|
The
Business Side of Agricultural Imagery:
A Conversation with Gerry Kinn
Adena Schutzberg
Gerry Kinn is the Manager, Operations and Business
Development, Integrated Systems Group at Applanix, a
Trimble Company. An engineer who grew up in rural America,
he’s spent the last 27 years developing remote sensing
and GIS applications, many of which address agriculture
and the environment. He’s also a member of the
magazine’s Editorial Advisory Board.
There’s one thing Gerry Kinn wants to make clear
from the outset. Despite the lack of significant
commercial success for businesses trying to sell remote
sensing data and services to agriculture, the technology
does work. And, there’s no doubt of its benefits to the
industry—saving money, time, increasing yield, enhancing
best practices, protecting the environment, and more.
Despite that, he admits, “we’ve just not made a
business of it.” Exactly why that was the case quickly
became the theme of our conversation.
History
Kinn notes that selling remote sensing to
agriculture is not a new idea. He pulled out reports from
1996 that reference efforts in 1992 outlining business
ideas. In point of fact, government efforts date back to
the late 1970s with federal efforts such as AgRISTARS
which aimed to keep an eye on international agriculture.
That program used the best data of the time, Landsat and
some AVHRR data (Figure 1). The latter provided access to
daily vegetation indices. On the commercial side Kinn
points to two “early” examples of commercial success:
Cropix and Kass Green’s Pacific Meridian (acquired by
Space Imaging a few years ago and recently sold to
Geo360). Cropix, based in Hermiston, Oregon, drew on a
1988 NASA grant to explore the use of Landsat imagery to
predict crop yields. The company was active in the mid
1990s, using Landsat and SPOT imagery to monitor potato
fields in Oregon and Washington. At the time, the
company’s 24-hour data turn around was particularly
impressive. It was a small business, and though the fellow
behind it made a living, once he retired, the business
shut down.
Pacific Meridian, founded in 1988, on the other
hand, used imagery to support land use/land cover mapping.
Always poised as a services firm, the goal was to bring
remote sensing to business. A focus on the water rights
side of agriculture
in the dry west kept the company
going, as did policy issues related to water. Pacific
Meridian is perhaps best known as the data keeper that
allowed the two sides (loggers and conservationists) to
come together over the spotted owl habitat in Oregon.
Markets
The market for remote sensing products and services
related to agriculture, Kinn explains, runs from the
consumer and the commercial side on one end to state and
federal departments and agencies on the other. The
consumer, for now, is not really in the picture. The
commercial side can be broken down into the growers, who
produce the food (potatoes, grapes, corn, etc.), coops
that may include growers but also produce food products
(here in Massachusetts we always point to Ocean Spray
which is known for cranberries, but also juice, raisins,
sauce, and more) and food providers (those who produce
food products, such as Archer Daniels Midland and ConAgra,
companies behind raw materials, such as lecithin, and
final products, such as Healthy Choice frozen meals,
respectively) (Figure 2).
While some states have huge agricultural bases, few
have money to invest in remote sensing, which leaves quite
a lot of burden on the federal government. Its
agricultural interests can be organized around two key
agencies: the US Department of Agriculture (USDA) and the
Environmental Protection Agency (EPA). Other corners of
the federal government do keep an eye on agriculture, but
haven’t really been targeted by industry.
Over the years, Kinn continues, different players
have tackled different customers. Cropix targeted the
grower directly, as did some of the larger agricultural
players best known for machinery, but who also play in
seeds, fertilizer and other products. Chemical companies,
food producers and large conglomerates involved in
commodities trading, also tried their hands in remote
sensing services. Some offered commercial services while
others used the data internally. Services companies, like
Pacific Meridian, sat in between the commercial and
government customers, serving both. One long term player
Kinn identifies in this area, is Earth Satellite
Corporation, EarthSat, founded in 1969. Today the company
focuses a bit more on environmental issues than
agriculture, and does “big picture” and policy work
for federal agencies. (To be complete, EarthSat also sells
data products, such as NaturalVue to both the government
and the commercial sector.)
