I'm the lead software architect at a major agricultural and construction equipment manufacturer. I deal with this stuff every day.
I'm not aware of any efforts to make collected data available but what I can say about the company I work for is that we largely view the data the farmer or operator's personal property. We don't transmit it back to any data center presently but we do have ways to copy the captured data off to a USB stick for post processing which a number of 3rd party precision farming applications are on the market to read it.
What I can say about the data is that it's voluminous. Our software collect all kinds of data besides just precision farming data. For example, CAN bus faults, GPS history (vehicle guidance), weather conditions (temperature, humidity), etc... And the way the data is collected follows no industry set standard - it's pretty much proprietary formats.
Also, much of our data collection is stored locally in a binary format to save space and improve performance. We have tools that move the data back to a flat file/text format for post processing.
I'm not involved with the 3rd parties that use the data but my guess is that someone has shared with them the formats and such.
Yes, this is a common problem to all industries. Formats can be figured out or obtained from the OEM, then a gateway is created to transform into a digestible format (my guess some XML schema). In the article, the author states his company [1], where in the tech section he describes this very scenario.
I've dealt with something similar, working for a small publisher in niche markets. We've got dozens of customers, some on third party platforms, and others with their own proprietary data formats. I built a platform to convert their data to our own format. Usually that part is easy. Mapping their categorization to our own is the time-consuming part.
>Farm data is expected to be a $20–$25 billion revenue opportunity, but we haven’t yet determined how the data can be collected, structured, stored and shared, let alone monetized.
One can know that a system can be improved by a certain amount without knowing who will pay for it; knowing how much it can be improved and how expensive it currently is, is enough to place an upper bound on how much someone would pay to improve it.
> Farmers tell me, “Look, I know that my data has value because everyone wants it. The part that frustrates me is that companies want my data in return for some undefined ‘value.’ Meanwhile they are often upselling me for additional services that would not exist without my data."
Sounds like a good description of what's been going on across the economy for 20 years. Farmers are asking the same question many others are asking: how can I keep more of my own data's value for myself?
1. This is tough, because the value of a single data point ("my data") is pretty much zero. The value is in the aggregate. Of course, without any data points, the value of the data is then zero. So the sum of the value of the individual data points does not equal the value of the data as a whole.
2. There's value to you in my ability to sell you more things. Which sounds nonsensical, but hear me out. If I'm able to sell you, it means you see value in that product or service greater than its cost. I would not be able to create that value for you without your data. So it's a help-me-help-you situation.
For #1, I'd imagine that farmers already belong to (or could form?) a farming or agriculture association, and one of the [additional] responsibilities could be data collection, anonymizing and selling to other parties. The association, similar to a labour union, could also potentially hire a 3rd party to do this work on their behalf provided they still retain ownership over the data, but at least the overall coordination and aggregate value would be captured on behalf of the farmers and not the vendors.
#2 would be fulfilled more easily by the association provided there was enough true representation of member interests and concerns.
Well, but neither the farmer or agriculture association are capable to actually collect the data - the data is generated as a byproduct when all kinds of systems do work on the farmer's fields, but it is generated by and within all these third party systems.
Aggregation and cleaning up of all that heterogenous data would be even harder, and that is definitely not the task for anyone who doesn't make their core business (I mean, none of the involved big companies is able to do that yet) creating and maintaining that capability can be practically done by a couple competitors worldwide, but not by every regional association duplicating the effort.
#1 - You're talking adjacency (moving up the value chain), which might not be a bad idea, but is the incremental value worth it? It might be: I don't know. But I know that it implies a level of sophistication and focus that may not be present.
#2 - I don't know if there's a distinction there. Either the value is lower (because the data doesn't include anyone outside the association) or dollars just flow in a different way (e.g., increased association dues for increased "free" services vs. fee-for-service).
Farmers join a local Co-Op, say "Iowa's Data Co-Op". The Data Co-Op then sells to the highest bidder, then after a small management fee returns dividends to the participating farmers.
The Data Co-Op partners with equipment integrators to build data collection devices and data transformations.
Local farmer groups are already a thing, where farmers get together to share research, look at properties with trials running on them etc. [1]
The question your have to answer before trying to sell any of this "info", is who, other than the farmers themselves is going to find it valuable enough to buy?
I could see myself dedicating my life to making software to help farmers organize their farm data and sell it. Anyone know of companies that do this?
edit: I'm a full-stack application developer with a lot of experience working with non-techy business owners and proven success as a problem-solver. I'm interested in getting into this industry as a technical project manager. Email me if you're looking for such a person (it's in my profile). I'll be in contact with the companies in the comments below, too. Thanks!
Yes, FarmersEdge (disclosure: I know the CTO) is a big player. I think, in a roundabout way, Monsanto is one of their competitors through acquisitions, though I'm not sure of their market placement wrt FE.
I think ag is an underrated software industry. It's ripe for disruption, but you need to do it in a way that makes sense to farmers, which can be tricky for at least a couple of reasons!
It is a commodity business with very thin margins. Farmers for the most part have zero pricing power for their commodities. Whatever you propose must "pencil out", as farmers say, and it needs to be a clear benefit. It should reduce or option-away risk, not add to risk.
Professional farm management companies might be a good place to start, because they have business analysts that have the time to understand your product and can deploy at scale across multiple clients.
Hertz farm management (Iowa, Illinois, Nebraska) is one example. Typical client is the children that inherited dad's farm but don't have the time and expertise to manage the land, which is typically leased to farmer-operators.
Most of the software on market is basically data collecting apps (be it manual or automatic data entry). What you get out of those apps is just another piece of (already abundant) information. You might get good insights out of it, but most farms are seriously understaffed and there might not be enough time left for doing another data entry job or analysing it.
