𝗙𝗼𝗿 𝘁𝗵𝗲 𝗳𝗶𝗿𝘀𝘁 𝘁𝗶𝗺𝗲 𝗶𝗻 𝗵𝗶𝘀𝘁𝗼𝗿𝘆, 𝗮𝗴𝗿𝗶𝗰𝘂𝗹𝘁𝘂𝗿𝗮𝗹 𝗳𝗶𝗲𝗹𝗱 𝗯𝗼𝘂𝗻𝗱𝗮𝗿𝗶𝗲𝘀 𝗵𝗮𝘃𝗲 𝗯𝗲𝗲𝗻 𝗺𝗮𝗽𝗽𝗲𝗱 𝗮𝘁 𝗴𝗹𝗼𝗯𝗮𝗹 𝘀𝗰𝗮𝗹𝗲. I've been collaborating with the Fields of the World (FTW) organization for over a year through Taylor Geospatial and it's finally released publicly in cloud storage for anyone to use, at no cost.

Play with FTW Explorer and download field boundaries for your AOI

Read the accompanying paper here:

Amazing work! Congratulations on the release

This is genuinely huge! For the first time, smallholder farmers, agtech startups, and researchers can access field‑boundary data at global scale, for free. No more relying on incomplete or expensive proprietary datasets. What’s the first use‑case you’d build with this?

We provide 2 years of boundary geometries and pixel-level predictions as well so computing field change seems like a great next step.

If FTW could integrate satellite imagery frequency (Sentinel‑2, Landsat) with those boundaries you’d have a dynamic crop intelligence layer that insurers commodity traders & carbon credit platforms would pay heavily for. Are you planning to add temporal frequency?

You should cite Christoph Rieke (@chrieke). Together in Germany in 2016, we pioneered the application of U-nets to Sentinel-2 for agricultural field segmentation. I was his advisor for the AI component.

@chrieke Could you provide a PDF so I could read more? It seems to not be open access

@chrieke You bet! The code and thesis PDF are available here:

We have looked into this and it's great, and quite comparable and in some areas better than the inhouse model we had built! Very valuable work.

Glad you’re finding it useful! Please reach out if you have any other feedback!

Absolutely will!

How well does it work for small fields in places like India?

The model runs on 10m Sentinel-2 imagery so your mileage may vary for small fields. Please see the paper for more details on individual country evals

Incredible release! Global field boundary data at this scale changes how we develop precision ag tools for challenging regions. Open data drives real innovation.

Unreal

Wow, this is amazing!

@giswqs Great Work.

@urbanhobbit Thanks İsaac Corley

@giswqs Es hermoso. Mejor que VARDA

This is great

Nice work! Be awesome to overlay economic indicators. For example agricultural output per capita

@Julian_H0ffmann So more like crop boundaries?

Thanks Isaac and Congratulations, this is really impressive.

@ThruTheHayes @TyrantsMuse osint tool

Cc:@nudpiedo

Great Work!

this is pretty incredible stuff -- love it

very cool. Can we integrate the data into PlantVillage platform. We would want to then track carbon increase or regen ag

Absolutely! The data is available on Source Cooperative S3 buckets in Zarr, COG, GeoParquet, and PMTiles format for 2024 and 2025. There's a README for how to load and slice it. Taylor Geospatial and the FTW org are happy to discuss and provide engineering support if needed!

many thanks

Great release. Went through the paper, sent you an email.

how's you olive field coverage in the west bank

@yohaniddawela

@grok what is all this tell me

@WELLO

So what kind of data are you extracting here?

Congrats, but I just wanna understand how it works? Like I am in Pakistan and as a young farmer how can I used it at my place where tech. is not very much advanced?

I tried to practise witj my farm in dominican replublic but honestly i do not know what case use for my farm this has- help