[Senate Hearing 115-645]
[From the U.S. Government Publishing Office]
S. Hrg. 115-645
TECHNOLOGY IN AGRICULTURE:
DATA-DRIVEN FARMING
=======================================================================
HEARING
BEFORE THE
SUBCOMMITTEE ON CONSUMER PROTECTION,
PRODUCT SAFETY, INSURANCE,
AND DATA SECURITY
OF THE
COMMITTEE ON COMMERCE,
SCIENCE, AND TRANSPORTATION
UNITED STATES SENATE
ONE HUNDRED FIFTEENTH CONGRESS
FIRST SESSION
__________
NOVEMBER 14, 2017
__________
Printed for the use of the Committee on Commerce, Science, and
Transportation
[GRAPHIC NOT AVAILABLE IN TIFF FORMAT]
Available online: http://www.govinfo.gov
__________
U.S. GOVERNMENT PUBLISHING OFFICE
37-228 PDF WASHINGTON : 2019
--------------------------------------------------------------------------------------
For sale by the Superintendent of Documents, U.S. Government Publishing Office,
http://bookstore.gpo.gov. For more information, contact the GPO Customer Contact Center,
U.S. Government Publishing Office. Phone 202-512-1800, or 866-512-1800 (toll-free).
E-mail, [email protected].
SENATE COMMITTEE ON COMMERCE, SCIENCE, AND TRANSPORTATION
ONE HUNDRED FIFTEENTH CONGRESS
FIRST SESSION
JOHN THUNE, South Dakota, Chairman
ROGER F. WICKER, Mississippi BILL NELSON, Florida, Ranking
ROY BLUNT, Missouri MARIA CANTWELL, Washington
TED CRUZ, Texas AMY KLOBUCHAR, Minnesota
DEB FISCHER, Nebraska RICHARD BLUMENTHAL, Connecticut
JERRY MORAN, Kansas BRIAN SCHATZ, Hawaii
DAN SULLIVAN, Alaska EDWARD MARKEY, Massachusetts
DEAN HELLER, Nevada CORY BOOKER, New Jersey
JAMES INHOFE, Oklahoma TOM UDALL, New Mexico
MIKE LEE, Utah GARY PETERS, Michigan
RON JOHNSON, Wisconsin TAMMY BALDWIN, Wisconsin
SHELLEY MOORE CAPITO, West Virginia TAMMY DUCKWORTH, Illinois
CORY GARDNER, Colorado MAGGIE HASSAN, New Hampshire
TODD YOUNG, Indiana CATHERINE CORTEZ MASTO, Nevada
Nick Rossi, Staff Director
Adrian Arnakis, Deputy Staff Director
Jason Van Beek, General Counsel
Kim Lipsky, Democratic Staff Director
Chris Day, Democratic Deputy Staff Director
Renae Black, Senior Counsel
------
SUBCOMMITTEE ON CONSUMER PROTECTION, PRODUCT SAFETY, INSURANCE, AND
DATA SECURITY
JERRY MORAN, Kansas, Chairman RICHARD BLUMENTHAL, Connecticut,
ROY BLUNT, Missouri Ranking
TED CRUZ, Texas AMY KLOBUCHAR, Minnesota
DEB FISCHER, Nebraska EDWARD MARKEY, Massachusetts
DEAN HELLER, Nevada CORY BOOKER, New Jersey
JAMES INHOFE, Oklahoma TOM UDALL, New Mexico
MIKE LEE, Utah TAMMY DUCKWORTH, Illinois
SHELLEY MOORE CAPITO, West Virginia MAGGIE HASSAN, New Hampshire
TODD YOUNG, Indiana CATHERINE CORTEZ MASTO, Nevada
C O N T E N T S
----------
Page
Hearing held on November 14, 2017................................ 1
Statement of Senator Moran....................................... 1
Prepared statement........................................... 2
Statement of Senator Blumenthal.................................. 3
Statement of Senator Inhofe...................................... 4
Statement of Senator Young....................................... 4
Statement of Senator Fischer..................................... 51
Press Release dated November 8, 2017 entitled ``Lindsay Corp
President and CEO Speaks to U.S. Senate Committee''........ 51
Report entitled ``FieldNET AdvisorTM Irrigation
Management Decision Support Tool'' by Lindsay Corporation.. 52
Statement of Senator Klobuchar................................... 61
Statement of Senator Lee......................................... 63
Witnesses
Justin Knopf, Vice President, Kansas Association of Wheat Growers 5
Prepared statement........................................... 6
Jason G. Tatge, Co-Founder, President and CEO, Farmobile......... 9
Prepared statement........................................... 11
Shannon L. Ferrell, J.D., M.S., Associate Professor, Oklahoma
State University Department of Agricultural Economics.......... 15
Prepared statement........................................... 18
Todd J. Janzen, President, Janzen Agricultural Law LLC........... 34
Prepared statement........................................... 36
Dr. Dorota Haman, Ph.D., Professor and Chair, Agricultural and
Biological Engineering, Institute of Food and Agricultural
Sciences, University of Florida (UF/IFAS)...................... 45
Prepared statement........................................... 46
Appendix
Hon. Bill Nelson, U.S. Senator from Florida, prepared statement.. 77
Timothy Hassinger, President and CEO, Lindsay Corporation,
prepared statement............................................. 78
Deere & Company, prepared statement.............................. 78
Response to written questions submitted to Justin Knopf by:
Hon. Jerry Moran............................................. 81
Hon. Catherine Cortez Masto.................................. 83
Response to written questions submitted to Dr. Shannon Ferrell
by:
Hon. Jerry Moran............................................. 84
Hon. Catherine Cortez Masto.................................. 89
Response to written questions submitted to Todd J. Janzen by:
Hon. Jerry Moran............................................. 92
Hon. Catherine Cortez Masto.................................. 93
Response to written questions submitted to Dr. Dorota Haman by:
Hon. Jerry Moran............................................. 94
Hon. Bill Nelson............................................. 95
Hon. Catherine Cortez Masto.................................. 98
TECHNOLOGY IN AGRICULTURE:
DATA-DRIVEN FARMING
----------
TUESDAY, NOVEMBER 14, 2017
U.S. Senate,
Subcommittee on Consumer Protection, Product
Safety, Insurance, and Data Security,
Committee on Commerce, Science, and Transportation,
Washington, DC.
The Subcommittee met, pursuant to notice, at 2:32 p.m. in
room SR-253, Russell Senate Office Building, Hon. Jerry Moran,
Chairman of the Subcommittee, presiding.
Present: Senators Moran [presiding], Blumenthal, Blunt,
Fischer, Inhofe, Lee, Young, Klobuchar, Hassan, and Cortez
Masto.
OPENING STATEMENT OF HON. JERRY MORAN,
U.S. SENATOR FROM KANSAS
Senator Moran. Good afternoon, everyone. Thank you for
joining us this morning. Our Subcommittee's hearing is on
``Technology in Agriculture: Data-Driven Farming.'' That's this
Subcommittee, and that hearing will come to order.
The agricultural community's adoption of field sensors,
drones, satellite imagery, advanced machinery, and similar
technology is increasing at an incredible pace. Our Commerce
Committee and this Subcommittee have been paying a lot of
attention to those issues. And the result of that increasing
pace is greater crop yields and improved sustainable practices
in farming. The most profitable farms are often the most
sustainable ones. This rapidly evolving technology will have a
vital role in preserving farmers' most important assets--their
land--with the potential increase farmers' margins to
unprecedented levels.
The collection and analysis of data has enabled farmers to
reduce costs through more efficient applications of inputs like
fertilizers and pesticides; improve production decisions
through enhanced recordkeeping and more accurate yield
predictions; and enhance land stewardship and sustainable
practices by removing inefficiencies in planting, harvesting,
water use, and the allocation of other resources. With an
increasing volume of quality data, in tandem with improved data
analysis, data-collection technology has the potential to
dramatically increase farm productivity and profitability.
The collection and use of such data raises issues regarding
control of the data, transparency of agreements between farmers
and data firms, and barriers to expanding internet access to
rural America.
Additionally, as data collection and sharing practices
become more popular across the ag economy, farmers are well-
positioned to benefit from their ``commoditization'' of data
collected from their land, especially as equipment
manufacturers, service providers, cooperatives and other
businesses seek access to that data.
The goal for this hearing is to educate and empower our
nation's farmers to understand the value of the information
they are creating, and certainly to allow Members of Congress
to have a better understanding of the current lay of the land
and what the future holds.
It's my pleasure to introduce the panel today, and I thank
you all for being here.
Justin Knopf is a farmer from Gypsum, Kansas, right in the
middle of our state. He grows wheat, alfalfa, soybeans, grain
sorghum, corn, and multi-species of cover crops. As part of his
sustainability-focused farming operations, he practices what is
referred to commonly as ``no-till'' farming and utilizes a
variety of technologies that assist his monitoring efforts to
be a good steward of the land while improving his yield.
