That's
right: Monsanto is making a big move into big data. At stake is an
opportunity to adapt to climate change by using computer science
alongside the controversial genetic science that has been the company's
signature for a generation. Data stands to benefit Monsanto's bottom
line, too: In its
2013 annual report,
the company blamed lost profits on knowledge gaps about both the
climate and its customers' farming practices. And information services
could even help Monsanto get its foot (and its seeds) in the door of
untapped global markets from Africa to South America.
Seeds of a data company
Whatever your feelings about Monsanto, it's hard to argue that the
company isn't paying attention to climate change. When I met Fraley in
New York in September, he explained that since he joined the company in
1981, Monsanto scientists have observed corn production belts migrate
northward by about 200 miles. That means traditional strongholds like
Kansas are becoming less productive, while new markets for Monsanto
products are opening in places like North Dakota and southern Canada.
But for Fraley, who has spent his career digging through the minutiae of
microscopic nucleotides, the most interesting trends are emerging on a
much smaller scale.
"Just a couple degrees difference changes when insects will hatch, or
when diseases will break out," he says. "So that puts a real premium on
modeling microclimatic conditions, so you can become predictive on not
only which field, but which part of a field should someone be looking
at."
A screenshot of the Climate Basic app from my iPhone in October shows conditions on my family's farm in Iowa.
Last year, Monsanto made a major investment in big data analytics when it paid
$930 million
to acquire Climate Corporation, a San Francisco tech firm whose
original business was selling crop insurance to farmers with rates set
by some of the most detailed weather data available anywhere. These
days, Climate Corp.'s flagship product is a smartphone app called
Climate Basic. The screenshot to the left—from my iPhone, taken back in
early October—shows my family's corn and soy farm in Iowa. You can see
each of five individual fields highlighted. There are 30 million
agricultural fields in America, and the app has all of them, mapped with
soil and climate data to a 10-meter-by-10-meter resolution.
The app knows our fields' real-time temperature, weather, and soil
moisture, and what we can expect on those metrics for the coming week.
The green tractor tells me Saturday is the best day to work the fields.
If I were to input data about what kinds of seeds I planted and when, it
could tell me when to harvest them and how much yield to expect. A
premium, paid version of the app includes other detailed
recommendations—for example, how much water and fertilizer to use.
That advice, says Climate Corp. CEO Dave Friedberg, represents a
fundamental shift "from intuition-based decision-making to analytical
decision-making," combining real-time climate data with records from
Monsanto's trove of field trials.
"Ultimately all of this is the digitization of physical phenomena, and using that to better predict the future," he says.
Sprawling databases have long been an essential item in the Monsanto
toolkit. Locating the genes for favorable traits in plants—drought or
insect resistance, for instance—so they can be bred into new seed
varieties requires sifting through the billions of base pairs in a
genome, which is one reason why biotechnology has grown in tandem with
computer processing power over the last two decades. Since the '80s,
Monsanto has amassed one of the world's largest agricultural databases,
gleaned from the results of countless field tests of countless seed
varieties under countless experimental conditions. That's what made
Climate Corp. such an attractive investment: The opportunity to
integrate that company's climate database with Monsanto's seed data.
More than a third of US farmland is now cultivated with guidance from Monsanto's climate data.
With Monsanto now at the helm, a team of a dozen Climate Corp.
researchers are working full-time to pull climate data from government
satellites and weather stations, university research sensors, and any
other source they can find. This gets fed through a pipeline of data
analysts and software engineers and emerges at the other end as Climate
Basic.
The payoff for growers can be huge: Monsanto estimates that farmers
typically makes 40 key choices in the course of a growing season—what
seed to plant, when to plant it, and so on. For each decision, there's
an opportunity to save money on "inputs": water, fuel, seeds, custom
chemical treatments, etc. Those savings can come with a parallel
environmental benefit (less pollution from fertilizer and insecticides).
These decisions can also help farmers make money by squeezing more
yield from the same acreage. This year, as
my colleague Tom Philpott reported,
high input costs and low commodity crop prices have led Midwestern
farmers to lose $225 per acre of corn and $100 per acre of soybeans. Big
data, Friedberg says, reveals how seemingly small choices—a four- or
five-day difference in planting time, for instance—can have a
significant impact on the final harvest and on farmers' bottom lines.
A new tool for climate adaptation
Precision agriculture is also an essential climate adaptation
strategy. Monsanto's data tools could become invaluable for farmers
struggling to cope with changing conditions, says Rebecca Shaw, a
scientist at the Environmental Defense Fund who studies the link between
agriculture and ecosystems. With less water available, "we can't afford
to be wasteful," she says. "It's really important that we get better at
understanding what the crop needs and when, and apply only that."
The promise of data-driven farming, Shaw says, is to streamline the
whole process and get the same or better output while cutting back on
inputs.
Of course, data is a two-way street: Every new user of Climate Corp.
software is a new source of real-time information for Monsanto about its
customers—what kinds of products they're using and how much they're
producing (i.e., how much money they're making). Matt Erickson, an
economist with the American Farm Bureau Federation, says
farmers need to proceed with caution
when they sign on to share proprietary data about their business
operations with an outside party like Monsanto. The key, Erickson says,
is for farmers and Monsanto to be on the same page when it comes to
deciding if and how that data will be shared with third parties—other
retailers, for example, or crop insurance companies. It's just like how
we all want to be kept apprised of the ways Google or Facebook use our
data.
