Facilitator Ben: I am interested in having people talk about their thoughts and works in terms of interoperability.
How do our systems relate, how could they connect, and how do they need to relate?
Birgit taking notes: please let me know if I got something gravely wrong so I can update the notes 
Shep: Small grant to build out FarmOS additional tooling, maybe data cleaning, clean out pesticide data or identify what is missing.
We need to pick a data solution, nothing is a perfect fit, so we get to figure out what systems are interoperable.
Most growers work with several certifiers and commodities, which means to have to work with many tools.
Evan: Is it different data formats or is it actual tool compatibility? Both.
Greg has been pushing for common schemas and ontologies but it’s unclear where that’s at.
Ben: Working on common farm convention - data transfer standard based off of FarmOS’ model. Turns into something you can use for data validation.
Sam from Clever & team have done something on standards.
Jennifer: Wanna hear people talk about this because I need tools to be interoperable.
“Recoding America” is a really important book here.
It is slow to load data because of the server capacity.
Ben: First thing is documenting the standard, and that draft is currently built on FarmOS,
2nd is a having that rolled out as a data standard,
3rd is tool to put data in one format and it gets exported into another one.
Shep: LiteFarm is here, so a good start would be to find a data format between FarmOS and LiteFarm.
—> possible follow-up item to talk about for Ben & Kevin/Divya (not present in session)?
Evan: What types of data are we talking about?
Jennifer: Soil map, topography map, conversation map, plan, direction of rows, tillage type, crop type, historic crop data for insurance purposes.
Shep: Sometimes it’s fillable PDFs, sometimes it’s in spreadsheets, ultimately it comes back to farm management data.
Often it could be pulled out of a farm management system to then be put into a grant application but that’s currently not possible.
Sometimes the data formats change from year to year, e.g. now we also want tillage depth to be recorded.
Jennifer: Yes, fencing stuff also changes all the time. The farmers often don’t have access to these platforms, they are governmental.
We often comb through a 4000 line spreadsheet to find data, e.g. the slope, low-till, the direction of lines all per individual field - in a conservation plan.
Shep: And often it’s not even spreadsheets but a fillable PDF.
Jennifer: So it needs a lot of experience on the part of the person looking at the data, e.g. retired NRCS people with maximum experience.
Shep: Ideally we’d have a data collection place, and if data is structured in a certain way in the USDA system, then you could get to interoperability.
Jennifer: Yes, so it has to be a really big deal and think “erase everything- what if we didn’t have this system?” to redesign it.
Ben: Do these systems even have ways to import data? You’d still be copying back and forth.
Shep: Yes, potentially. Public good idea: We can go to funders or other organisations who have their own proprietary systems and say that we care about making this more accessible for growers and we want to establish a data push-pull system with a schema to match.
“Better cotton” is a fund that has their own data collection system - so we can talk to them about setting up a schema with an open source convention, then we can build something.
Say a grower is already working with John Deer, what is the potential to have an open source schema that would allow to interface and allow data exchange with their software.
Evan: What is the chance of John Deer being open to that? Birgit: We’d probably have to find a win-win for them and how that brings them value as well.
Ben: Where we overlap is the soil health.
MODUS standard for lab methods. A lot of methods aren’t measuring soil & carbon but something else that will then calculate it.
MODUS gives a name to all those different methods.
AgGateway now holds that standard as steward and we think about how to make it useful and how to get soil labs to adapt it.
We are figuring out how to have that help farmers choose which lab to use and which methods to ask for.
Jennifer: Why not look at the farm practices instead of soil carbon indicators? It takes more work to farm regeneratively, and there’s no meaningful carbon market there.
I think it’s a waste of time & money that could be of better use somewhere else, namely directly with the farmers.
Shep: Soil data is about 2 cents per 2 pounds of cotton when we bring them back 60 cents on the pound, so it’s worth it for them. We can’t get 60 cents back to the grower without that test.
Ben: interesting larger question of if the USDA runs everything through one server then maybe they can create one proper new system to run everything through.
Jennifer: If we used our local participation system well… the whole qualitative data work is hopefully good for USDA investment into community scale agriculture - and they gave us a million for our watershed within four months. It’s a sort-of functional system - but no one reads the law, so our district was the first one to send a report like this.
We have put it into our conversation action plan to get chemicals out of our environment - why do we spread pesticides when establishing pollinator habitats? We need an alternative.
Mechanical buffers take a lot of manual labor - it’s a shift in what we value.
The only way I see to make that change is to go to USDA and say we talked to 1 million local farmers and this is what we want and need.
If we can have a policy coming out of this community with strong qualitative and quantitative data, we can make them listen.
Ben: How much data would it take?
Jennifer: I can share what we collected — these are our reports from the Local Working Group in our state that we utilized ChatGPT and survey stack to turn qualitative data into quantitative data and policy recommendations: https://www.whiterivernrcd.org/reports
And here is the report from the same process in the USVI: VI Local Working Group Data 2.2024 - Google Docs