Jason Tatge is the Co-Founder and CEO of Farmobile, a
technology firm based in Overland Park, Kansas; that's a suburb
of Kansas City. His company's services provide farmers with
real-time access to ownership of current and historical data
pertaining to their land. By providing a user-friendly,
simplified, and comprehensive overview of relevant data,
Farmobile's customers are able to make educated decisions in a
much more timely fashion.
Dr. Shannon Ferrell is an Associate Professor at Oklahoma
State in the Department of Agricultural Economics. He also
serves as an agricultural industry representative to the
Oklahoma Environmental Quality Board, which oversees operations
of the Oklahoma Department of Environmental Quality. And the
Senator from Oklahoma will have an opportunity to introduce Dr.
Ferrell shortly.
Mr. Todd Janzen is President of Janzen Agricultural Law,
LLC, and the Administrator of the Ag Data Transparency project.
This project makes available the Ag Data Transparency
Evaluator, which aims to provide clarity to consumers as to
what businesses do with the data that is shared with them all.
And, finally, Dr. Dorota Haman is Professor and Chair of
the Department of Agriculture and Biological Engineering at the
University of Florida. She specializes in irrigation water
management and efficiencies, and has been an active leader in
providing irrigation technologies in developing countries, in
the Americas, and in Africa.
[The prepared statement of Senator Moran follows:]
Prepared Statement of Hon. Jerry Moran, U.S. Senator from Kansas
Good afternoon. Welcome to the Subcommittee's hearing on
``Technology in Agriculture: Data-Driven Farming.'' The Subcommittee
will come to order.
Thank you for being here today to discuss the advancements and
benefits of agricultural technology and the potential of ``Big Data''
in farming.
The agricultural community's adoption of field sensors, drones,
satellite imagery, advanced machinery and similar technology is
increasing at an incredible pace to increase crop yields and improve
sustainable practices. The most profitable farms are often the most
sustainable ones. This rapidly evolving technology will have a vital
role in preserving farmers' most important asset, their land, with the
potential to increase farmers' margins to unprecedented levels.
The collection and analysis of data has enabled farmers to reduce
costs through more efficient applications of inputs like fertilizers
and pesticides; improve production decisions through enhanced
recordkeeping and more accurate yield predictions; and enhance land
stewardship and sustainable practices by removing inefficiencies in
planting, harvesting, water use, and the allocation of other resources.
With an increasing volume of quality data, in tandem with improved data
analysis, data-collecting technology has the potential to drastically
increase farm productivity and profitability.
The collection and use of such data raises issues regarding control
of the data, the transparency of agreements between farmers and data
firms and barriers to expanding Internet access in rural areas.
Additionally, as data collection and sharing practices become more
popular across the agriculture economy, farmers are well-positioned to
benefit from the ``commoditization'' of data collected from their land,
especially as equipment manufacturers, service providers, cooperatives,
and other businesses seek to access and utilize this data.
My goal for this hearing is to educate and empower our Nation's
farmers to understand the value of the information they are creating.
It is my pleasure to introduce our panel today. Thank you all for
being here.
Mr. Justin Knopf is a farmer from Gypsum, Kansas, and he grows
wheat, alfalfa, soybeans, grain sorghum, corn, and multi-species cover
crops. As a part of his sustainability-focused farming operations, he
practices what is commonly referred to as ``no-till'' farming and
utilizes a variety of technologies that assist his monitoring efforts
to be a good steward of the land while improving his yield.
Mr. Jason Tatge is the Co-Founder and CEO of Farmobile, a
technology firm based in Overland Park, Kansas. His company's services
provide farmers with real-time access to and ownership of current and
historical data pertaining to their land. By providing a user-friendly,
simplified yet comprehensive overview of relevant data, Farmobile's
customers are able to make educated decisions in a timely fashion.
Dr. Shannon Ferrell is an Associate Professor at Oklahoma State
University Department of Agricultural Economics. He also serves as the
agricultural industry representative to the Oklahoma Environmental
Quality Board, which oversees the operation of the Oklahoma Department
of Environmental Quality.
Mr. Todd Janzen is President of Janzen Agricultural Law, LLC and
the Administrator of the Ag Data Transparency project. This project
makes available the Ag Data Transparency Evaluator, which aims to
provide clarity to consumers as to what businesses do with the data
that is shared with them.
Dr. Dorota Haman is a Professor and Chair of the Department of
Agricultural and Biological Engineering at the University of Florida.
She specializes in irrigation water management and efficiencies, and
has been an active leader in providing irrigation technology to
developing countries in the Americas and Africa.
I look forward to hearing the testimonies of this expert witness
panel. I now turn to my colleague Ranking Member Blumenthal for his
opening remarks.
Senator Moran. I look forward to hearing the testimony of
these expert witnesses. And before we do that, let me turn to
the Ranking Member, the Senator from Connecticut, Mr.
Blumenthal.
STATEMENT OF HON. RICHARD BLUMENTHAL,
U.S. SENATOR FROM CONNECTICUT
Senator Blumenthal. Thank you, Mr. Chairman.
And thank you to the witnesses for being here today. And
thank you to the Chairman for having this hearing.
I represent the State of Connecticut in the United States
Senate. And I want to welcome witnesses from the states that do
a different type of farming. We do have farming in the State of
Connecticut. And I have an additional connection to what you
folks do for a living, which is that my grandfather had a farm,
and my first job literally was shoveling manure on his farm at
the age of probably about 7 or 8 years old.
I would bet the most complicated piece of machinery on his
farm in the 1950s and 1960s was the radio in his house, and the
idea of data and farming being in the same sentence would have
totally perplexed him. But, in fact, data has enabled us to
increase yield and productivity in ways that would have been
unimaginable to him and many farmers of his generation and
maybe the generation afterward. And the benefits have been
widely shared by America and the world because America's
farmers have led the world in using technology to raise
productivity and yield.
At the same time, the advances in data have raised
questions about who owns it, who controls it, how do we protect
privacy, and how do we prevent others from, in effect,
profiteering at the expense of our farmers, who really should
be the ones who own that data and control it?
So these kinds of questions bring us here today. And I
thank you for shedding some light on an enormously important
and complex topic that occupies this Committee in a number of
different realms, and this one is certainly one of the
preeminently important ones.
Thank you very much.
Senator Moran. Mr. Blumenthal, thank you very much. Thank
you for explaining how you got your start in politics.
[Laughter.]
Senator Moran. I would recognize the Senator from Oklahoma
for purposes of an introduction.
STATEMENT OF HON. JIM INHOFE,
U.S. SENATOR FROM OKLAHOMA
Senator Inhofe. Well, since I don't know anything about Dr.
Ferrell, I won't introduce him except to say that he is here on
behalf of Oklahoma State University, and he's got to be a good
guy if he's with the Oklahoma State University. Thank you.
Senator Moran. He was smart enough, Senator Inhofe, to wear
the KU tie, however.
[Laughter.]
Senator Moran. Perhaps that's just patriotic.
Senator Young has the floor for purposes of introduction.
STATEMENT OF HON. TODD YOUNG,
U.S. SENATOR FROM INDIANA
Senator Young. Thank you, Mr. Chairman. I appreciate you
holding this very timely hearing on the agricultural technology
and digitalization, if I can say that word, of the farm.
I am happy to introduce Todd Janzen, a fellow Hoosier and
expert in the field. He is the President of Janzen Agricultural
Law in Indiana, and his experience in the industry began at an
early age. Todd grew up on a grain and livestock farm in
Kansas, where he learned the ins and outs of the industry.
After graduating from Bethel College in Kansas, Todd made his
way to the great Hoosier state and attended my alma mater,
Indiana University's McKinney School of Law, where he began his
law career in Indianapolis.
In addition to his work at Janzen Ag Law, Todd serves as
General Counsel for the Indiana Dairy Producers. He's a member
of the Indiana Farm Bureau Property Rights Policy Committee,
and previously he sat on the Board of the Council for
Agriculture, Science, and Technology.
Todd maintains a blog on law and technology issues facing
the agricultural industry, and his writing has been republished
by numerous journals and news sources. Todd is the
Administrator of the Ag Data Transparency project, which is
making ag data publicly available to farmers all across the
world. These insights will be especially relevant today I know
as the Committee discusses ag technology and examines the many
potential benefits and considerations that lie ahead.
I want to thank Mr. Janzen for taking the time to testify.
And I look forward to hearing the entire panel's discussion
this afternoon.
Senator Moran. Thank you both for those introductions, and
let us now hear from our witnesses. We'll start with Mr. Knopf
and work our way to his left.
STATEMENT OF JUSTIN KNOPF, VICE PRESIDENT,
KANSAS ASSOCIATION OF WHEAT GROWERS
Mr. Knopf. Mr. Chairman and Ranking Member Blumenthal,
members of the Subcommittee, thank you for the opportunity to
address you today. My name is Justin Knopf. I am a fifth
generation farmer from Gypsum, Kansas, as Chairman Moran
mentioned.