The Emerge Experience
Kinn speaks with great pride about Emerge, a
company he and others launched as a division of then
Litton-owned TASC, based outside Boston in 1996. (TASC is
now part of Northrop Grumman and Emerge, most recently
owned by ConAgra has since been split up and the pieces
acquired by Applanix, a Trimble Company and LJT, Inc.)
“We planned to serve the commercial market by bringing
the benefits of remote sensing, a database, and timely
weather information to the farmer. Combined with a crop
model (a predictive model of yield based on input [water,
fertilizer, etc.]), it could really change a farmer’s
prospects.” The vision was simple: the farmer would
build his own database documenting fertilizer, watering
and planting details in the software (something typically
done in a notebook by most farmers, even today). Then,
he’d use a crop model to predict the expected yield
(Figure 3). The check was to compare that expected outcome
to the reality of the remotely sensing images. “For that
level of detail, you needed high resolution imagery, one
foot, and in timely fashion,” Kinn notes. “That means
planes and digital sensors.” The target farmer had more
than 1,000 acres, meaning he had money to invest in such a
service, and too much acreage to manage with paper and
pencil.
Emerge and other players in the late 1990s offered
such services with limited success. Why? “Some of it was
cultural. Computer technology had not yet reached the
farm,” Kinn recalls, “but equally important was a kind
of confusion caused by state registration procedures. In
many states farmers must pay a fee to the state to
register the amount and type of fertilizers applied to
fields. While the fees are not large, the idea that a
payment is already due to the state, and this service
would cost more, led to understandable confusion in the
market.”
What’s Remote Sensing Good
for?
To follow up on his “the technology works”
statement, Kinn outlined what remote sensing can bring to
agriculture. First off, it can be used to monitor yields,
early in the season, and then later. In fact, there are
three key times to capture data: emergence
(“sprouting,” the source of Emerge’s name),
mid-season, and senescence (plant degeneration that
generally occurs at the end of the growing season). Each
period lasts about two weeks. One of the appeals of this
market, Kinn adds, is the idea that each field will need
to be flown three times per season, for many seasons. That
translates into more dollars for the service provider. If
accepted by the market, this can be the basis of a healthy
remote sensing business.
Second, remote sensing can delineate soil
zones. This is not the same type of the soil type maps
provided by the U.S. Soil Conservation Service (now
Natural Resources Conservation Service [NRCS]) with named
soil types, but rather a large scale mapping of
differences in composition based on percentage of sand,
loam, etc. The interesting thing here is the resulting
data doesn’t describe those percentages, but in fact,
simply creates lines on the map to distinguish between
different areas. The farmer then need only sample once or
twice in each polygon to determine the detailed makeup.
Consider that each soil sample costs about $6 to be
processed today, knowing where to sample can save money
over even a relatively small field.
A third area of benefit is in field delineation,
that is finding the boundaries of crops, in support of
regulatory procedures. “If you tell the USDA you are
farming 100 acres of soybeans, remote sensing is an
effective way to be sure that’s in fact what’s going
on,” says Kinn. In fact, he points out, aerial imagery
has been used for regulatory purposes since the 1940s.
The final area is what’s called precision
agriculture (Figure 4). That term has gotten a bit muddled
in recent years, so I asked Kinn for an up-to-date
definition. “It’s about best practices, giving a crop
just the water, fertilizer, herbicide, etc. it needs to
grow to its potential, but no more.” He feels strongly
that precision agriculture could have a significant impact
on the environment. “Consider that a farmer could
uniformly apply 130 lbs/acre of nitrogen fertilizer onto a
corn field. The plants would soak up what they need, and
the rest would run off into the water table. Whatever
residual there was from past practice would be excess. The
extra nitrogen does not benefit the plants, and, even
worse, contaminates the environment if it is not bound to
the soil in some way. If we can provide ‘just enough,’
the impact on the environment is reduced and the plants
should still provide their maximum yield.” Applying this
“just enough” vision should save the farmer money on
herbicide, water, fertilizer and other such costs. Of
course this needs to be balanced with the cost of
measuring crop needs and the technology to apply
fertilizer and other products.
