Azure FarmBeats client library for Python - version 1.0.0b2

FarmBeats is a B2B PaaS offering from Microsoft that makes it easy for AgriFood companies to build intelligent digital agriculture solutions on Azure. FarmBeats allows users to acquire, aggregate, and process agricultural data from various sources (farm equipment, weather, satellite) without the need to invest in deep data engineering resources.  Customers can build SaaS solutions on top of FarmBeats and leverage first class support for model building to generate insights at scale.

Use FarmBeats client library for Python to do the following.

  • Create & update parties, farms, fields, seasonal fields and boundaries.
  • Ingest satellite and weather data for areas of interest.
  • Ingest farm operations data covering tilling, planting, harvesting and application of farm inputs.

Source code | Package (PyPi) | API reference documentation | Product documentation | Changelog

Getting started

Prerequisites

To use this package, you must have:

Install the package

Install the Azure FarmBeats client library for Python with pip:

pip install azure-agrifood-farming

Authenticate the client

To use an Azure Active Directory (AAD) token credential, provide an instance of the desired credential type obtained from the azure-identity library.

To authenticate with AAD, you must first pip install azure-identity and enable AAD authentication on your FarmBeats resource. If you followed the installation docs when creating the FarmBeats resource, this is already covered.

After setup, you can choose which type of credential from azure.identity to use. As an example, DefaultAzureCredential can be used to authenticate the client:

Set the values of the client ID, tenant ID, and client secret of the AAD application as environment variables: AZURE_CLIENT_ID, AZURE_TENANT_ID, AZURE_CLIENT_SECRET

Use the returned token credential to authenticate the client:

from azure.agrifood.farming import FarmBeatsClient
from azure.identity import DefaultAzureCredential

credential = DefaultAzureCredential()
client = FarmBeatsClient(endpoint="https://<my-account-name>.farmbeats.azure.net", credential=credential)

Key concepts

Basic understanding of below terms will help to get started with FarmBeats client library.

Farm Hierarchy

Farm hierarchy is a collection of below entities.

  • Party - is the custodian of all the agronomic data.
  • Farm - is a logical collection of fields and/or seasonal fields. They do not have any area associated with them.
  • Field - is a multi-polygon area. This is expected to be stable across seasons.
  • Seasonal field - is a multi-polygon area. To define a seasonal boundary we need the details of area (boundary), time (season) and crop. New seasonal fields are expected to be created for every growing season.
  • Boundary - is the actual multi-polygon area expressed as a geometry (in geojson). It is normally associated with a field or a seasonal field. Satellite, weather and farm operations data is linked to a boundary.
  • Cascade delete - Agronomic data is stored hierarchically with party as the root. The hierarchy includes Party -> Farms -> Fields -> Seasonal Fields -> Boundaries -> Associated data (satellite, weather, farm operations). Cascade delete refers to the process of deleting any node and its subtree.

Scenes

Scenes refers to images normally ingested using satellite APIs. This includes raw bands and derived bands (Ex: NDVI). Scenes may also include spatial outputs of an inference or AI/ML model (Ex: LAI).

Farm Operations

Fam operations includes details pertaining to tilling, planting, application of pesticides & nutrients, and harvesting. This can either be manually pushed into FarmBeats using APIs or the same information can be pulled from farm equipment service providers like John Deere.

Examples

Create a Party

Once you have authenticated and created the client object as shown in the Authenticate the client section, you can create a party within the FarmBeats resource like this:

from azure.identity import DefaultAzureCredential
from azure.agrifood.farming import FarmBeatsClient

credential = DefaultAzureCredential()
client = FarmBeatsClient(endpoint="https://<my-account-name>.farmbeats.azure.net", credential=credential)

party_id = "party-1"

party = client.parties.create_or_update(
    party_id=party_id,
    party={
        "name": party_name,
        "description": party_description
    }
)

Create a Farm

from azure.identity import DefaultAzureCredential
from azure.agrifood.farming import FarmBeatsClient

credential = DefaultAzureCredential()
client = FarmBeatsClient(endpoint="https://<my-account-name>.farmbeats.azure.net", credential=credential)

party_id = "party-1" # Using party from previous example

farm = client.farms.create_or_update(
    party_id=party_id,
    farm_id="farm-1",
    farm={
        "name": farm_name,
        "description": farm_description
    }
)

Create a Season

Creating a Season object, spanning from April to August of 2021.

from azure.identity import DefaultAzureCredential
from azure.agrifood.farming import FarmBeatsClient

credential = DefaultAzureCredential()
client = FarmBeatsClient(endpoint="https://<my-account-name>.farmbeats.azure.net", credential=credential)

season_id = "contoso-season"
season_name = "contoso-season-name"
season_description = "contoso-season-description"
year = "2021"
start_date_time = "2021-01-01T20:08:10.137Z"
end_date_time = "2021-06-06T20:08:10.137Z"

season = client.seasons.create_or_update(
        season_id=season_id,
        season={
            "name": season_name,
            "year": year,
            "startDateTime": start_date_time,
            "endDateTime": end_date_time,
            "description": season_description
        }
    )

Create a Boundary

Creating a Boundary for the Seasonal Field created in the preceding example.