I also currently serve as Vice President for Kansas
Association of Wheat Growers. Working alongside my dad and
brother, we grow wheat, alfalfa, soybeans, grain sorghum, corn,
and multi-species cover crops across our 4,500-acre farm.
Like most U.S. farms, we are a family farm. As my father
has always said, the most important thing we raise on our farm
is children. My wife Lindsey and I have two daughters and a
son, and my brother and his wife have two sons. We utilize a
holistic approach in our farm management, rooted in values of
faith and family with a multi-generational view. This
decisionmaking process examines not only our economic returns,
but also the returns to natural and human resources.
For the past 15-plus years, our dry land operation has
utilized a cropping system focused on continuous no-till
practices and cropping rotations. This system protects the
soil, allows biology in the soil to thrive, is more resilient
to extreme weather, and increases carbon content in soils by
sequestering carbon dioxide from the atmosphere.
Never before has our society had the access to data and
information that we have today. The same is true in
agriculture. Data collection, processing, and the utilization
of data for improved decisionmaking has become a core
competency for many, if not the majority of farmers. The amount
of knowledge per acre and amount of knowledge about each acre
are significant drivers in the amount of profit per acre.
There are three main areas on our farm where data shapes
our decisionmaking and has impact: economic sustainability,
environmental stewardship, and transparency with consumers. And
I'd like to share a brief example of each with you.
As you know, the current low commodity prices equate to a
difficult economic reality on the farm. Managing costs is
critical right now. We are utilizing data to divide fields into
specific management zones; for instance, high-, average-, and
low-producing areas of the field. When growing corn, seed is
one of more expensive inputs. We are prescribing different
planting populations for each yield zone, planting more seed in
areas of the field that consistently produce more, maximizing
our return on that investment, and planting less seeds in areas
of the field that are consistently lower producing, lowering
our costs on those acres.
Farmers understand the importance of environmental
stewardship and protecting our natural resources for future
generations. That being said, we've made our fair share of
mistakes and always have room for improvement. On our farm, we
have a significant focus on protecting the soil and improving
its health and resiliency. Cover crops are an important tool in
this endeavor. Utilizing data collection equipment on our
machinery to carry out on-farm research trials has enabled us
to better quantify the impact of cover crops to subsequent crop
yields and other agronomic factors.
Consumers today, and, therefore, the supply chain, are
increasingly interested in how their food is produced and that
it's done in a way that corresponds with their values. This
past year, we enrolled our wheat acres into a sustainability
program with ADM. Basically, we entered field information,
things such as the amount of fertilizer rates used in each
field and yield data into a web-based software. The software
utilizes sustainability metrics designed by the collaborative
group field to market, and then gives us quantifiable
environmental impact metrics on our farm, benchmarks on how our
farm compared to other farms, and then the supply chain
receives the aggregated data.
The agricultural economy is at a crossroads right now,
depressed prices, increased costs, and rising debt levels are
creating economic angst. The average age of farmers continues
to increase while the number of us continues to decrease. There
is this great challenge of intensifying our farming system and
doing so in a way that is sustainable, if not restorative, to
our natural resources. Consumers are increasingly removed from
the farm and wary of technological innovations in farming.
These are all significant challenges. However, the minds
and spirits engaged in what will be the next generation of
agriculture are as bright as ever. There will be a record
percentage of farms transitioning to this next generation in
the coming decade. This transition represents a great
opportunity for change and innovation not only in improved
productivity, but also in environmental stewardship. It is
critical that we collaborate, learn, and adapt in order that we
may have continual improvements.
I appreciate genuinely the opportunity to share the value
of data with you today. And I appreciate that Congress is
willing to listen to the people who may be impacted by future
legislation.
[The prepared statement of Mr. Knopf follows:]
Prepared Statement of Justin Knopf, Vice President,
Kansas Association of Wheat Growers
Mr. Chairman, Ranking Member Blumenthal, and Members of the
Subcommittee, thank you for the opportunity to address you today. My
name is Justin Knopf and I am a fifth generation farmer from Gypsum,
Kansas. I also serve as the Vice President of the Kansas Association of
Wheat Growers. Working alongside my dad and my brother, we grow wheat,
alfalfa, soybeans, grain sorghum, corn, and multi-species cover crops
across our 4500 acre farm.
As my father has always said, we also grow people on our farm. For
my brother and I, farming is a lifelong learning process. The local
young people who find a summer job with us grow in responsibility, work
ethic, and perspective. My wife Lindsey and I have three young
children, two daughters and a son, and my brother also has two young
sons. Our farm is not unlike most farms in the United States. According
to the USDA, farming is still overwhelmingly comprised of family-owned
businesses. 99 percent of U.S. farms are family farms, and they account
for 89 percent of farm production.\1\ The USDA also estimates that
small farms make up 90 percent of the farm count and operate nearly
half of America's farmland.\2\
---------------------------------------------------------------------------
\1\ https://www.ers.usda.gov/webdocs/publications/eib164/eib-
164.pdf
\2\ https://www.ers.usda.gov/webdocs/publications/eib164/eib-
164.pdf
---------------------------------------------------------------------------
We utilize a holistic approach to our farm management, rooted in
values of faith and family, with a multi-generational view. Farmers
understand the need for good stewardship and conservation. This is what
we do every day. We depend on clean water and healthy soils to make a
living and feed the world. This decision making process examines not
only our economic returns, but also the returns to natural and human
resources. For the past fifteen plus years, our dry land operation has
utilized a cropping system focused on continuous no-till practices and
crop rotations. This system protects the soil, allows soil biology to
thrive, is more resilient to extreme weather, and increases Carbon
content in soils by sequestering CO2 from the atmosphere.
Although we have been blessed with some bountiful harvests in the
last few years, the current economic reality on the farm is difficult
and the coming years are shaping up to be some tough times. Farm income
levels are at their lowest point since 1985. Net farm income dropped 95
percent from 2014 to 2015, and net farm debt levels have increased 25
percent over the last 3 years.\3\ This downturn has largely been caused
by low commodity prices, which are due to record highs in both local
and worldwide production over the past two years.\4\ These production
levels have increased supply, while overall demand has waned, due to a
strong U.S. dollar and decreasing exports.\5\ Another major factor is
that while revenues have only gone down, the cost of production and
expenses have gone up. From 2009 to 2015 the cost of production has
increased almost 50 percent.\6\ This rise in costs has forced farmers
to look for ways to find efficiencies and minimize costs. Our ability
to adapt to changes is what will keep us going when times get tough.
---------------------------------------------------------------------------
\3\ http://www.agmanager.info/kfma/state-summaries
\4\ https://www.wsj.com/articles/whats-behind-the-glut-in-
agricultural-commodities-1476670
020
\5\ https://www.wsj.com/articles/the-next-american-farm-bust-is-
upon-us-1486572488
\6\ http://www.agmanager.info/kfma/state-summaries/2015-state-
summary-detailed-cost-summary
---------------------------------------------------------------------------
According to the United Nations there will be 9.1 billion people on
the planet in the year 2050.\7\ One of the more significant long term
challenges facing our world is how we feed a growing global population.
Food security isn't just an agricultural issue; it is a national
security issue. As farmers, we must find a way to produce more food, on
less land, with less water, all while protecting our soils and natural
resources. As stewards of the land it is our job to find ways to do
more with less. It will take all available tools to meet these
challenges. Agricultural innovations, like technological improvements,
seed technology, and on farm efficiencies, are all important. Research
within private entities and public institutions is critical. Perhaps
most fundamental is collaboration with others, an eagerness to learn,
and a willingness to adapt.
---------------------------------------------------------------------------
\7\ http://www.fao.org/fileadmin/templates/wsfs/docs/expert_paper/
How_to_Feed_the_World
_in_2050.pdf
---------------------------------------------------------------------------
Never before has our society had the access to data and information
that we have today. Data is all around us, and there is value in it
all. While a record of Google searches and websites visited may be
useless history to me, analysts and marketers see valuable information
that allows them to adjust the content they create. The same, of
course, is true in agriculture. While some may see a jargon-filled
spreadsheet or just a bunch of various colors on a field map, I see
ways to maximize efficiency in my operation, both for my pocketbook, as
well as for the land that provides the livelihood of my family. Data
collection, data processing, and the utilization of data for improved
decision making has become a core competency for many, if not a
majority of, farmers.
The obvious benefit of data is the ability to make improved
management decisions. Data has become an important layer in our
decision making process and a driver in our economical sustainability
and environmental stewardship. The amount of knowledge per acre, and
amount of knowledge about each acre, are significant drivers in the
amount of profit per acre.
There are three main types of data we utilize on our farm.
Microdata is data we collect and produce that is specific to our farm.
Service provider data is data that is provided to us by service
partners that is specific to our farm. And Macrodata, or big data, is
data we provide to others and they, in return, give us an idea of what
is happening in the industry on a larger scale.
