fond, approved under the Startup Denmark scheme, and building with peaq. Our own network is terrestrial and ship-borne, with satellite feeds for open ocean.

  • Manipulated signals arrive flagged, not corrected.Every station runs the same dedicated standardised hardware, encrypting each message at the instant of capture. Every position is nanosecond time-stamped and assigned a trust score by our anomaly detection models, so a spoofed or manipulated signal is flagged rather than absorbed into an apparently clean track.

  • We supply the source material, not the interpretation.You receive the source material and the record of its origin, in under five seconds from antenna to your API. Delivery is raw NMEA over TCP, a REST API, a WebSocket stream, or SAPIENT compliant output to BSI Flex 335. Interpretation and the resulting signal remain with your quantitative researchers.

Run us alongside your current feed.Send an MMSI or IMO number and a date range, and we will return what our network collected, with receiver metadata and trust scores attached.

Benchmark against your feed

What to look for in an AIS data provider for commodity and energy trading.

These are the questions we would put to a provider if we were buying rather than selling. They apply to every provider on your shortlist, including us.

  • Establish where each position originated

    A cargo figure or a continuous track is the output of decisions taken upstream by someone else. Ask whether the provider can return the underlying message, the receiving station and the moment of capture, for any position in the archive. Where provenance is not carried through to delivery, an analyst cannot distinguish a genuine draught change from an interpolation, and compliance cannot evidence to an examiner what was screened. Our positions arrive with receiver metadata, a nanosecond capture timestamp and a trust score attached.

  • Define where the latency clock starts and stops

    A response time quoted at the API boundary measures the final hop, not the full path. For anything that has to move intraday, the figure that matters is capture to availability. Require every provider to measure between the same two points, otherwise the numbers describe different things. Ours is under five seconds from antenna to your API.

  • Test coverage in your own waters, against your own vessels

    Global coverage claims cannot be compared on paper. Select fifteen or twenty MMSIs the desk genuinely trades, a load port or a chokepoint, and a window for which you already hold data, then ask each provider to return what they collected for precisely that. We track 100,000+ vessels a day across 40+ countries, and we will state what we hold for a named area rather than answer in the abstract.

  • Establish whether a second feed is a genuinely independent source

    Redundancy is only redundancy where two feeds share neither upstream ownership nor the same underlying receivers. Ask directly whether the data is collected on the provider's own infrastructure or acquired and blended, and who owns that infrastructure. It matters technically, because an independent comparison is the only means of corroborating a suspicious track, and commercially, because genuine substitutability is what gives a data team leverage at renewal.

  • Distinguish spoofing from GNSS interference

    These are two distinct problems. A transponder can broadcast a false position, while area-wide interference distorts genuine ones, and the two leave different patterns. Ask whether suspect positions are flagged and passed through with the evidence, or corrected before delivery. A model cannot account for a correction it cannot observe. Every position we deliver carries a trust score from our anomaly detection models.

  • Match delivery formats to the stack the desk already operates

    Establish the shape of the payload before the commercial conversation, not after. Engineers who need the AIVDM sentence untouched ingest raw NMEA over TCP, researchers work from a REST API, and intraday work requires a WebSocket stream. Teams operating alongside defence or critical infrastructure may need SAPIENT compliant output to the BSI Flex 335 interface, the architecture for which is owned by the UK MOD. We deliver all four.

How we produce the data
a trading desk receives.

Our own network, a single hardware standard, and a trust score on every position before it reaches your models.

Monitor

Collect the traffic your models price

Tankers, gas carriers and dry bulk across 40+ countries, 100,000+ vessels a day, collected on our own terrestrial and ship-borne receivers, plus satellite feeds for open ocean rather than sourced from someone else's network.

Verify

Score it before it reaches the desk

Each message is encrypted on the hardware at the moment of capture, nanosecond time-stamped, and scored by our anomaly detection models, so a manipulated track arrives flagged with the evidence rather than tidied up.