from azure.identity import DefaultAzureCredential
from azure.agrifood.farming import FarmBeatsClient

credential = DefaultAzureCredential()
client = FarmBeatsClient(endpoint="https://<my-account-name>.farmbeats.azure.net", credential=credential)

party_id = "party-1"
boundary_id = "boundary-1"

boundary = client.boundaries.create_or_update(
    party_id=party_id,
    boundary_id=boundary_id,
    boundary={
        "geometry": {
            "type": "Polygon",
            "coordinates":
                [
                    [
                        [73.70457172393799, 20.545385304358106],
                        [73.70457172393799, 20.545385304358106],
                        [73.70448589324951, 20.542411534243367],
                        [73.70877742767334, 20.541688176010233],
                        [73.71023654937744, 20.545083911372505],
                        [73.70663166046143, 20.546992723579137],
                        [73.70457172393799, 20.545385304358106],
                    ]
                ]
        },
        "status": "<string>",
        "name": "<string>",
        "description": "<string>"
    }
)

Ingest Satellite Imagery

Triggering a Satellite Data Ingestion job for the boundary created above, to ingest Leaf Area Index data for the month of January 2020. This is a Long Running Operation (also called a 'Job'), and returns a Poller object. Calling the .result() method on the poller object waits for the operation to terminate, and returns the final status.

from azure.identity import DefaultAzureCredential
from azure.agrifood.farming import FarmBeatsClient

from isodate.tzinfo import Utc
from datetime import datetime

credential = DefaultAzureCredential()
client = FarmBeatsClient(endpoint="https://<my-account-name>.farmbeats.azure.net", credential=credential)

party_id = "party-1"
boundary_id = "westlake-boundary-1"
start_date_time = "2021-01-01T20:08:10.137Z"
end_date_time = "2021-06-06T20:08:10.137Z"

# Queue the job
satellite_job_poller = client.scenes.begin_create_satellite_data_ingestion_job(
    job_id=job_id,
    job={
        "boundaryId": boundary_id,
        "endDateTime": end_date_time,
        "partyId": party_id,
        "startDateTime": start_date_time,
        "provider": "Microsoft",
        "source": "Sentinel_2_L2A",
        "data": {
            "imageNames": [
                "NDVI"
            ],
            "imageFormats": [
                "TIF"
            ],
            "imageResolution": [10]
        },
        "name": "<string>",
        "description": "<string>"
    }
)

# Wait for the job to terminate
satellite_job = satellite_job_poller.result()
job_status = satellite_job_poller.status()

Get Ingested Satellite Scenes

Querying for the scenes created by the job in the previous example.

from azure.identity import DefaultAzureCredential
from azure.agrifood.farming import FarmBeatsClient

from datetime import datetime

credential = DefaultAzureCredential()
client = FarmBeatsClient(endpoint="https://<my-account-name>.farmbeats.azure.net", credential=credential)

party_id = "party-1"
boundary_id = "boundary-1"

scenes = client.scenes.list(
    party_id=party_id,
    boundary_id=boundary_id,
    start_date_time=start_date_time,
    end_date_time=end_date_time,
    provider="Microsoft",
    source="Sentinel_2_L2A"
)

for scene in scenes:
    bands = [image_file["name"] for image_file in scene["imageFiles"]]
    bands_str = ", ".join(bands)
    print(f"Scene has the bands {bands_str}")

Troubleshooting

General

The FarmBeats client will raise exceptions defined in [Azure Core][azure_core] if you call .raise_for_status() on your responses.

Logging

This library uses the standard logging library for logging. Basic information about HTTP sessions (URLs, headers, etc.) is logged at INFO level.

Detailed DEBUG level logging, including request/response bodies and unredacted headers, can be enabled on a client with the logging_enable keyword argument:

import sys
import logging
from azure.identity import DefaultAzureCredential
from azure.agrifood.farming import FarmBeatsClient
# Create a logger for the 'azure' SDK
logger = logging.getLogger('azure')
logger.setLevel(logging.DEBUG)
# Configure a console output
handler = logging.StreamHandler(stream=sys.stdout)
logger.addHandler(handler)
endpoint = "https://<my-account-name>.farmbeats.azure.net"
credential = DefaultAzureCredential()
# This client will log detailed information about its HTTP sessions, at DEBUG level
client = FarmBeatsClient(endpoint=endpoint, credential=credential, logging_enable=True)

Similarly, logging_enable can enable detailed logging for a single call, even when it isn't enabled for the client:

client.crops.get(crop_id="crop_id", logging_enable=True)

Next steps

Additional documentation

For more extensive documentation on the FarmBeats, see the FarmBeats documentation on docs.microsoft.com.

Contributing

This project welcomes contributions and suggestions. Most contributions require you to agree to a Contributor License Agreement (CLA) declaring that you have the right to, and actually do, grant us the rights to use your contribution. For details, visit cla.microsoft.com.

When you submit a pull request, a CLA-bot will automatically determine whether you need to provide a CLA and decorate the PR appropriately (e.g., label, comment). Simply follow the instructions provided by the bot. You will only need to do this once across all repos using our CLA.

This project has adopted the Microsoft Open Source Code of Conduct. For more information see the Code of Conduct FAQ or contact opencode@microsoft.com with any additional questions or comments.