Specifically on our farm, we collect and utilize this data in a
number of different ways. As on many farms, our seeding, spraying, and
harvesting equipment all has hardware and software that measures and
records spatially what is being done or happening in the field.
Performing on-farm research with sound scientific and statistical
principles is one way we use this technology. For example, we were able
to quantify the impacts of cover crops on subsequent crop yields, which
has led to a broader adoption of cover crop practices. We utilize
satellite imagery as a way to help us identify management zones within
a field and predict yield variability. These management zones allow us
to modify our seeding rates based on the productivity of the land,
which lowers our seed cost on acres that are less productive. We use
zone soil sampling to quantify soil fertility levels in differing areas
of the field, which allows us to fertilize based off specific soil
conditions and fertility levels. It allows us to focus inputs on the
areas that need them and to avoid applications on areas that don't. We
enter data into ADM's sustainable wheat program, and in return we
receive sustainability metrics based off the field to market
calculator. Through this program we are able to use key sustainability
outcomes and metrics and benchmark our farm's performance to others in
the program.
I also share economic, cost, and revenue data with the Kansas State
University (K-State) Farm Management program and receive informational
data back on my farm's profitability in relation to other like-sized
no-till farms. This program allows me the ability to know how my
business is doing in relation to others in the industry, what business
strategies I should implement to become more profitable, or what
investments I should make in my business.
The quality and the quantity of data in agriculture, and its
importance, is driving the improvement of farming practices and its
value will only continue to grow. It is vital that our stakeholders and
collaborators work alongside our public research institutions, such as
K-State, to continue to develop the tools farmers need to be
successful. Private industry is rapidly expanding in this space and the
technology is changing by the day. Competition for the ``digital acre''
is increasing and it is rapidly driving innovation.
For example, we can now use crop sensors mounted to sprayers that
utilize algorithms, developed by K-State and other land-grant
institutions, that tell us in real time how much nitrogen each plant
needs, while giving credit to biological nitrogen that already exists
in the plant. As the sprayer travels through the field the sensors will
tell us in real time how much nitrogen the plants in that spot need.
This technology allows us to put the right amount of nitrogen in the
right place which saves money and increases environmental stewardship.
There are also proprietary tools from companies such as Pioneer and
Monsanto that utilize soil and weather data to predict a crop's
nitrogen needs and the amount of available nitrogen in the soil. This
data helps farmers tune the timing and quantity of fertilizer
applications to increase efficiencies.
A researcher from Kansas State has been utilizing land on our farm
and others to test and develop a sensor that can quickly and
efficiently quantify soil water-holding capacity differences across a
field. As water is typically one of our most limiting factors for crop
production on many farms in the Great Plains, efficient access to this
information would be very valuable in developing management zones and
insight in how to best manage each area within the field.
Data is also important to those off the farm as well. Consumers
have an ever growing interest in their food. They want to know more
about how their food is produced, how it is processed, and if it is
being grown in a way that aligns with their values. Our use of data
allows us to tell our story to the consumer and enables us to do so
with transparency like never before.
However, as we begin to find new ways to collect and utilize this
valuable data we need to make sure we protect the ownership interests
and rights of farmers. We need to make sure that government and
regulatory agencies do not try and access proprietary data that is
critical to a farmer's business. We need to make sure third party
dealers and vendors do not try and take ownership of data that was
generated and collected by the farmer. Finally, we need to ensure that
the privacy rights and ownership interests of the farmer are respected
by all those who may want to access this data. The last thing we need
is for those who are not aligned with our farm interests to twist and
misconstrue what we do on the farm.
The agricultural economy is at a crossroads right now. Depressed
prices, increased costs, and rising debt levels are creating economic
angst. The average age of farmers continues to increase while the
number of us continues to decrease. There is this great challenge
intensifying our farming system, but doing so in a way that is
sustainable if not restorative to our natural resources. Consumers are
increasingly removed from the farm and wary of technological
innovations in farming. However, the minds and spirits engaged in
agriculture and farming are as bright as ever. There will a record
percentage of farms transitioning to the next generation in the coming
decade. This transition represents a great opportunity for change and
innovation, not only in improved productivity, but also in
environmental stewardship. It is critical that we collaborate, learn,
and adapt in order that we may have continual improvement. I appreciate
the opportunity to share the value of data with you today, and I
appreciate that congress is listening to the people who may be impacted
by future legislation. There is an immense amount of technology, both
here and on the horizon, that will allow American farmers to continue
to meet these challenges. This drive for continual improvement and
understanding of the complex biological ecosystem we farm with is what
will allow my children, and their children after them, to continue
feeding the world and protecting the natural resources long after I am
gone. I urge you to continue to listen as your shape future
legislation.
Senator Moran. Thank you, Justin.
And now Mr. Tatge.
STATEMENT OF JASON G. TATGE, CO-FOUNDER, PRESIDENT AND CEO,
FARMOBILE
Mr. Tatge. Chairman and members of the Subcommittee, thank
you for the invitation today to share my ideas about how to
improve the collection, standardization, and interoperability
of agricultural data for the benefit of every American.
My name is Jason Tatge, Co-Founder, President, and CEO of
Farmobile. We help farmers collect and organize data so they
can use it to better manage their own operations, share the
data with their trusted partners, and even license their data
to vetted third parties to create a brand new revenue stream
for their business.
I've been on over 200 farms in the Midwest in the last few
years, and I've had hundreds of conversations with farmers
defining the future of global agriculture. I'd like to share
what I've learned from listening to the men and women on the
ground.
One of the most important topics we discuss on the farm is
data. Who has it? Who does it belong to? And what's it being
used for?
Through my conversations, I've come to believe that for the
sake of a safer, more productive food future, farmers need to
be able to own their data outright. Farm data needs to be
accessible in real time and compatible with multiple systems.
In short, we need a standard for the agricultural industry.
I will start with the question of ownership. Big
agricultural companies know it will benefit them to own the
digital content coming out of the farmers' fields. When a U.S.
farmer spends hundreds of thousands of dollars on a new piece
of equipment, the largest manufacturers profit from the initial
sale of the equipment, then they profit again from the data
generated from the farmer using that equipment. This is wrong,
because the data being collected by many big ag companies is
the farmers' proprietary intellectual property. It is a unique
formula or secret recipe for operating their successful
businesses.
Attempts to get big players in the ag industry to
voluntarily enact transparent data policies have been slow. In
fact, organizations who played a big role in drafting the best
practices for data transparency have failed to sign on or adopt
them. Asking a farmer for their secret recipe would be bad
enough, but tricking them into signing it away, the unique
formula, with complicated legal agreements, is appalling to
most, and the main reason I sit here.
The potential value in this data to farmers and
agricultural communities writ large cannot be overstated. If
farmers own their data and can license it multiple times, we're
talking about an opportunity to create an estimated billion
dollar annual new revenue stream returning that money to rural
agricultural communities of America.
To further illustrate the importance of data ownership, I
want to talk about the value data represents for a farmer.
Here's an example. We know the genetic yield potential for corn
in the U.S. is over 500 bushels per acre, yet the national
average for corn is only about 170 bushels. Why is this? The
answer lies in the data.
Farmer-controlled digital records to document the farming
practices will help U.S. agriculture and individual farmers
determine best practices for maximizing their yields. Those
records enable farmers to make better decisions, identify
efficiencies, boost productivity, and mitigate risks as well as
aid the industry in streamlining the manual process required to
participate in Federal programs and crop insurance.
These kinds of ag data benefits, however, require
agriculture to get past the roadblocks of data
interoperability. Many farmers operate mixed fleets today, and
that means they have separate data systems for each equipment
brand. Ag companies make it incredibly labor-intensive to move
data out of one system and into another. This stifles
competition and customer choice in an already dramatically
shrinking landscape of agricultural giants.
I believe we need to properly align incentive structures
and drive standardization across the industry. Every farmer has
a right to access and use their data regardless of where it
came from or what system contains it. We should encourage the
flow of information that could help farmers and their trusted
advisers to make the best decisions for their farms and the
food industry at large.
The power of data can go way beyond the field. When farm
data is available in real time in a standardized portable
format, like Farmobile's electronic field records, there's a
huge potential to reduce the volatility in commodity markets.
Agricultural markets are volatile for good reason: there's
a massive lag time in getting the information. The USDA is the
gold standard, but even the USDA sometimes takes 3 weeks to get
out information in the right form for release, but most row
crops are only alive for about 90 days. Faster information will
dramatically reduce the volatility in the markets and enable
traditional risk management strategies, like hedging, to work
again for farmers and the agricultural businesses that rely on
these commodities to produce their products.
In 1960, John F. Kennedy said, ``The farmer is the only man
in our economy who has to buy everything at retail, sell
everything at wholesale, and pay the freight both ways.'' It's
unfortunate that this statement is still very true today.
