AKS Cluster deployment, field boundary identification, irrigation detection, and data ops

[renatolfc]

Dear FarmVibes.AI users, hello! 👋

We are happy to announce we've just merged what are the results of four months of work on the FarmVibes.AI platform!

In this release, we've implemented many features you requested over the past few months. A highlight is data operations, which will allow you to delete data produced by workflow runs.

In this release, we've migrated the installation program from bash to python, and that's the most disruptive change in this release.

Now, to update your existing clusters, if you cloned the repo, do a

git pull

to update your local copy. Then, you will have to update (or install) the vibe_core library, with

pip install ./src/vibe_core

Finally, you can upgrade your cluster by running

farmvibes-ai local update

Thank you for using FarmVibes.AI!

---

Here's a summary of the changes from the last pull request:

Backend:

• [🎉 NEW] We added support for setting up a remote FarmVibes.AI cluster on Azure Kubernetes Service (AKS). Refer to our Remote Cluster User Guide for information on the architecture, services and setup instructions.
• [🎉 NEW] We added a new DataOps service to the cluster responsible for managing and keeping track of assets related to workflow execution. It is now possible to delete all data produced by a workflow run, as explained in the Data Management section of the documentation.
• [📈 IMPROVEMENT] We have fully rewritten the FarmVibes.AI script in Python. It now supports running in multiple operating systems (Windows, Linux, and macOS), and local and remote cluster setup. For more information, run the farmvibes-ai with the --help flag.
• [📈 IMPROVEMENT] We made several backend improvements on overall cluster stability and workflow state tracking, addressing issues #104, #110, #112, and #113.

Workflows:

• [🎉 NEW] We added a new farm_ai/water/irrigation_classification workflow that estimates 30m pixel-wise irrigation probability maps using Landsat-8 and Copernicus DEM data.
• [🎉 NEW] We added a new ml/segment_anything/point_prompt_sam workflow that employs Segment Anything Model with user-provided points as prompts to segment Sentinel-2 raster images.
• [🎉 NEW] We added a new data_ingestion/user_data/ingest_smb workflow that allows users to import their own rasters stored on a SMB share into the cluster storage.
• [📈 IMPROVEMENT] We improved the farm_ai/agriculture/heatmap family of workflows to allow users to create nutrient heatmaps with less samples (~<50 for 100 acre area, in comparison to 600 samples in the previous version). The enhancements help identify the candidate locations for sampling nutrients information and use them to generate the heatmap.
• [📈 IMPROVEMENT] We improved the farm_ai/agriculture/weed_detection workflow to operate on rasters with an arbitraty number of bands (with or without an alpha band). Related to issue #102.

Notebooks:

• [🎉 NEW] We added an irrigation classification notebook that demonstrates the use of the farm_ai/water/irrigation_classification workflow.
• [🎉 NEW] We added two notebooks (segmentation anything and FarmBoundary) showing the integration of Segment Anything Model and FarmVibes.AI.
• [🎉 NEW] We added an identification of optimal sensor/sample location notebook that demonstrates the use of the farm_ai/sensor/optimal_location workflow.

Client:

• [📈 IMPROVEMENT] We added the ability to simultaneously monitor multiple workflow runs. The client now has a monitor method that receives a list of run objects, and displays a summarized monitor table with each run's information.

Documentation:

• [🎉 NEW] We added a new Data Management section to the documentation, explaining how FarmVibes.AI caching service works and the data management operations available.
• [🎉 NEW] We added a Remote Cluster User Guide with an overview of the remote setup process for FarmVibes.AI, its main components, configuration options, and customization possibilities.