Skip to main content

Sensor Things API

Purpose​

The Sensor Things API connects IoT devices to AgriFoodData and manages continuous measurements in real time.

Underlying standard​

OGC SensorThings API v1.1 — an open, widely adopted standard of the Open Geospatial Consortium for the management of sensors and their observations. The platform implements an OGC-STA-conformant endpoint, so any STA-capable client can talk to it directly.

Core resources​

ResourcePurpose
ThingA physical object — e.g. a sensor box at the edge of a field
SensorA single measurement device with its properties (vendor, model, range, unit)
DatastreamA logical channel: sensor + observed property + unit + a series of observations
ObservationA single timestamped measurement
LocationThe geographic position of a Thing
FeatureOfInterestWhat is being observed (often linked to a Field or Region from the Farm API)

Capabilities​

  • Register and manage sensors — onboarding, metadata, lifecycle (active, in service, retired).
  • Define datastreams per sensor and observed property.
  • Live data ingest — high-frequency streams via MQTT; batch and manual updates via REST.
  • OGC-STA queries — filter, order, paging — for time-series analysis.
  • Spatial context — link a Datastream to a Region in the Farm API so every measurement is grounded on a real piece of land. The main API also exposes region-scoped helpers (/v1/regions/{id}/sensors, …/datastreams, …/observations) and composite endpoints that create a Thing + Sensor + ObservedProperty + Datastream in a single call.

Minimal example​

client = AgriFoodData(token=...)

thing = client.sensor_things.things.create({
    "name": "SoilBox-42",
    "location": {"type": "Point", "coordinates": [78.49, 17.39]},
})

sensor = client.sensor_things.sensors.create({
    "name": "TDR-100",
    "encodingType": "application/pdf",
    "metadata": "https://example.com/datasheet.pdf",
})

stream = client.sensor_things.datastreams.create({
    "thing": thing.id,
    "sensor": sensor.id,
    "observedProperty": "soil-moisture-10cm",
    "unitOfMeasurement": {
        "name": "Volumetric Water Content",
        "symbol": "m³/m³",
    },
})

client.sensor_things.observations.publish(
    datastream=stream.id,
    result=0.27,
    phenomenonTime="2026-05-25T10:00:00Z",
)

See it in action​