Big Data Comes to the Farm
Although the prospect seems far-fetched, the second-oldest profession is on the cusp of a technological revolution. Farmers have faced a long transition over the past three centuries in which their ancient art has become more and more of a science. But the pace of that change may soon increase radically, driven by forces which have been transforming other sectors of the economy as well. The watershed emergence of automation and big data is poised to come to the American agricultural heartland. Ultimately, this could be a period of disruptive innovation in farming not unlike the 70s rise of tech culture in Silicon Valley.
The overarching description of this new farming revolution is “precision agriculture.” A new product from Monsanto (NYSE: MON), which is now available to corn farmers in Iowa, Illinois, Minnesota, and Indiana, illustrates the trend.
Monsanto Assembles the Pieces of a New Paradigm
Monsanto’s FieldScripts is the result of a marriage that might have seemed unlikely, but may drive profits for the company in a wholly new market beyond its hybrid and GMO seeds and crop protection chemicals. Last October, Monsanto bought a private company, the Climate Corporation, which had been founded in 2006 by two Google alumni. Following in the footsteps of projects like Google Earth, the Climate Corporation assembled a massive database — topographical maps of 25 million American fields, combined with weather and climate data, weather simulation modeling, and soil fertility data.
The Climate Corporation had been in the business of providing those data to farmers. But Monsanto, which paid $1 billion for the company, realized that by marrying its database with Monsanto’s own deep database about its corn and soybean varieties and their yield performance under various conditions, they could potentially provide an extremely precise “prescription” for farmers.
In 2012, Monsanto had made another strategic acquisition: a company called Precision Planting, which builds farm equipment that can sow, fertilize, and reap automatically according to a precise program.
Adding all these pieces together — Monsanto’s yield data, the Climate Corporation’s soil and weather data and models, and Precision Plantings’ hardware, and you have the basis for FieldScripts: an automated, big-data-driven way to maximize yield for farmers. The program Monsanto prescribes can tell farmers what varieties of corn are appropriate for which of their fields. It can tell them which parts of their fields are likely to be more productive, and concentrate more planting there. And it can execute the whole program with machinery that is largely automated — often needing human babysitters rather than operators.
But Does It Work?
On average, farmers who tested FieldScripts over the past two years of its development saw corn yields rise some 5 percent — by about 5 to 10 bushels per acre. That’s an impressive feat.
And of course, the system generates still more data as it is deployed. Think of a crowd-sourced application such as Google’s traffic-savvy Waze, which crunches user data (speed and location) to help drivers avoid congested roads. The more drivers use it, the more useful it becomes to all of them, because the more accurate its picture of current traffic conditions.
More Use, More Data, Higher Effectiveness
FieldScripts will also continue to generate data that when refined and applied, may allow the system to create even more incremental gains for farmers. Monsanto believes that precision agriculture will ultimately allow American corn farmers to increase their average yield by 25 percent. This may be why it thinks that the total addressable market for farm data analytics and applications could be $20 billion annually.
There are other competitors of Monsanto’s size entering the space as well. DuPont’s (NYSE:D) Pioneer division, like Monsanto a heavyweight in crop protection and hybrid and GMO seeds, has launched a product called Encirca, which is similar to FieldScripts though not as ambitious in its data sweep. Nevertheless, it believes that the system could generate $500 million in revenues.
Both Monsanto and DuPont are viewing the space as an area of key importance. The Climate Corporation’s vice president of marketing said of FieldScripts, “We view this as a platform that is as important to Monsanto as biotech.”
Proprietary Data?
The fact that these systems, once deployed, continue to generate data, creates some questions in the minds of farmers, as expressed by groups such as the American Farm Bureau.
For one thing, FieldScripts, for example, gives farmers a program telling them which seed will perform the best in which piece of ground under which conditions and with which crop protection products. But will it only be telling farmers about which Monsanto seed varieties, and which Monsanto products, will work best? Monsanto says no; the system will suggest a competitor’s product if it would work better, and that if the system functioned otherwise, it would lose credibility.
With the firm deploying proprietary data, though — and harvesting ever more of those data from farmers — some are skeptical. Some fear that with proprietary systems, farmers will get “locked in” to one of the big competitors, since data will not be portable from one company’s system to another.
Who Owns the Data?
This raises another keen issue for farmers, who apparently are as concerned about privacy and data security as any other segment of the population. They want to know who will own the data they generate, as year after year their data feed back to Monsanto or DuPont increases the company’s trove and becomes more and more valuable. So far, companies have agreed in principle that farmers should own their own data — that those data should not be sold or given to third parties, or used for other purposes than the farmers’. But contracts have not consistently reflected those principles — so farmers are forming co-operative groups to advocate for their data.
Farmers’ caution may be due in part to the poor reputation that Monsanto has acquired for the aggressive manner in which it has pursued what it regards as infringement of its intellectual property. Farmers may think that a company that has played hardball in such matters before may also be willing to play hardball with them.
A Swarm of Startups
Both of these issues — data freedom and data ownership — show how this nascent transformation of agriculture could be similar to the Silicon Valley culture of the 70s which gave birth to a host of hardware and software innovators, permanently removing computing from the private domain of the previous generation’s giants such as IBM (NASDAQ: IBM).
Although the field is now dominated by big players, the whole thrust of precision agriculture may mitigate against them, and in favor of upstarts.
Farmers and Capital Expenditures
For one thing, farmers are notoriously wary of big capital expenditures, and so far, the precision agriculture offerings are in the same mold as other large, expensive machinery. It won’t make sense for many farmers — and even if it could boost yield, if the expenditure would be too much of a stretch, they won’t invest. The wave of farm consolidation that some readers may remember from the 80s “Farm Aid” era instilled some bitter lessons.
But this technology intrinsically lends itself to smaller machines, to homespun solutions, to the open-source
The Dashboard of FarmBot, an Integrated Open-Source Precision Agriculture System
“maker” culture (think “farm hacking”), and to small, innovative startups offering products along those lines.
Entrepreneurial farmers are already leveraging off-the-shelf tech components and ubiquitous smartphone adoption to create sensors, collect data, and feed them alerts. One such farmer we read about was ahead of the game — creating a DIY drone to fly over his 250-acre family farm monitoring the light reflected off plants’ leaves. Those data can be crunched to determine the “normalized difference vegetation index,” a value that helps determine which plants and which areas of a field are healthy. Although such agricultural drones have been in use in Japan for some time, they are not yet in wide use in the U.S.
These individual makers and hackers may be the forerunners of the nimble startups that ultimately take over the landscape — just as upstart Apple (NASDAQ: AAPL) and Microsoft (NASDAQ: MSFT) rose from garages to conquer the territory of IBM, and just as the ubiquitous pocket-sized smartphone is the ultimate progeny of the mainframe.
Big Data and the Return of the Small
Instead of giant machines moving over huge swaths of land, the future of agriculture may be hosts of small machines doing precision work in precision areas. They may be like fieldworkers of the past, but they won’t be humans — they’ll be robots. Just not giant ones. This also is the direction that agriculture’s new developments are pointing in.
Good for Efficiency… and Good for the Environment
If you have a twinge of regret at this vision of a big-data driven automated agricultural future, with a romantic notion of farming being challenged, it’s worth noting as well that these trends have the potential to markedly improve efficiency. And that doesn’t just mean high yields; it means lower inputs, especially of agricultural chemicals. Precision application, in this case, driven by data about what farming methods are effective where and under what conditions, could mean the application of less pesticides, herbicides, fungicides, and fertilizers — since they will be more likely to go where they’re needed. Farmers will like the efficiency and the increased yields.