Limited Success
With all those clear benefits, why has agriculture
not jumped at remote sensing from commercial through
government customers? Again, Kinn returns to culture.
“The link has not yet been made between these best
practices and a cultural imperative. I firmly believe that
growers want to be good stewards of the land, if for no
other reason than they know they have to return to the
same land next year.” These days, culturally, it’s not
“cool” to smoke. Society has made that change of
attitude. It’s not yet there in agriculture.
From the business side, Kinn notes that farmers
typically demand a minimum 3:1 return on investment to
take up a new technology (or any improvement). The
businesses offering such services could show that return,
but with inexpensive commodity crops (like corn) the
payoff might not be attractive enough. In the early 1990s
corn prices approached $4/bu, by the late 1990s they were
below $2/bu. With yields varying between 125 and 160 bu,
the revenue per acre could vary between $250 and almost
$700 per acre. That means that the desirability of buying
information products is often driven by the price of corn
on the commodities exchange. Higher cost crops, such as
citrus, have had some measure of success with remote
sensing, as have small scale uses. Kinn suggests too that
many businesses who entered the market hoped to serve
virtually any crop. Instead, he suggests, it might have
been wiser to specialize, and really learn one, preferably
high value, crop.
Another challenge can be linked to liability.
“Many players simply ‘don’t want to know’ about
the details such imagery provides.” Growers might be
liable should they mistakenly mistreat crops and imagery
later reveals it. That, he offers, in today’s litigious
society, could result in lawsuits not so different from
those aimed at doctors who misread x-rays. He also
believes that while vendors worked hard to simplify and
streamline software for the non-technical users, “most
apps were still too complex and too time consuming. In the
end, we failed to capture the early adopters that carry
you till you can reach the mass market.”
The Future
So, what has to change for the wider uptake of
remote sensing in agriculture? Kinn recalls that you can
never predict how a market will unfold. The best you can
do is look for analogous situations. That, he says, may
simply mean those in the industry need to be patient. Some
markets simply take a long time to mature. Consider, he
said, that the patent for the fax machine was granted in
the 1850s. It was operational in the 1860s, but was not in
widespread use until after 1960. That’s 100 years of
patience! Of course, he admits, technology adoption has
sped up since then, so there’s hope for increased
uptake, in “our lifetime.”
He notes too that the ultimate funding for the
breakthrough will be from consumers. He alludes to digital
imaging. The money in that is not from remote sensing, but
from people taking “pictures of Aunt Martha.” Money in
computers comes from little girls and grandmas wanting
those items for communication, games, and personal
enjoyment. “High value consumer uses for remote sensing
are possible, we’ve just not found them,” Kinn states.
To prove his point, he quizzes me: “Would you pay $10
for a detailed image of your yard to use for planting and
landscaping?” “Sure,” I reply. “Would you pay $30?
Would your neighbors pay? Would those in the upscale
neighborhoods?”
Kinn also notes that remote sensing is concentrated
on “valuable” geographies, ones on which the most
dollars are spent per square mile. Where are those today?
“Cities,” I suggest. He provides more examples: “We
don’t remote sense the desert, but the highway or
pipeline that runs across it. Or we might capture data on
a specific stand of trees with high value wood.”
Agricultural crops, for the most part, are not perceived
to have that high value, at least today.
In closing Kinn reminds me that from a business
perspective remote sensing needs two things: large areas
of interest (to demand a plane or satellite for coverage)
and the need for regular revisits. “Agriculture has both
of those properties. That’s what drew so many to
agriculture. Will it ever become a market for remote
sensing? ABSOLUTELY! However, I have no idea when.”
Back
|
