The SensorThings API Data Model

In order to understand the SensorThings API, it is essential to understand the underlying data model. This model consists of the following classes:

• Things: Some real-world thing, in this context usually the thing where the sensor is located. This could be a monitoring station or just the room where the sensor has been mounted.
• Locations: Location of the Thing.
  The Location is provided separately from the Thing as a Thing can move to a different Location. When a Thing moves, a new Location is created. The old Location stays as it was, and is still linked to the Thing through a HistoricalLocation entity.
• HistoricalLocations: Links to the previous Locations where a Thing was in the past.
• Datastreams: a Datastream links together a Thing with a Sensor that is measuring an ObservedProperty to provide an entry point for a time series.
  All Observations created by the Sensor pertaining to the ObservedProperty are linked to this Datastream.
• Sensors: A description of a sensor that provides values, including information on the measurement method.
  While formally designed for sensors in the IoT context, the Sensor could also be a human observer.
• ObservedProperties: The ObservedProperties describe what is being measured by a Sensor at a specific Location.
• Observations: A single measurement value.
  Information on what ObservedProperty was measured with which type of Sensor is provided by the Datastream to which this Observation is linked; the object on which the measurement was performed is provided by the FeatureOfInterest of this Observation.
• FeaturesOfInterest: the FeatureOfInterest is the object on which the measurement was performed.