Soil-moisture and temperature probes from different stream data to VAS through secure telemetry. Each probe’s time-series is harmonised in CeRDI’s collector and validated before public display. Users can view profile “bucket” readings, rainfall, and temperature through an interactive interface with verified and unverified sensor filters.

System Flow

Acquisition → Ingestion → Validation → Visualisation

1. Sensor Data Acquisition Sources

   • Sources: Sensors from Southern Farming Systems (SFS), Agriculture Victoria, Mackillop Farm Management Group (MFMG), and other regional networks.

   • Telemetry: Data are transmitted via third-party services such as Adcon and Wildeye, then harvested by CeRDI systems.

   • Types: Soil moisture, temperature, rainfall, and derived indices (e.g., plant available water).

Figure: Screenshot from the Rainfields demonstration site

2. Data Ingestion & Harmonisation

   • Sensor data streams are imported into CeRDI’s Multi-Provider Probe Collector, replacing fragmented legacy import scripts.

   • Built using Laravel + TimescaleDB, it supports efficient time-series storage and live aggregates.

   • A token-secured REST API exposes endpoints to list devices, sensors, observed properties, units of measure, and to retrieve or export observations (CSV ≤ 500 k rows).

   • API Examples:

     • POST /api/login – Create token

     • GET /api/devices/list – List devices

     • GET /{provider}/observations/device/{id}/list – Retrieve time-series

     • GET /{provider}/observations/export – Download CSV

Figure: Agriculture Victoria’s soil moisture probes in the VAS portal (8 August 2024).

3. Validation & Quality Controls: The system distinguishes:

   • Verified Sensors: Calibrated soil-moisture and temperature probes; validated against field standards.

   • Unverified Sensors: Other environmental readings awaiting context or metadata.

   Automated checks detect inactive probes, stale data, inconsistent naming, and calibration errors. Manual review ensures only approved streams appear on public maps.

4. Visualisation & Portal Integration

   • Validated soil-moisture probe data are displayed live in the VAS interface.

   • Interactive “speedo” gauges show current soil-moisture levels for each probe site.

   Users can overlay rainfall fields, probe locations, and temperature data, filter by provider, network, verification status, and date range, and access summary statistics for each site. These visual elements provide an intuitive overview of real-time soil-water dynamics, supporting decision-making for researchers, farmers, and environmental managers.

Figure: An example of the interface using verified data.

API Access

The soil moisture probes section uses a REST-based architecture to expose sensor readings, metadata, and device information securely. Access is governed by API tokens that define the scope and ownership of datasets. Core Endpoints:

/api/login

POST

Generates API token for secure access

/api/devices/list

GET

Returns list of all registered devices with location and status

/{provider}/observations/device/{id}/list

GET

Retrieves time-series data for a specific probe

/{provider}/observations/export

GET

Exports validated readings (CSV ≤ 500,000 rows)

/api/sensors/list

GET

Lists available sensors and their measured properties

Example Response (JSON)

{

  "device_id": "SFS_204",

  "location": [-38.255, 143.490],

  "observed_property": "Soil Moisture (0–10cm)",

  "unit": "v/v",

  "timestamp": "2025-10-28T06:00:00Z",

  "value": 0.34,

  "provider": "Southern Farming Systems"

}

Data Sources

This section unifies live soil and climate data from key regional networks: Southern Farming Systems (SFS), Mackillop Farm Management Group (MFMG), and Agriculture Victoria, along with other CMA and private providers (Adcon, Wildeye).