Thursday, February 12, 2015

Defining "Big Data" in Agriculture

A recent comment on one of my articles addressing big data concerns in farm leases asked why I used the term "big data" instead of just "data." This got me thinking about the definition of "big data" and what the term really means with respect to agriculture.

Before looking at what "big data" is, we need to define what "farm data" is. When I look at the agriculture industry today, I see three different types of farm data that are being collected:
Agronomic Data.  This is information derived about activities and conditions on farm fields. Examples include soil analysis, nutrient information, hybrid selection, plant populations, and yield data.
Machine Data.  This information is associated with how equipment is functioning. Examples include fuel consumption, machine health indicators, diagnostic codes and engine performance.
Weather Data. This is information about precipitation, wind, temperature and other climate conditions. 
My prediction is that courts and legislative bodies will give each of these categories on the farm different levels of protection, as the law develops to address farm data privacy and ownership issues. Agronomic data is probably afforded the most protection under existing laws, since agronomic data is similar to a traditional "trade secret."  Equipment manufacturers would also like to view machine data as a proprietary trade secret--but owned by the manufacturer--not the farmer.  And weather data, I'm not sure anyone can claim to own that.

So with this understanding, what is agriculture's "big data"?  Here is my definition:
Big data is the ability to aggregate information to discover trends and find patterns.
Agronomic data becomes "big data" when multiple farmers upload their data to the same place, then that data is analyzed to discover, for example, that applying a certain amount of fertilizer at a certain time in the growing cycle of a certain hybrid provides the highest yield. A single farmer analyzing his own data might also discover this, but his isolated data can not tell him whether that formula will yield similar results in other fields.  Big data proves it does (or does not).

Do you have a different definition of "big data"?  Let me know.


  1. I would define Ag Big data as the classic 3V+A definition applied to agriculture. So a huge Volume of highly Variable data computerized at high Velocity level by Analytics components.

    Best example I can think of is Climate Pro computing tons and tons of agronomic data (including historic data) coupled with meteorological data in order to generate personnalized recommandation to farmers. All the 3V+A ingredients are in there and it is only beginning!


    1. Thanks Jacques for your feedback. I think you are correct that "big data" also means the ability to make recommendations for the future.

  2. Rather than “Big Data” I prefer to call individual farmer data “Small Data”. Farmers use small data to author “Crop Stories” that become chapters of their own book with every pass of the field. Just like an author uses their mind, paper, pen and computer to create a unique story…A farmer also uses their mind, land, inputs and machinery to create a unique story.

    Both of these stories are unique in a way that I really doubt two persons would independently get to the same conclusion by selecting the exact same variables along the way.

    The author has several different opportunities to turn their unique work of art into a revenue stream. In the old days, a publisher would offer the author a contract, acquire most, if not all rights to the book, and in turn assume the risk to print, publish and sell the book through book stores and retailers thus paying the author royalties from the sales. Today, the ability to self-publish using services like Amazon allow the author to retain many more rights over the final product and substantially reduce the costs print, publish and sell the book to anybody interested.

    The farmer, at the end of a growing season, has created a unique work of art called a “Yield Map”. Many “Big Ag” companies want these unique “Crop Stories” so they can turn them into “Big Data” by combining a whole bunch of it to their advantage. They will then use the knowledge gathered to sell farmers more products that are more expensive and provide more value but rarely share this value and risk equally with the farmers who knowingly or unknowingly conceded this data originally. Rather than giving this valuable creation to anyone who shows up with a thumb drive or worse yet having the equipment manufacturers automatically collect it, store it, pool it, and use it for their own value creation, we believe there is a vast growing energy and momentum for the farmers to take control of this valuable commodity and let a market determines its worth.

    What would anyone, not in agriculture, do if someone showed up to their place of business or office with thumb drives in hand to collect data? Probably call the cops!

    As of today, I have not heard of a paper, ink, typewriter or computer companies claim to own a copy of everything created using their products that were fully bought and paid for by the authors before this unique work was created. That being said, isn’t it kind of insane that “Big Ag” companies who make seed, chemicals, fertilizer and equipment all claim to have access to “their” data and are legally wrangling to lock up the rights to this farmer created unique data? Please “Big Ag” stop and take a look at how “Big Brother-ish” you are appearing and ridiculously silly this argument seems. At Farmobile we are looking to partner with the companies that share vision of creating shared value and creating additional revenue with our customers not off of our customers.



Note: Only a member of this blog may post a comment.