But disruption is coming in the form of technology. We need
to make sure that our legal system keeps up with the technology
available and is informed on the formative debates that will
define the future in my industry. Most farmers I talk with
believe we have 2 to 3 years to figure this out, or they will
lose. Farmers need policies that safeguard data rights, are
interoperable, and improve data access to drive efficiencies in
innovation and food production. As you consider these issues
further, ask yourselves, ``Why is it ever okay for others to
own or control farmers' data? And how do we enact policies that
create true data interoperability?''
I firmly believe that done right, data is the answer to
advancing agriculture and the entire food industry while
protecting America's farmers.
I appreciate your openness to ideas and action from the
private sector as well as administrative and legislative
change. I look forward to working with the industry as well as
the members of the Committee to advance this vision. Thank you.
[The prepared statement of Mr. Tatge follows:]
Prepared Statement of Jason G. Tatge, Co-founder, President and CEO,
Farmobile
Chairman and Members of the Subcommittee:
Thank you for the invitation, today, to share my ideas about how to
improve the collection, standardization and interoperability of
agricultural data for the benefit of every American.
My name is Jason Tatge, co-founder, president and CEO of
Farmobile--a relatively small agtech startup company, from Kansas, with
a creative business model that turns our farmer customers data into a
monetizable commodity and shares the revenue with the farmers. We
recently celebrated our four-year anniversary and employ over 40 people
with plans to add at least 20 more over the next 12 months. Farmobile
offers a ``data as a service'' subscription that properly aligns our
company's future success with our farmers' success. Practically
speaking, we help farmers to collect and organize their data so they
can use it to better manage their own operations, share their collected
data with their trusted partners and/or sell their data to interested
third parties, the same way musicians can sell their music.
I've been on over 200 farms in the Midwest in the last few years
and am completely amazed by how ridiculously awesome these people are
at growing our food. These folks are the ``rock stars'' of global
agriculture.
One of the most common topics we discuss when on the farm is data
ownership. Many are confused over how we've gotten to this current
place or when data ownership even become a question They are confused
about who has access to their data and what they are doing with it. All
have an expressed interest in being able to establish a value for the
data they generate. As one farmer, David Seba from Cleveland, Missouri
told me, ``Big ag has been collecting our data for so long, that
there's this attitude that the way we farm carries no value. Well, it
does. For farmers, the field is our business and the way we manage it
is our formula for success. So, why is it okay for these companies to
claim the data as theirs and then sell it without our knowledge?''
At Farmobile we are proud to be working alongside some of the most
innovative farmers in the world, and we are passionate about providing
these farmers the opportunity to establish ownership and directly
profit from the data generated from their field activities, if they
choose to sell licenses to their data.
Introduction
Whether you represent the 2 percent of the U.S. population who farm
or the non-farm constituents who--like all of us--eat, farm data will
become a digital currency that impacts both farmers and food buyers.
Today, I'll share my thoughts about the state of the industry, and the
needs, risks and opportunities we have.
Data and analytics are disrupting and changing most industries.
From grocery shopping to political campaigns, the world is forever
changed by data. Farming is no different, although I'd suggest we are a
few years behind other industries when it comes to data collection.
That's changing fast, and we have a lot to learn from other industries
that have already made the move from analog to digital, like
healthcare.
A big part of farming today is being able to manage a large mixed
fleet of equipment. Real-time data connectivity empowers farmers to
remotely manage their logistics like never before by using any Internet
connected device. While adoption of precision agriculture technologies
has been on the rise for years, now that it's available in real-time,
adoption has accelerated because farmers quickly ``see'' the value of
data.
Real-time data is the ``game changer'' for the future of farming
because of the ability to gain insights and react ``right now'' during
the season. This is the foundational driver to improve yields, lower
input costs, strengthen stewardship and pave the way for cutting-edge
programs like yield guarantees that enable seed and chemical companies
to ``share the performance risks'' associated with their recommended
products.
Big ag companies certainly agree that the industry is going
digital. Look no further than public statements made by ag business
giants framing this opportunity for their shareholders. They absolutely
know it will benefit them to own the digital content coming out of the
farmers' fields.
But this comes with a cost to the farmers, not only do these big
companies expect to get the data for free, but they also create
``silos'' for the data and make it very difficult to get that data back
out from their systems. This stifles competition and customer choice in
an already dramatically shrinking landscape of agricultural giants,
whose recent mergers have reduced the big six to the big four.
Farmers are just beginning to understand that their data has value
outside the perimeter of their operations, and that data ownership and
a neutral digital strategy is necessary to be competitive today.
Because of this, some farmers are starting to ask tough questions about
data: Who owns it? How will it be used? How do I extract maximum value
from it? And, most importantly, how do I put a ``fence'' around my data
so that it's protected for future generations?\1\
This brings me to the first opportunity and risk for farmers--data
ownership.
Farmers and Data Ownership
As business owners, farmers face a very real risk from many ag
companies with whom they do business because: 1) companies gain access,
control and sometimes ownership over the farmer's private data; and 2)
these companies can lock farmers into their data policies.
At the center of this growing concern is the method in which ag
companies typically collect and use a farmer's data. To understand it,
let me provide a consumer-facing illustration that everyone
understands--Google.
When I choose to use Google to search the web (for free), I
understand that Google is collecting information about me through my
interaction with their technology. I know that Google turns this
collected data into information by combining it with other datasets.
Further, I realize Google makes money from selling this information to
marketers that want to learn more about me. In spite of this, I choose
to use Google search because it is of value to me, and it's free.\2\
On the other hand, when I purchase a license to use Microsoft
Office, I gain access to tools like Microsoft Word, Microsoft Excel and
Microsoft PowerPoint. These tools provide value to me. When I use these
tools, Microsoft does not get rights to the content I create. Could you
imagine the types of congressional hearings we'd be having on that
topic--if Microsoft treated its customers the way big ag treats their
customers?
When a U.S. farmer spends hundreds of thousands of dollars on a new
piece of equipment, the largest manufacturers profit from the initial
equipment sale PLUS they profit from the data generated from the farmer
using that equipment. The collection of this data often happens without
the farmer's knowledge due to complex and heavy-handed user agreements.
While the fact that Google is collecting search data doesn't bother
me as a consumer, the stakes are much higher and far different in the
farmer example. The data being collected by many big ag companies is
the farmer's Intellectual Property--the special and unique formula or
``secret recipe'' for operating their successful business.
Imagine if we, as a user of Google, asked for its search engine
algorithms. Or, as a customer of Microsoft, if we asked for its source
code to the Microsoft Office Suite? Asking a farmer for their ``secret
recipe'' would be bad enough, but tricking them into signing away that
unique formula with complicated legal agreements is appalling to most
and the main reason I am here today.
We believe farming practices represent Intellectual Property that
could be copyright protectable. Yet, today, it is difficult to
establish who owns this information because farmers are caught in the
habit of unknowingly giving this data for free when they sign
complicated legal agreements pertaining to an entirely different
subject. It is my personal motivation to help farmers by providing
alternatives with upside potential.
I've been working for the better part of three years with the
American Farm Bureau Federation (AFBF) to address these issues. The
AFBF has shown great leadership in trying to bring transparency to
these confusing legal contracts farmers are required to sign. Working
with commodity groups, farm organizations and agriculture technology
providers, the AFBF established the Privacy and Security Principles for
Farm Data in November 2014. Thirty-seven different organizations
participated in drafting the ``Core Principles'' document. Many of
these organizations were very opinionated around the wording of the Ag
Data Transparency Evaluator's ten questions, but only nine of these
companies have agreed to become Ag Data Transparent! The ones who
haven't signed are challenging the very need for ``ownership'' of farm
data to be defined in the ``Core Principals.''
Make no mistake about it, these companies are intentionally
delaying participation because they hope this issue will blow over and
farmers will continue to operate the way they have in the past--by
unknowingly checking a box in a legal contract in order to take
delivery of their product.
Missouri Farm Bureau President Blake Hurst of Tarkio, Missouri
describes the situation like this: ``So much of what we do is done by
habit. As soon as we get in the habit of giving that data away, no
company is going to remark on the fact that it is a heck of a good deal
for them. If we don't start out doing it the right way, it will be very
harmful to farmers in the future.''
In agriculture, we are at a point in time where there is a great
opportunity to ``do the right things for the right reasons'' on behalf
of the people who produce our food.
Data Interoperability
We know the genetic yield potential for corn in the U.S. is over
500 bushels per acre, yet the national average for yield is about 170
bushels per acre. Having farmer-controlled digital records (such as
Electronic Field Records) to document farming practices will help U.S.
agriculture better determine best practices farmers. Those records
enable farmers to make better decisions, identify efficiencies, boost
productivity and mitigate risks, as well as aid the industry in
streamlining the manual processes required to participate in Federal
programs and crop insurance.
These kinds of ag data benefits, however, require agriculture to
get past roadblocks to data interoperability and over the ``not-
invented-here'' syndrome. Farmers need a uniform standard that allows
data to be portable and enables them and their trusted service
providers to make real use of the information.
