Many pieces of farm machinery have been capable of collecting and storing agronomy and production data for more than a decade. Extracting usable and attractive tools for producers from all that information has proven challenging.
The development of cloud computing and data storage, as well as the internet of things in which objects including farm equipment collect and upload telemetry, have automated the collection and processing of data.
This was a necessary step before many growers considered using these datasets to help in their farm management decisions.
Numerous data points are now automatically brought into agricultural modelling and analysis programs, including rainfall, temperature, soil saturation from soil sensors, seeding date, varieties, row spacing, pest forecasts, market trends, input, rotations and production.
Growers can now dive into their own data to see the economic analysis of on-farm trials of products or fertility programs and how they are affected by weather, for example.
Smart farming has become a bit of a catch phrase in the industry, but it basically describes farming techniques that use these information and communication systems to help make production decisions.
Machine learning, in which statistical techniques give computer systems the ability to learn from data without being explicitly programmed, is just beginning to be implemented in some agricultural data management systems, and it’s set to help make farm management decisions easier.
Big data enters the picture in the analysis of production information from not just one farm but across dozens, hundreds or even thousands of farms to see agronomic trends.
Terry Griffin, who works in Kansas State University’s agricultural economics department with a focus on big data in agriculture, said if you have information from only a few farmers, you cannot clearly understand how the farmer’s management affects the system.
“But when you have hundreds and thousands of farmers on millions of acres, we can start treating the farmer as a variable in the analysis instead of the constant,” Griffin said.
“That’s the way I would define big data in agriculture.”
He said there are competing philosophies when it comes to big data at the farm level.
Some farms believe if data from their farm is shared into a community and analyzed property, they’ll be able to make better decisions based on that information — so it’s a win-win situation.
Other farmers believe they are at a disadvantage if their data is shared because the information may be used against them. These farmers may be concerned that growers competing for the same farmland could use the information to their own advantage.
“Some of these fears are real, some are perceived,” Griffin said.
“Will a landowner get access to some of this information and charge me more rent, or use this information to negotiate more rent from a competing farmer who might farm the land next year, or if my retailer might get this data and charge more for the products that make me more money, or will Deere use this to price their equipment such that I can’t afford it but another farmer can, or will the government start requesting proof that I am not polluting environmentally sensitive areas?” he said.
“There is all this fear about privacy, and these are things that people really do worry about.”
He said a cost-benefit analysis needs to be performed before a grower decides to join a company or group that uses on-farm production information.
“As an economist we always compare the benefits of a decision versus a cost of a decision and sometimes the costs are real (and) sometimes they are perceived,” Griffin said.
“Sometimes they have a fear of something that may happen, although it never has happened. That’s the category I put some of these things into.”
One of the more obvious risks of joining a big data-based company comes from the high amount of turnover in these companies — there is a good chance a farmer will choose a product that will be out of business in a few years.
“If you chose poorly you will need to do something again in a few years — join another company, go through those hoops with a different platform with different processes,” he said.
“Not only that but the ability to exclude others (from your data) is even more diminished because the data that they share now potentially could be in the public domain, or at least shared freely or available to many different groups because of the company going out of business.”
Griffin said farmers receive many opportunities to join the dozens of companies who are trying to obtain their data, but if a clear benefit at the farm level isn’t apparent now, their best bet is to wait until they find a company or community that has immediate benefits that outweigh any of the perceived costs.
He said growers should look closely at the benefit side of the equation because there are few examples of how big data has benefited the farmer at the farm level.
Like a lot of new ideas, he said, big data falls into the area of overemphasizing the short-term benefits, which can be over-sold at the farm level.
However, long-term benefits from working with companies that use big data to improve on-farm efficiencies are difficult to see now but will likely be significant.
Griffin is convinced big data is here to stay, not only for making growers more efficient but also to improve management and to help growers tap into speciality markets.
“This is not a passing fad, it’s not going away any time soon,” he said.
“And farmers need to be prepared to participate in the farm data world. Going off the grid is probably not going to be sustainable in the long haul. The sooner that they acknowledge the reality here, the sooner they will be able to start taking advantage of it.”