connect OceanLab catalogue as ODIS node

[jmckenna]

thanks @lara-veylit ! I'll do my best to summarize, but please expand/correct where necessary:

Here's a summary (also beneficial for other OIH/SINTEF team members) :

• OceanLab is a sensor network (recording currents, salinity, temp, etc.)
  • schema: https://oceanlab.azure.sintef.no/d/YinybPjnk/list-of-all-data?orgId=1
  • stored in InfluxDB
  • already has SensorThingsAPI (https://github.com/ILIAD-ocean-twin)
• no current metadata catalogue, but examining options, including pygeoapi
  • pygeoapi does connect to SensorThingsAPI as a provider
  • pygeoapi does automatically generate and embed JSON-LD
• possible next steps:
  • create a JSON-LD template together at the next meeting using an ODIS pattern
  • possible ODIS templates to use:
    • dataset
    • services (to describe the SensorThingsAPI)
  • then, SINTEF can either manually generate the JSON-LD from the template, or leverage something like pygeoapi to automatically generate the JSON-LD
  • then provide the OIH team with a sitemap.xml that points to the JSON-LD (pages or actual files)
  • then, create an entry in the ODIS Catalogue for OceanLab, that points to that sitemap url
  • then, ODIS can start harvesting from the OceanLab endpoint (and Google Dataset Search will also harvest)

cc @UteBroenner @fils @pbuttigieg

[UteBroenner]

:-)

[jmckenna]

• ChatGPT output of JSON-LD from @UteBroenner :

  {
  "@context": {
  "sml": http://www.opengis.net/sensorML/2.0,
  "gml": http://www.opengis.net/gml/3.2,
  "wtr": http://www.opengis.net/waterml/2.0,
  "swe": http://www.opengis.net/swe/2.0,
  "swes": http://www.opengis.net/swes/2.0,
  "xlink": http://www.w3.org/1999/xlink
  },
  "@graph": [
  {
    "@type": "sml:System",
    "@id": "ctd_sensor_munkholmen",
    "sml:identification": {
      "sml:IdentifierList": {
        "sml:identifier": {
          "sml:Term": {
            "@definition": "urn:ogc:def:identifier:OGC:1.0:shortName",
            "sml:value": "CTD Sensor"
          }
        }
      }
    },
    "sml:classification": {
      "sml:ClassifierList": {
        "sml:classifier": {
          "sml:Term": {
            "@definition": "urn:ogc:def:classifier:OGC:1.0:platform",
            "sml:value": "Munkholmen"
          }
        }
      }
    },
    "sml:capabilities": {
      "sml:CapabilityList": {
        "sml:capability": {
          "sml:Abstract": {
            "sml:value": "This CTD sensor measures temperature and salinity of the water."
          }
        }
      }
    }
  },
  {
    "@type": "wtr:MeasurementTimeseries",
    "wtr:timeSeriesMetadata": {
      "wtr:temporalExtent": {
        "gml:TimePeriod": {
          "@gml:id": "temporal_extent",
          "gml:beginPosition": "2023-05-07T23:00:30.846161Z"
        }
      }
    },
    "wtr:point": {
      "gml:TimeInstant": {
        "@gml:id": "time_instant",
        "gml:timePosition": "2023-05-07T23:00:30.846161Z"
      },
      "wtr:MeasurementTVP": {
        "wtr:time": "2023-05-07T23:00:30.846161Z",
        "wtr:value": 7.8394,
        "wtr:metadata": {
          "wtr:MeasurementMetadata": {
            "wtr:observedProperty": {
              "@xlink:href": http://example.com/temperature
            },
            "wtr:procedure": {
              "@xlink:href": "#ctd_sensor_munkholmen"
            }
          }
        }
      }
    }
  }
  ]
  }

[pbuttigieg]

This use case also requires the following:

1. SensorNetwork: http://sensorml.com/ont/swe/system/SensorNetwork
2. AUVs
3. Sensors themselves
4. ...

[pbuttigieg]

It also looks like we'll need a SSSOM to map some of the more generic SensorML properties (some of which are undefined, like Platform) to the schema.org