The need for data interoperability is not a new issue. My written
remarks contain an excerpt from the testimony of the late Neal
Patterson, who spoke before the Senate Committee on Health, Education,
Labor and Pensions in June of 2015.\3\ Neal was a personal friend and
mentor of mine as well as co-founder and CEO for Cerner, a leading
health information technology company.
Neal believed, as I do, in the parallels between Electronic Health
Records and Electronic Field Records. His testimony stated: ``The
intersection of healthcare and IT is one of the most important in
modern society. Every citizen touches and depends on both.'' \3\
I absolutely believe the same is true for agriculture, everyone
eats. Every farmer has a right to access and use their data, regardless
of where it came from or what system contains it. We should encourage
the flow of information that could help farmers--and their trusted
advisors--to make better-informed decisions about their businesses and
food production.
In agriculture, sensor technology and communication protocols exist
for data to move quickly across different systems; however, many
existing companies are not interested in building tools that would
allow standard data to move efficiently. At Farmobile, we build
technology that supports interoperability; we are a neutral provider
that enables farmers to compare ``apples to apples'' when looking at
products and services offered to them.
It is not by accident that big ag companies use their war chests of
cash to hold farmer data hostage in their platform. They make it very
labor intensive to move the data from one system to another. I believe
in properly aligning incentive structures to drive standardization and
financially benefit farmers--who are the creators of Electronic Field
Records. The Electronic Field Record is a universal commodity in
support of digitizing agriculture, and both farmers and consumers
benefit.
Farmobile is the first company to build a business model around the
monetization of standardized farm data whereby farmers share in the
revenue, and data buyers can drive further innovation as the consumers
of this valuable information. This is a powerful new revenue
opportunity--a true win-win for farmers and the industry. (Figure 1)
\4\
The idea of farmers harvesting their data and selling it as a new
``crop'' is a game-changer. It adds economic strength to rural
communities, and also contributes to food safety--which is in the
national security conversation.
Real-time Data and the Impact on Commodity Markets
After graduating with a Bachelor of Arts in Financial Economics
from Gustavus Adolphus College in St. Peter, Minnesota, I spent the
next 20 years trading agricultural commodities--the pure economic
theory of supply and demand fascinated me and still does today. I first
traded for the Pillsbury Company and then for a large regional player,
The Scoular Company. When my career began, commodity trading was done
``in the pits'' using an open outcry system. There was an inherent time
delay to disseminate pricing data--first from the pits in Chicago to
the local grain buyers, then from the grain buyers to the farmers. This
created an unfair advantage for those, who could afford to pay for the
real time pricing feeds. For years, this opportunity was used to take
advantage of additional margin--and the farmers paid the freight for
decades.
The last 10 years of my trading career were all about challenging
the status quo in the commodity trading world and changing sides from
being the buyer to helping the seller. My company helped farmers become
better grain marketers by utilizing new technology, which enabled them
to take advantage of real-time data feeds in their marketing plans.
This opportunity was fueled by the Chicago Mercantile Exchange
acquisition of the Chicago Board of Trade which rapidly accelerated the
use of electronic trading and hedging.
Today a similar opportunity exists to ramp up the creation of farm
data into ``tradeable'' information in the form of Electronic Field
Records. To work, the data must be interoperable and available in real-
time to those who desire to purchase it. This data liquidity will
dramatically accelerate the foundational science to help solve the
looming global food challenge and identify best practices, minimize
environmental impact and maximize nutritional content of food being
produced. Every time this information is ``sold,'' it is with the
explicit permission of the farmer, and the farmer who created it shares
in the revenue. The same digital information can be sold multiple times
with an opportunity to create an estimated $1billion annually of new
revenue returned to rural agricultural communities of America.
Once you get something faster, you rarely go back. The commodity
markets are no different. I will challenge anyone to debate the notion
that real-time data, data ownership and interoperability would not be
good for the farmer.
The reason that there is so much volatility in the agricultural
markets is because of the massive time lag in getting the information.
The USDA is the gold standard in historical commodity information.
However, this information is released three weeks after it is observed
due to the process required to get that information in the right place.
The delay causes much of the volatility given the fact that most row
crops are alive about ninety days and it takes about 21 days to get the
data from the county offices to the markets in the form of USDA
reports.
The technology exists today to get that information to the market
daily. Faster information will dramatically reduce volatility in the
markets and enable traditional risk management strategies, like
hedging, to work again for farmers and the agricultural businesses that
rely on these commodities to produce their products.
Many large commercial grain trading companies have reported
significant losses in the markets recently as traditional hedging
practices are introducing more risk than they are reducing. Faster
access to better information will help normalize markets and monetarily
benefit the farmers who choose to sell licenses to their information.
Conclusion
I'd like to conclude by revisiting history. It is 1960 and John
Fitzgerald Kennedy is running for president when he visits a group of
farmers in Senator Thune's home state of South Dakota and he says:
``The farmer is the only man in our economy who has to buy everything
at retail, sell everything he sells at wholesale, and pay the freight
both ways.''
It's pretty incredible to think that--with all the change we've
seen in the last 57 years--this statement is, unfortunately, as true
today as it was then.
But disruption is coming and it's coming in the form of technology.
We need to make sure that our legal system keeps up with the technology
available. Most farmers I talk with think we have probably two to five
years to figure this out, or they will lose.
Thank you for your time today. I hope my testimony sheds some light
on what is happening in the industry and I look forward to continued
conversations about the many ways we can help the farmer finally stop
paying the freight both ways. I firmly believe that, done right, data
is the answer.
1. Farmers need policies that safeguard their data rights, and allow
interoperability and accessibility to drive efficiencies and
innovation in food production.
2. As you review this topic, ask yourselves:
Why is it ever ``o.k.'' for others to own or control a
farmer's data?
How do we affect policies for true data
interoperability?
I appreciate your openness to ideas and action from the private
sector, as well as administrative and legislative change. I look
forward to working with the industry, as well as members of the
Committee, to advance this vision.
Thank you.
cc: Addendum
Addendum
Additional References
\1\ The Problem of Vendor Lock-In for Ag, http://bit.ly/2xHeie5
\2\ Farmobile: Changing the Game in Ag Data, http://bit.ly/2oObquw
\3\ Testimony of Neal L. Patterson, Co-founder, Chairman and CEO of
Cerner Corporation, U.S. Senate Committee on Health, Education, Labor
and Pensions, Hearing: Health Information Exchange: A Path Towards
Improving the Quality and Value of Health Care for Patients, June 10,
2015, http://bit.ly/2zLI1YB
\4\ Farmobile's Business Model (Figure 1)
(Figure 1, Farmobile's Business Model)
Senator Moran. Thank you, Jason.
Dr. Ferrell.
STATEMENT OF SHANNON L. FERRELL, J.D., M.S., ASSOCIATE
PROFESSOR, OKLAHOMA STATE UNIVERSITY DEPARTMENT OF AGRICULTURAL
ECONOMICS
Dr. Ferrell. Subcommittee Chairman Moran, Ranking Member
Blumenthal, and members of the Subcommittee, thank you for the
opportunity to present my observations in the collection and
utilization of data in agriculture, the opportunities and
challenges that presents, and the legal issues surrounding
agricultural data collection, transmission, and use.
The new frontier in agriculture presents a fascinating and
sometimes paradoxical mix of cutting-edge technology, recent
legal changes, and centuries-old common law.
Farm equipment rolls off the assembly line with a suite of
sensors and transmitters enabling it to share unprecedented
amounts of farm-level or ``Small Data,'' that gives farmers the
ability to proactively manage risks that heretofore they may
have found unmanageable and, in some cases, even unknowable.
At the same time, this perfusion of Small Data can now be
aggregated by many means into what we call agricultural ``Big
Data.'' Analysis of big data in agriculture holds many
potential advantages for producers, who can apply Big Data
insights to their individual operations, and at the same time,
we now have the opportunity for better market analysis, as Mr.
Tatge was just saying, and Mr. Knopf as well, and we can now
manage agricultural risk at a national and potentially even
global scale in a few years.
Within the policy and academic realms, Big Data holds the
potential for us to provide more timely responses to industry
crises and have better evaluation of farm policy impacts and
food programs. So Big Data may eventually be able to predict
many crises before they even emerge.
However, agricultural data faces a peculiar chicken-and-
the-egg problem in that the development of datasets
sufficiently large to take full advantage of all the
opportunities ag data opposes requires participation by a large
number of producers. At the same time, farmers are often
reluctant to participate in those agricultural data systems if
they're concerned about the share of the value that they're
going to receive for the contribution of their data.
Further, despite a little potential shown by agricultural
data, the current technological, economic, and legal
environment raises some issues about how the value of
agricultural data will be shared between data aggregators and
producers, as Mr. Tatge just mentioned, so producers receiving
what they deem to be sufficient value is going to be a gateway
issue for us having a critical mass of producers making data
contributions to really truly understand the value that
agricultural data could pose for our industry.
