Presentation Information
[eHvadpaV-05]Semantic Digital Twins for Marine Observing Systems: a SOSA-based Approach to Machine-Actionable Platform Descriptions
*Alessandro Oggioni1 (1. National Research Council of it (CNR) (Italy))
Marine observing infrastructures, including fixed platforms, moored buoys, and instrumented chains, generate vast amounts of heterogeneous data whose provenance, quality and context remain only partially machine-readable. This limits their discoverability, interoperability and reusability within marine data infrastructures.
This work presents a methodology for semantically describing marine observing platforms and their associated sensor systems using the Sensor, Observation, Sample and Actuator (SOSA) ontology and its extension SSN, complemented by PROV-O for deployment and calibration provenance, GeoSPARQL for geometric representation, and the NERC and QUDT vocabularies for standardized property types and units. The approach was evaluated through real-world case studies conducted at marine sites of the Integrated European Long-Term Ecosystem, Critical Zone and Socio-Ecological Research Infrastructure (eLTER RI).
The resulting RDF descriptions constitute semantic digital twins of observing systems, representing both the physical structure of platforms, including sub-components, sensor cages, mooring chains and anemometer masts, and the full lifecycle of sensor instances, including acquisition, calibration history, deployment context and spatial position within a local coordinate reference system.
These machine-actionable descriptions support automated reasoning over sensor metadata. By linking sensor instances to manufacturer specifications, calibration records, deployment configurations and spatial context, reasoning agents can infer key information such as measurement accuracy and precision. Furthermore, the approach enables the integration of observations with established frameworks, such as the Essential Ocean Variables (EOVs) and the Essential Climate Variables (ECVs), by considering not only observed properties but also contextual information, including sampling frequency, sensor position in the water column and deployment characteristics. This enables the automated assessment of whether and to what extent an observation contributes to specific monitoring objectives.
The approach supports FAIR data principles and lays the groundwork for interoperable, machine-actionable and robot-ready marine observation registries within research infrastructures.
This work presents a methodology for semantically describing marine observing platforms and their associated sensor systems using the Sensor, Observation, Sample and Actuator (SOSA) ontology and its extension SSN, complemented by PROV-O for deployment and calibration provenance, GeoSPARQL for geometric representation, and the NERC and QUDT vocabularies for standardized property types and units. The approach was evaluated through real-world case studies conducted at marine sites of the Integrated European Long-Term Ecosystem, Critical Zone and Socio-Ecological Research Infrastructure (eLTER RI).
The resulting RDF descriptions constitute semantic digital twins of observing systems, representing both the physical structure of platforms, including sub-components, sensor cages, mooring chains and anemometer masts, and the full lifecycle of sensor instances, including acquisition, calibration history, deployment context and spatial position within a local coordinate reference system.
These machine-actionable descriptions support automated reasoning over sensor metadata. By linking sensor instances to manufacturer specifications, calibration records, deployment configurations and spatial context, reasoning agents can infer key information such as measurement accuracy and precision. Furthermore, the approach enables the integration of observations with established frameworks, such as the Essential Ocean Variables (EOVs) and the Essential Climate Variables (ECVs), by considering not only observed properties but also contextual information, including sampling frequency, sensor position in the water column and deployment characteristics. This enables the automated assessment of whether and to what extent an observation contributes to specific monitoring objectives.
The approach supports FAIR data principles and lays the groundwork for interoperable, machine-actionable and robot-ready marine observation registries within research infrastructures.
Comment
To browse or post comments, you must log in.Log in