"We can't afford to be wasteful," says Shaw.
"It's really important that we get better at understanding what the
crop needs and when, and apply only that."
"The big thing we're striving for is transparency," Erickson says.
"Making sure farmers are aware of any secondary or tertiary uses."
Climate Corp.'s
privacy policy
dictates that farmers continue to own their data after it is shared,
and that data won't be used for purposes not explicitly approved by the
farmer. Still, Fraley says he envisions using data to market custom
products and services to farmers; a farmer whose crops are suffering
from a disease or insect infestation might get an ad for appropriate
chemical treatments, for example.
"There's a huge opportunity on the marketing side," he says. "We're just exploring it."
Privacy concerns aside, American farmers seem to be buying in.
According to Friedberg, prior to the Monsanto acquisition, less than 10
million of the 161 million acres of US farmland were being farmed with
help from Climate Corp. software. Today, that number has grown to over
60 million. In other words, more than a third of US farmland is now
cultivated with the guidance of Monsanto climate data. And it's
continuing to grow rapidly: According to Climate Corp., the number of
Climate Basic accounts has grown from 30,000 to over 70,000 just since
the spring of this year.
"The information itself becomes the business"
For Monsanto's data push, the US market is only the beginning.
Efforts to increase crop yields are especially important in the
developing world. The global population is projected to swell to more
than 9 billion people by 2050—with
over half of that growth in Africa. The UN Food and Agriculture Organization
estimates that as a result, food production will have to increase by 70 percent. But
according to the UN's latest climate change report, South Asia and sub-Saharan African—the two regions with the most food insecurity—are actually expected to see an 8 percent
drop in crop yields by 2050, thanks to
rising temperatures and increasingly sporadic rainfall.
Data services could help Tanzanian farmers adapt to unpredictable climate patterns. Karel Prinsloo/AP
Fraley estimates that in central Africa, a typical corn farm produces
less than a tenth of what the same-sized plot would yield in the US,
despite having roughly equivalent soil and weather quality. Poorly bred
seeds are part of the problem, he says, but biotechnology can't fill the
gap by itself. That's because African countries also suffer from a
chronic lack of basic weather data that US farmers take for granted.
Sub-Saharan Africa (excluding South Africa) has roughly one weather
station per 239,000 square miles; the US has one station per 14,000
square miles.
The need for better data is becoming even greater as climate change
alters rainfall patterns that have informed farming practices for
generations. A
2012 UN study found,
for example, that between the 1970s and the 2000s, precipitation during
Tanzania's main rainy season fell by nearly 30 percent. The rainy
season now starts a month earlier, as well.
Sub-Saharan Africa has roughly one weather station per 239,000 square miles. The US has one station per 14,000 square miles.
What farmers in the developing world
do have, though, is
cell phones. And that's where Monsanto sees an opportunity to distribute
its data. "There will be value in information on crops and geographies
where we currently don't do business, so it's a great opportunity to
expand our footprint and get us into new spaces," Fraley says. "Where
the information itself becomes the business, we see a lot of
opportunity."
Beyond weather predictions, mobile data services could help African
farmers identify pests and diseases—text in a photo of a mysterious bug,
and get advice on how to get rid of it. They could also help connect
farmers in remote villages to urban markets for their crops, says Andrew
Mattick, a former World Bank agricultural consultant who is currently
based in Mozambique. Rural farmers are often so isolated from markets,
he says, that it's impossible to know whether they are getting a fair
price for their produce.
"When you're not linked to markets," he says, "you have no idea how
much a cabbage costs in Maputo [the capital]." Mobile data could help
bridge that gap.
Monsanto is already active with data services in India, where
according to Friedberg some 3 million smallholder farmers have signed up
for text message updates that are essentially a simplified version of
the Climate Basic app. Field-by-field mapping efforts are underway in
South America, Fraley told me, and SMS services are being developed for
Africa, where the company's reach is currently relatively small.
Without access to market data, "you have no idea how much a cabbage costs in Maputo," says Mattick.
Data tools "are going to transform agriculture for smallholders
across Africa," Fraley says. "Even farmers who cannot read or write can
intuitively digest the information on a cell phone."
Of course, data products are only part of the business model. There
are strict limitations on GMO crops in India and in all but four African
nations, as Michael Specter recently reported in
The New Yorker. But
Fraley envisions that as African farmers begin to share data about
their farms, the company "will initially use that approach to support
our seed business. There's a huge opportunity to provide African growers
with better seeds."
In other words, free data services provided in Africa could
eventually feed back information about growing practices, disease and
insect problems, and climate conditions. That information could be
converted into seeds developed specifically for African fields and sold
to African farmers. The ball is already rolling: In May the company
announced the results
of the first harvest in Kenya of a new variety of drought-resistant
corn that had been in the works since 2008. Yields from the new seeds
were twice the national average, the company reported.
The real test for Monsanto's data push will be closer to home. For
farmers who have already signed on to the company's services, the next
few years' worth of harvests will show whether a smartphone app really
can pull extra corn from their ground—and put extra cash in their
pockets. Friedberg is confident that it will.
"As you start to realize that the upside is there," he says, "everything starts to change."
No comments:
Post a Comment