So addressing the concerns of producers with respect to
their rights and data, the value it creates, and their privacy
if they choose to share their information is vital. Farmers
often express concerns like this collectively under the
question of, ``Who owns their data?'' And that may not be the
question that has a clear answer in our current intellectual
property framework, although, arguably, there is a colorable
argument to be made that they do own their data. The question
is, ``What does that ownership actually mean?''
So that question of ownership may not be as important as
ensuring farmers always have access to their data once it has
been shared, that they can receive value from its use, and they
can feel comfortable with the level privacy or that they're,
conversely, comfortable with the lack of privacy that they're
going to experience as a result of sharing those agricultural
data platforms.
Well, agricultural technology making use of data grows at
an exponential rate, but the technology and policies for
protecting data has not. For example, opting out of data
collection is going to grow increasingly difficult as more and
more of even the used equipment machinery fleet has embedded
technologies that make data sharing something you must opt out
of rather than opting into. And, indeed, many producers may not
even know that they have an option of opting out of data
sharing if they so choose.
Further, in some circumstances, it may be almost impossible
to truly anonymize data once it has been shared, because with
the addition of some publicly available data, we could almost
interpolate everything that you would want to know about that
operation from what data has been shared, even though,
ostensibly, it was supposed to be amalgamated with others and
rendered anonymous.
The resolution of these issues may depend on the relative
bargaining power of those at the table when data use agreements
are negotiated, and historically farmers have been at something
of a disadvantage in that regard. However, significant steps
are already underway to facilitate consensus among industry
stakeholders regarding those issues, and you'll hear about some
of those steps today, and you already have heard some of those
in passing from the previous witnesses.
There are a number of ways I think the Subcommittee and
Congress, as a whole, can facilitate the realization of
agricultural data's true potential.
First, Congress can support continuing efforts to build
industry consensus between farmers, equipment manufacturers,
and data service providers. Whether through consensus or with
legislation, Congress could also consider support of a clear
framework of right-to-know issues with respect to how your data
is being used, the right to opt out of data collection if you
so choose, guidelines for the disclosure of agricultural data
uses by service providers, and protections against the
disclosure of data. It can also fund research and educational
efforts to help agricultural producers make informed decisions
about how to engage agricultural data systems and how to
develop protections for agricultural data shared with service
providers and the government.
Finally, if the agricultural data revolution is to realize
its true potential, sustained efforts to build and maintain a
robust broadband Internet infrastructure for rural America must
be sustained, as you mentioned, Chairman. The current support
of rural broadband access from a supply perspective is having a
positive impact, but we also need to support demand-side
drivers for rural Internet access as well. Widely available
wireless and hardwired broadband connectivity both are crucial
to realizing the potential of agricultural data as well as
maintaining the economic opportunities that can revitalize
rural America.
Chairman Moran, Ranking Member Blumenthal, and members of
the Committee, thank you for the opportunity to share, and I
look forward to helping you explore this issue further.
[The prepared statement of Dr. Ferrell follows:]
Prepared Statement of Shannon L. Ferrell, J.D., M.S., Associate
Professor--Oklahoma State University Dept. of Agricultural Economics;
Jointly
prepared by the witness and Dr. Terry Griffin, Assistant Professor,
Kansas State University Dept. of Agricultural Economics
Executive Summary
Today's technology affords farmers the ability to instantaneously
collect data about almost every facet of their cropping (and
increasingly, their livestock operations) year-round. As a result,
there has been unprecedented growth in the amount of data collected at
the farm level. This farm-level ``Small Data'' increasingly provides
management insights to agricultural producers allowing them to manage
more risk factors than ever before. At the same time, this profusion of
Small Data can now be aggregated by many means to create agricultural
``Big Data.'' Analysis of Big Data in agriculture holds many potential
advantages for producers and creates the opportunity for better
macroeconomic analysis of farm policy tools, food programs, and
management of agricultural risk at a national scale.
The current technological, economic, and legal environments raise
issues about how the value of agricultural data will be captured among
the agricultural producers generating the data and the agricultural
technology providers (ATPs) aggregating it. Producers receiving what
they deem to be sufficient value for their data contributions is
critical as a potential gateway issue for making those contributions;
without large, robust participation in agricultural data systems, such
systems will fail to reach their full potential.
Thus, addressing the concerns of agricultural producers with
respect to their rights in data, the value it creates, and their
privacy if they choose to share their information is vital to see that
the agricultural industry collectively maximizes the value of these
data technologies. Farmers often express these concerns collectively as
a concern about who ``owns'' their data, and there are no clear answers
in the current intellectual property framework. However, the question
of agricultural data ownership may not be as important as ensuring
farmers always have access to their data can receive value from its
use, and can feel comfortable with the level of privacy--or lack
thereof--that can be afforded to those participating in Big Data
platforms.
Significant steps are already underway to facilitate consensus
among industry stakeholders regarding these issues. This Committee and
Congress as a whole may best be able to facilitate the realization of
Big Data's potential advantages to U.S. agriculture through support of
this consensus effort, support of educational efforts to help
agricultural producers make informed decisions about how to engage with
Big Data systems, continued development of more robust protections for
agricultural data shared with the government, and continued support of
improved broadband access in rural areas.
Acknowledgements
Dr. Terry Griffin of Kansas State University's Department of
Agricultural Economics was instrumental in the preparation of this
testimony, and his assistance in the creation of this document is
gratefully acknowledged. Dr. John Fulton of Ohio State University's
Department of Food, Agricultural, and Biological Engineering, Ms.
Maureen Kelly Moseman, Adjunct Professor of Law at the University of
Nebraska College of Law, Mr. Todd Janzen of the Janzen Agricultural Law
firm, Mr. Ryan Jenlink of the Harness, Dickey & Pierce law firm, Dr.
Keith Coble of the Mississippi State University Department of
Agricultural Economics, Dr. Ashok Mishra of the Arizona State
University Morrison School of Agribusiness, and Mr. Matthew Steinert of
Steinert Farms, LLC also provided vital input in the development of
this testimony.
Perhaps the greatest contribution to this testimony and my
understanding of agricultural data systems, though, was made by Dr.
Marvin Stone. Dr. Stone was a giant in the agricultural data field,
contributing tremendously to the development of the Green Seeker
technology that significantly advanced machine-sensing of plant health.
He was also instrumental in the development of the SAE J1939 standard
that forms the foundation for many of the machine data technologies at
the heart of this discussion. Beyond being a giant in the field we
examine here today, Dr. Stone was a mentor to myself and hundreds of
other students at Oklahoma State University. He and his wife were both
killed in the tragic Oklahoma State University homecoming parade
accident of October 24, 2015. I hope this testimony honors his memory,
the contributions he made to this field, to the U.S. agriculture
industry, and to all his students.
Issue Analysis
1. Introduction
I would like to thank Subcommittee Chairman Moran, Ranking Member
Blumenthal, and the Members of the Committee for the opportunity to
present my observations on the collection and utilization of data in
agriculture and the legal issues surrounding the concept of Big Data
and its application to U.S. farmers and ranchers. This new frontier in
agriculture presents a fascinating and sometimes paradoxical mix of
cutting edge technology, recent legal changes, and centuries-old
doctrines of common law. In my testimony today, I will discuss how both
``Small Data'' and ``Big Data'' in agriculture are being utilized by
agricultural producers and what lies just over the horizon for those
technologies. I will also discuss some of the opportunities and
challenges posed by the advancements in agricultural data technology.
Then, I lay a framework for discussing the legal issues surrounding Big
Data in agriculture, discuss how the current U.S. legal environment
addresses ownership and privacy rights in agricultural data, and
suggest some potential avenues for policy responses that may facilitate
the economic advantages to be gained from the application of Big Data
principles to agricultural data while dealing with the concerns
associated with such applications.
2. The growth of Small Data and Big Data in production agriculture
The concept of Big Data has exploded in a relatively short period
of time. However, there would be no Big Data in agriculture were it not
for Small Data. Since these definitions and the issues surrounding data
use in agriculture continue to evolve, my testimony today will provide
some framing for both.
2.1 Defining core terms in the Data-Driven Farming discussion
Three terms immediately rise to the top in an examination of the
agricultural data discussion: agricultural data, Small Data, and Big
Data. Taken together, the use of Small Data and Big Data in agriculture
is increasingly referred to as ``digital agriculture.''
The concept of agricultural data is almost too broad to define, but
looking at research in the field and conversations surrounding
agricultural data indicates the term centers around two more specific
concepts: ``telematics'' or ``machine'' data and ``agronomic'' data.
Telematics data (sometimes called ``machine data'') refers to the
information an agricultural implement (such as a planter) or self-
propelled vehicle (such as a tractor or combine) collects about itself.
Almost by definition, telematics data comes from agricultural equipment
owned, operated, or hired under contract by the agricultural producer.