Act

Deliver into the pipes you already have

Raw NMEA over TCP, a REST API, a WebSocket stream or SAPIENT compliant output to BSI Flex 335, in under five seconds from antenna to your API, with the provenance still attached to every position.

How Worldwide AIS Network compares.

The questions worth putting to every provider under evaluation, including us.

What to checkWhat to ask any providerWorldwide AIS Network
Where the data comes fromDo you operate the receivers yourselves, or buy and blend feeds from others?Our own independent physical infrastructure. Terrestrial and ship-borne, plus satellite feeds.
Provenance on each positionCan you tell me which receiver collected this specific position, and when?Every position carries its originating receiver and a nanosecond time stamp.
Handling of spoofingHow do you detect a manipulated or spoofed position before it reaches me?Anomaly detection models assess the validity of every position in real time.
LatencyWhat is the delay from reception to the data being available in my system?Under five seconds, antenna to API.
Delivery formatsWhich formats do you deliver, and will they drop into what I already run?Raw NMEA over TCP, REST API, WebSocket stream, and SAPIENT compliant output to BSI Flex 335.
Coverage where it matters to youCan you show me real data for my specific waters before I commit?100,000+ vessels a day across 40+ countries. Ask for a sample of your own area.
Running alongside what you haveCan I add you next to my current provider rather than replacing it?Yes. Running side by side is the intended pattern, not a compromise.

We have deliberately not filled in a column for any other provider. Coverage, latency and sourcing change, and a claim we cannot evidence is worth nothing to you. Put these questions to everyone you are evaluating and compare the answers yourself.

Questions we get from desks part-way through an evaluation.

Can we run Worldwide AIS Network alongside our existing provider?

Yes, and it is the pattern we would recommend. Most desks that begin with us replace nothing at the outset. They run an independent feed alongside their incumbent, compare the two on vessels and windows they know well, and decide from there. A second source collected on separate infrastructure is the only practical means of corroborating an unexpected track, so parallel running is a legitimate long term arrangement rather than a trial stage to be exited.

How much integration work is involved in adding or switching a feed?

It depends almost entirely on the point of ingest. If your pipeline already consumes raw NMEA over TCP, repointing it at another stream is a configuration change. A REST API or a WebSocket stream into a research environment is usually a small connector. The effort sits in the mapping rather than the transport: reconciling field by field how positions, timestamps and identifiers correspond to what your models expect, and deciding how the trust score is applied. We will not quote a timeline we cannot stand behind, but we will take your engineers through the formats before any commitment.

Do you cover the waters our desk actually trades?

Interrogate the specific areas rather than accepting a global figure at face value, ours included. We track 100,000+ vessels a day across 40+ countries, and coverage in any given area follows where our own network has reception rather than a single global claim. Name the load ports, chokepoints or basins that matter to your book, and we will state what we collect there and what we do not.

Are you an independent AIS data provider or a reseller?

Independent in the infrastructure sense, not the marketing sense. We own and operate the physical receivers, terrestrial and ship-borne, with satellite feeds for open ocean, and the same dedicated standardised hardware at every station. The data is neither bought nor blended from a third party feed. Worldwide AIS Network ApS is a Danish company, CVR 45911748, backed by Den Danske Maritime Fond and Startup Denmark.

How do we compare two AIS feeds fairly?

Fix the variables first: the same MMSIs, the same window, the same area, with agreement in advance on what is being counted. Message volume per vessel per hour, the timestamp of first receipt for a given event, the length and placement of gaps, and the behaviour of each feed in waters where interference is common are more informative than any feature list. Then establish whether each provider can explain a disagreement between the two, because a feed that can attribute a position to a receiving station and a capture time can be adjudicated, and one that cannot must simply be trusted.

Past the comparison stage?

Detail on flows and balances modelling, floating storage, chokepoint counts, and pre-trade and pre-fixture screening is set out on our main page for AIS data for commodity trading.

AIS data for commodity trading

Put an independent feed next to the one you are evaluating.

Send an MMSI, a chokepoint or a date range. We will return what our network collected and where each message originated, for assessment against everything else under evaluation.

Benchmark against your feed