The Global Precision Network for Physical AI

A decentralized network of GNSS reference stations delivering centimeter-level positioning for drones, robotics, autonomous vehicles, and industrial systems.

GEODNET is a decentralized GNSS reference network delivering centimeter-level positioning worldwide. Thousands of independently operated base stations continuously collect multi-constellation satellite and space-weather observations, which the network aggregates into RTK corrections, post-processing data, and raw observation streams — with cryptographic guarantees over data transmission, location, and ownership authenticity.

RTK

Stream centimeter-level corrections to any compatible GNSS receiver over the internet — no local base station required.

  • Standard NTRIP delivery, RTCM 3.2
  • 1–2 cm achievable precision
  • WGS84, ITRF2020/2014, NAD83 + 25 datums
Open RTK Console
PPK

Download historical observation and correction data to post-process trajectories when a real-time link isn't available or maximum accuracy is the priority.

  • Historical data by station & time window
  • Post-processing-ready observations
  • Drone mapping, photogrammetry, survey
Open Post -Processing
Raw GNSS Data

Direct, continuous 1 Hz raw observation streams (RTCM 3.2 MSM4) from a customized pool of stations — the wholesale layer beneath RTK. Build your own VRS, PPP, and POD engines.

  • Per-station raw RTCM streams
  • Station API + live station inventory
  • QC metrics & enterprise console
Open Raw Data Console

How GEODNET Delivers Precision

GNSS Satellites

Raw Signals

Base Station Network

Receive satellite signals

Calculate local errors

GEODNET Network

Aggregates station data

Computes RTK corrections

Cloud / Internet Delivery

Streams correction data (NTRIP / API)

Rover Collect GNSS Signals

Receive GNSS + corrections

Results

cm-level positioning

No drift

Real-time accuracy

The most complete backend set of web services and APIs in the industry

RTK · Real-Time

RTK · Standard NTRIP Access

Deliver centimeter-level positioning to any compatible GNSS receiver over the internet—no local base station required.

Best for:

Surveying

Robotics

Drones

Precision Agriculture

Autonomous Vehicles

Real-time
GGA · Real-Time Location

Access live rover position data through the GEODNET network.

Best for:

Fleet tracking

Asset management

Analytics

Real-time

PPK · Post-Processing

PPK · Post-Processing

Download historical correction data for post-processing workflows.

Best for:

Fleet tracking

Asset management

Analytics

Historical

RAW GNSS DATA · DEVELOPER

Raw GNSS Data

Access raw GNSS observations for custom positioning applications.

Best for:

Research

RTK engines

AI & ML

Foundation
Quality Monitoring

Monitor station health and network data quality in real time.

Best for:

Enterprise deployments

Network validation

Verified

How businesses integrate with GEODNET

Filling coverage gaps, building a positioning product from scratch, or expanding into new markets overnight — none of it has to start with deploying hardware.

Augment your network

Already run RTK infrastructure? GEODNET data fills coverage gaps instantly — no new hardware, no site negotiations.

Build from scratch

Skip the infrastructure buildout. Use GEODNET as your data backbone and go straight to the software layer.

Expand overnight

Extend service into rural, international, and remote areas without a single field deployment.

Research & development

Academic and R&D teams use raw GNSS data for atmospheric research, algorithm development, and satellite analysis.

Raw GNSS Data

GEODNET operates a global backbone station network supplying data to leading precision GNSS organizations — fueling R&D that can't be replicated in simulation.



Each backbone station runs a Hemisphere P40 GNSS receiver, with cellular 4G, battery backup, and a primary wired internet connection. Positioning and monitoring use Natural Resources Canada's CSRS-PPP service.

The Raw Data Service delivers direct, continuous 1 Hz raw observation streams (RTCM 3.2 MSM4) from a customized pool of reference stations — not rover-specific point corrections — so you can build independent Network RTK (VRS), PPP, and Precise Orbit Determination engines on your own infrastructure.

GEODNET uses Natural Resources Canada's CSRS-PPP service as part of its station positioning and monitoring.

Network & data specifications

Active stations
20,000+ globally (live count on map)
Coverage
Global · 25+ regional datums
Update rate
1 Hz
Achievable precision
1–2 cm (RTK)
Stream format
RTCM 3.2 · MSM4
Delivery protocol
NTRIP v1.0
raw caster
rtcm.geodnet.com:2201
Constellations
GPS · GLONASS · Galileo · BeiDou · QZSS · IRNSS/NavIC
RTCM messages
1005 coords
1033 rx/antenna/fw
1074/1084/1094
1114/1124/1134 MSM4
Observables (MSM4)
Code (pseudorange) · carrier phase · C/N0 · lock-time indicator · half-cycle ambiguity
Coordinate frames
WGS84 (G2139) · ITRF2020 · ITRF2014 · NAD83(2011) + 25 regional datums
Station metadata
ECEF coordinates · receiver & antenna model · firmware version (via 1033) · QC quality flags
APIs
/mountpoints
/coordinates
/stations
 /qc
HTTP Basic Auth
Quality control
Hourly TEQC / RTKLIB: Availability >99.9% · MP12 / MP21 <0.5 m · IOD slips · o/slps
Station states
Active (in caster) · Online (streaming, pre-QC) · Offline

Build on the Precision Network for Physical AI

GEODNET is already deployed across real markets, real customers, and real-world positioning demand.