Nexus Observatory Sign inCreate account
Nexus Observatory

/

Earth observation & GeoAI

/

GeoAI

Models trained and checked against field truth.

GeoAI, inside Earth observation & GeoAI: semantic segmentation and land cover, change detection, object detection and counting, training data and labels and more. Every item is traced to its specification, from ISO 19144-2 (Land Cover Meta Language) to OGC TrainingDML-AI Part 1, and runs on the open Observatory Protocol.

Plan with usArchitectureThe protocol
Technologies
9
At the frontier
2
Protocol functions
3 of 8
Industries served
16

Real elevation · Bow Valley, Alberta · from the Earth explainers

01What it is

GeoAI.

The reference frames, formats, services, analysis methods, machine learning and 3D models that turn observations into located, queryable and decision-ready geographic information. Each technology below carries what it is, what we do with it, and the standard it is drawn against.

GIS, maps and GeoAI

Semantic segmentation and land cover

Pixel-level classes such as water, built-up, crops and burn scars from imagery.

We design, integrate and operate.

ISO 19144-2 (Land Cover Meta Language) · OGC TrainingDML-AI Part 1

GIS, maps and GeoAI

Change detection

Finding what changed between dates, from new structures to landslides and flood extent.

We design, integrate and operate.

CEOS ARD (consistent inputs) · ISO 19157-1

GIS, maps and GeoAI

Object detection and counting

Locating assets, vehicles, vessels and damage in imagery and video.

We design, integrate and operate.

STAC Label extension · OGC TrainingDML-AI Part 1

GIS, maps and GeoAI

Training data and labels

Curated, versioned, licensed training sets with provenance and class definitions.

We design and operate.

OGC TrainingDML-AI Part 1 · STAC Label extension

GIS, maps and GeoAI

Model metadata and model cards

Recording inputs, bands, resolution, training data and limits with every model.

We design and operate.

STAC Machine Learning Model (MLM) extension

GIS, maps and GeoAI

Accuracy assessment

Sampling design and area-adjusted accuracy so map errors are reported honestly.

We design and operate.

ISO 19157-1:2023 · Olofsson et al. (2014) good practices for accuracy assessment

GIS, maps and GeoAI

Edge inference on imagery

Running models beside cameras and drones so alerts arrive without moving raw video.

We design, integrate, deploy and operate.

OGC API – Processes

Frontier · GIS, maps and GeoAI

Geospatial foundation models

Large pretrained models for imagery and maps fine-tuned with few labels.

We design and integrate.

OGC TrainingDML-AI Part 1

Frontier · GIS, maps and GeoAI

Super-resolution and multi-sensor fusion

Learning to combine sensors and resolutions into sharper or gap-free products, with uncertainty flagged.

We integrate.

ISO 19157-1:2023

02Architecture

Where it sits, and what it speaks.

GeoAI draws on 2 domains and 16 layers of the spectrum; the highlighted layers carry this area. It speaks 3 of the protocol’s 8 core functions.

Platforms and orbits

8

Optical, spectral and thermal sensors

8

Radar, lidar and microwave sensors

10

Processing levels and analysis-ready data

9

Calibration and validation

8

Tasking, ground segment and access

9

Data cubes, analysis and archives

7

Frontier and research

8

Reference systems and geodesy

7

Vector, raster and array formats

10

Tiling, maps and visual services

7

APIs, catalogues and discrete global grids

8

Spatial analysis and siting

10

GeoAI

10

3D, BIM and digital twins

8

Metadata, quality and publication

5

01JoinIdentity, attestation quote, signed policy.

02AdmitAdmitted for a stated scope, nothing wider.

03DescribeWhat, where, how and by whom, with calibration.

04TimeOne clock across the grid, with its error bound.

05ObserveEvery value with quality and provenance.

06QualifyTested against the specification, state on record.

07FederateSigned exchange across grids and trust domains.

08OrchestrateSoftware and configuration by signed plan.

03On the grid

The ground, measured before it is modelled.

Real elevation, line of sight, an orbit swath and a GeoAI pass, each an interactive explainer on open terrain data with its source shown.

Satellite pass

One orbit, one swath, line by line.

GeoAI

A detection method on real slopes, simulated outputs.

Relief

Terrarium-encoded DEM, contours by marching squares.

Layer stack

Every layer registered to one grid.

04In the field

Where this runs, sector by sector.

5 deployment scenarios from the sector research that rely on geoai.

Government and public safety

Deployable communications and sensing for incident command

A cell on wheels with satellite backhaul, mesh telemetry and environment nodes at the incident base.
Development and humanitarian

School connectivity with a school observatory

Satellite or fixed wireless, school Wi-Fi, an open connectivity probe and an optional station.
Development and humanitarian

Displacement-site connectivity and sensing

Point-to-multipoint wireless, satellite backhaul, solar and water-point sensors with protection safeguards.
Scientific

Ocean and coastal low-cost buoy arrays

Satellite-backhauled wave buoys accepted in a national programme trial.
Transport

Highway corridor mobile coverage for safety

Small cells and satellite backhaul along unserved roads.

05Open and enterprise

Open by default. Enterprise by contract.

The open counterpart lives on the main site for anyone to read and run. The enterprise service is designed, deployed and operated with you, under contract.

Open, on the main site

  • Observatories atlas
  • World grid
  • Map sources with licences

Enterprise, on this site

  • Capture campaigns
  • Site twins

06The rest of the line

Earth observation & GeoAI, area by area.

Earth observation & GeoAI

Ground segment

Downlink and process at the edge.
Earth observation & GeoAI

Drones and imagery

Campaigns planned to the ground sample distance.
Earth observation & GeoAI

Site twins

The site, measured in 3D, to scale.

Bring us the site.

Survey, design, deployment and operation for geoai. The form names who receives your details.

Talk to uscontact@observatory.global

Models trained and checked against field truth.

GeoAI, inside Earth observation & GeoAI: semantic segmentation and land cover, change detection, object detection and counting, training data and labels and more. Every item is traced to its specification, from ISO 19144-2 (Land Cover Meta Language) to OGC TrainingDML-AI Part 1, and runs on the open Observatory Protocol.

01What it is

02Architecture

Where it sits, and what it speaks.

03On the grid

04In the field

Where this runs, sector by sector.

05Open and enterprise

Open by default. Enterprise by contract.

The open counterpart lives on the main site for anyone to read and run. The enterprise service is designed, deployed and operated with you, under contract.

Open, on the main site
    Enterprise, on this site
      06The rest of the line

      Bring us the site.