I run small web app for farmers (http://vitipad.com english website is not really up to date as I am active mostly on Czech market at the moment.
I am not selling the data, the app is used for storing their data and generating the reports for various govt agencies.
Selling to farmers is not easy, they don't have much time for administrative tasks, they are usually older and not really into new stuff anymore. Most of the software on market was created in 90s and there was quite a lot of acquisitions lately (mostly to wipe the market).
I can't wait to see new crop of tools entering the market, but mixing correct set of features will be hard.
Check out Agrible. agrible.com and morning farm report which is our flagship delivery platform. Instead of focusing on data collection/organization (a digital farm notebook) we are focused on predictive analytics which is essentially decision support for the grower. We combine high res weather / soil / field info and have a built a model that literally is growing virtual crops and we are using our model to predict yields at the end of the growing season!
FarmersEdge is definitely big in at least exploiting the data for higher yield through several techniques including avoiding disease, rotating crops, weather preduction, etc...
Really interesting work, and in a domain that is just opening up to the idea of using data. They're definitely recruiting and open to remote work if you're serious. I used to work for a startup they acquired.
There's loads of opportunity in helping farmers gleen new insights.
Look at Quantified Ag, making fit bits for cows. http://quantifiedag.com/ Groundbreaking stuff, which came about from a simple discussion. It seems boring on the surface, but there's some cool stuff going on in the sector.
Check out our startup Pycno. We do a smart sensor for farmers, since getting the data out of the farm is the first problem. On the software side we have a lot of analytics and predictions that the farmer can benefit from. We are in HAX, here in Shenzhen!
Precision Planting HQ is 2 corn fields away (~5 miles) from the house I grew up in in rural Illinois. Maybe I shouldn't have moved to the west coast to chase startups...
Check out climate.com
We are building the tools for farmers to use all these data.
We have offices in SF, Seattle, Chicago and St Louis if you are interested.
What I think is going to be most interesting is companies/banks/insurances agencies being able to gleam information without going through the farmers at all. Not by subverting them and going to the machinery manufacturers, which is already happening, but just by using other sources of information, such as satellite imagery.
It's already possible to predict yield potential through biomass/NDVI monitoring via satellite, and given that imagery from a source such as LandSat is freely available, or something like Planet isn't that expensive once your a bank/insurance agency I'm certain they must already be using.
Once you can predict yields reliably via satellite, you can start to predict supply for certain types of grain in different regions, or to a lesser extent, countries as a whole. This data is useful to investment banks and insurers, as well as organisations such as CBH[1], that handle the entirety of Western Australia's grain export. They currently determine potential yield by sending out a grower survey, and asking the farmers to predict yields themselves.
I think the idea of anyone "stealing" our data is overrated. The data from our farm[2] is specific to our farm, and as long as where not farming in such a way that is detrimental to our partners (such as banks/suppliers), such that we were trying to lie about yield potential, or resource usage, I'm not sure what value could be stolen by anyone with our info. In fact we openly share all the trial work we do, so that everyone in the community can learn and benefit.
This seems to confirm just how important the price of data is for ML & AI based startups. Methinks there is a market alone for data companies, not even those that are necessarily doing anything productive with said data, just collecting it.
Also raises the question of whether its possible to bootstrap a data science, ML, or AI startup, given the high expense of purchasing that data.
There is actually a thriving industry in data collection and these range from companies that provide tools for you to do your own data collection and those who will collect the data for you (or provide pre-made datasets). Full disclosure, I work for one such: Scrapinghub where we provide tools, platform, and data services.
But surely we already know that there's a huge market for information storage. Cartographers are the big ones, who can produce maps to sell to the public, but also often produce custom excerpts of their datasets for building planning surveys, etc.
The other group are information brokers, like those behind credit check firms who again mainly exist to collect and sell data.
I'm not sure the farmer should be getting in the way of data collection. Regional and global optimizations can be mined from a large enough set of ag data. Do we really want to nickel and dime research this way? Its a fair question.
I know; there's a sense of ownership. But farmers are just one participant in the chain of data management. They just happen to be the one person residing physically over the land. So they could charge rent on data. But should they?
The point is that everyone except farmers are profiting from the data the farmers are helping to collect. Perhaps the better question is "why should everyone except farmers get to nickel and dime research in this way?"
I didn't read that. I read that no money was changing hands; just agreements. And of course the farmers benefit greatly; that's the whole point of it all, to make better yields.
Why do you think increasing yields greatly benefits farmers? If the pencil industry doubles its output, do you imagine their profits will increase twofold?
The phrase "better yields" is exactly synonymous with "increased yields". A crop's yield is its volume, measured in bushels. There is literally no other way to interpret this.
Citation: I grew up on a farm, and my family has farmed for at least 200 years.
EDIT: This is a particularly weird reason to downvote me...
And that'd be pedantic. Better can be more cost-effective, cheaper in time or money or equipment or quality of land required. I grew up on a farm, live on one now, my family has farmed for 300 years, so I guess I win?
I'm not aware of any efforts to make collected data available but what I can say about the company I work for is that we largely view the data the farmer or operator's personal property. We don't transmit it back to any data center presently but we do have ways to copy the captured data off to a USB stick for post processing which a number of 3rd party precision farming applications are on the market to read it.
What I can say about the data is that it's voluminous. Our software collect all kinds of data besides just precision farming data. For example, CAN bus faults, GPS history (vehicle guidance), weather conditions (temperature, humidity), etc... And the way the data is collected follows no industry set standard - it's pretty much proprietary formats.
Also, much of our data collection is stored locally in a binary format to save space and improve performance. We have tools that move the data back to a flat file/text format for post processing.
I'm not involved with the 3rd parties that use the data but my guess is that someone has shared with them the formats and such.