Agronomic data refers to information about a crop or its environment,
such as ``as-planted'' information from a seed planter, ``as-applied''
information from a fertilizer sprayer, yield data from a grain combine,
and so on. While agronomic data resembles telematics data in that much
of it is gleaned directly from agricultural implements, agronomic data
can also be obtained from many other sources such as hand-held sensors,
aerial platforms such as manned survey flights or flights by unmanned
aerial systems (UAS, commonly called ``drones''), and even satellite
imagery.
Another piece of the agricultural data puzzle is so-called
``metadata,'' which includes management information such as seeding
depth, seed placement, cultivar, machinery diagnostics, time and
motion, dates of tillage, planting, scouting, spraying, and input
application. In addition to data on the products and how those products
are applied, information on external environmental circumstances such
as weather including precipitation events, evapotranspiration, and heat
unit accumulation help to round out the complete agricultural data
package.\1\
---------------------------------------------------------------------------
\1\ T. Griffin, et al., ``Big Data Considerations for Rural
Property Professionals.'' Journal of the American Society of Farm
Managers and Rural Appraisers, 2016:167, 168.
---------------------------------------------------------------------------
Beyond these data sources, numerous other data sources continue to
emerge in the agricultural data space. Work continues to build data
collection technology in the livestock industries, ranging from GPS-
enabled cattle ear tags to ``bolus'' sensors that can be swallowed by
animals to provide health data. Some would argue that vendor-generated
data about producers might also fit into this category; such data could
include everything from payment history data to customer relationship
management (CRM) information (does the producer try to negotiate input
prices, have preferences for some products over others, typically buy
inputs from one salesperson versus others, etc.).
Agricultural data is the foundation for Small Data systems. In
simplest terms, farms use ``Small Data'' when data are isolated to the
fields where the data originated. Farmers who use information
technology to conduct their own on-farm experiments, document yield
penalties from poor drainage, or negotiate crop share agreements are
using data that is considered ``small.''
Perhaps ironically, the evolution and revolution in agricultural
Big Data comes from the expansion of ``Small Data'' in agriculture.\2\
There has been remarkable growth in producers' ability to collect data
pertaining only to their own operation through the growth of techniques
and technologies such as grid soil sampling, telematics systems for
farm equipment, Global Positioning Systems (GPS)/Global Navigation
Satellite Systems (GNSS), farm aerial imagery acquired via small
unmanned aerial systems (sUAS) and the like. Producer adoption of these
information technologies has increased dramatically in recent years,\3\
giving rise to a profusion of agricultural data heretofore unseen.\4\
---------------------------------------------------------------------------
\2\ K. Coble, T. Griffin, A. Misrha, and S. Ferrell, ``Big Data in
Agriculture: A Challenge for the Future,'' forthcoming in Applied
Economics and Policy Perspectives (accepted for publication October 20,
2017).
\3\ T. Griffin, Miller, N.J., Bergtold, J., Shanoyan, A., Sharda,
A., and Ciampitti, I.A. 2017. Farm's Sequence of Adoption of
Information-Intensive Precision Agricultural Technology. Applied
Engineering in Agriculture 33(4):521-527 doi: 10.13031/aea.12228.
\4\ B. Erickson, and D. Widmar. 2015. Precision Agricultural
Services Dealership Survey Results. West Lafayette, Indiana, Purdue
University, August. Accessed June 21, 2016: http://
agribusiness.purdue.edu/files/resources/2015-crop-life-purdue-
precision-dealer-survey.pdf
---------------------------------------------------------------------------
The new abundance of field-level information provided by these
technologies could improve the ability of producers to make profit-
maximizing decisions benefitting the producer operating the field, i.e.
Small Data.\5\ However, pooling the datasets of hundreds or thousands
of fields could hold a much greater potential value both to individual
producers and the agricultural industry as a whole. Agricultural Big
Data--farm data that has been combined into an aggregate form--has the
potential to reveal undiscovered insights. Currently, only limited
quantitative evidence exists regarding the value of assembling data
from precision agriculture technology into a community; however,
indirect evidence suggests farm data has economic value.
---------------------------------------------------------------------------
\5\ Griffin, supra note 1.
---------------------------------------------------------------------------
While the term Big Data is relatively new, it refers to a concept
that is not. There are many definitions for the term, but a straight-
forward one might be ``a collection of data from traditional and
digital sources inside and outside your company that represents a
source for ongoing discovery and analysis.'' \6\ While this definition
sounds much like traditional data analysis (and it is), recent advances
in both data collection and transmission increase the analytical power
of data analysis procedures by orders of magnitude. The ``big'' in Big
Data comes from the fact data sets continue to grow exponentially both
in breadth (with more and more firms collecting data) and depth (with
data from more and more sources long the food supply chain being
aggregated by more firms). Conceptually, Big Data is defined as the
analysis of datasets requiring advanced tools to manage the data due to
four factors: volume, velocity, variety, and veracity.
---------------------------------------------------------------------------
\6\ L. Arthur. 2013. What is big data? Forbes, CMO Network blog
entry. Available at http://www.forbes.com/sites/lisaarthur/2013/08/15/
what-is-big-data/, last accessed November 15, 2014.
Table 1: Big Data defining factors
------------------------------------------------------------------------
Factor Definition
------------------------------------------------------------------------
Volume The sheer amount of data precludes its storage on a
single computer system; analytic software must
aggregate the data from multiple systems
------------------------------------------------------------------------
Velocity New data enters the analysis continuously at high
rates of transmission.
------------------------------------------------------------------------
Variety Data is aggregated from a variety of sources, many
of which may use different data formats.
------------------------------------------------------------------------
Veracity Accuracy of the data is vital to correct analysis,
while the data source may apply varying (or no)
methods of data validation. Thus the Big Data
system may have to independently validate the data
or make assumptions about its accuracy.
------------------------------------------------------------------------
Agricultural data has arguably already crossed over into the realm
of Big Data as measured by these factors.
Existing technologies can already generate over 10 MB of data per
acre, and when extrapolated over the 90 million acres of corn ground in
the U.S., this means 900 terabytes (TB, 1 TB being equal to 1,000,000
MB) of data could be generated on corn acres alone.\7\ A student at
Ohio State University recently completed the ``Terra Byte'' project to
determine how much data could be garnered from one corn plant, with a
resulting 18.4 gigabytes) of data; over a 100 acre corn field, this
would be the equivalent of 60 petabytes (PB, 1 PB being equal to
1,000,000,000 MB).\8\ Already, the commodity dataset has grown too
large to be transported via broadband connections, or even physically
via external hard drives, meaning analytical software must go to the
data \9\ Thus, the volume requirement for Big Data is satisfied.
---------------------------------------------------------------------------
\7\ T. Griffin, ``Can Agricultural or Farm Data Be Considered Big
Data?'' Kansas State University, https://www.agmanager.info/machinery/
precision-agriculture/precision-ag-farm-data-blog
/can%C2%A0agricultural%C2%A0or%C2%A0farm%C2%A0data%C2%A0be (last
visited November 8, 2017).
\8\ M. Brookhart and M. Reese. ``World Record for Data Collection
Set by OSU Precision Ag Team.'' Ohio Country Journal, October 11, 2017,
http://ocj.com/2017/10/world-record-for-data-collection-set-by-osu-
precision-ag-team/#.Wd45GMwR0Qo.twitter (last visited November 8,
2017).
\9\ Grifin, supra note 7.
---------------------------------------------------------------------------
Looking only at as-planted data collected from planters via
telematics, 5.5 MB of data on location, speed, cultivar, and other geo-
spatial and meta-data are collected for each acre planted. During
planting seasons, the size of the aggregated farm data community
becomes much larger every day. Although agricultural operations are
seasonal, it should be recognized that even for commodity crops like
corn, cotton, soybean, rice, and wheat that peak planting times differ
for each such that as-planted data are collected during several months
of the year rather than all at once. In addition to planting, other
field operations such as tillage, spray applications, and harvest occur
at other times during the season; each operation adding to the
community of data. Thus, Griffin observes, planting data alone would
satisfy the ``velocity'' component of Big Data.\10\ By the same token,
each of these data points are being collected by different brands of
equipment using different file formats and supplemented using manually-
collected data such as soil samples, all of which may be reported in
non-standard formats, satisfying the ``variety component.'' \11\
---------------------------------------------------------------------------
\10\ Id.
\11\ Id.
---------------------------------------------------------------------------
That leaves the ``veracity'' component and agricultural data can
certainly pose veracity challenges. Such challenges arise from the
problems inherent in trying to measure biological processes by
mechanical means. Data quality has been a contentious topic in
precision agriculture for decades; especially regarding raw yield
monitor data and other farm data collected by on-the-go sensors. A part
of the debate on the veracity of yield data involves whether the farmer
or combine operator properly calibrates the yield monitor. Therefore,
both sensors and human error influence farm data quality. Given this,
agricultural data appears to more than satisfy the Big Data test.\12\
---------------------------------------------------------------------------
\12\ Id.
---------------------------------------------------------------------------
Although not as prominent to the discussion as Big Data and
agricultural data, another important term to define is service
provider. Service provider (sometimes called an ``Agricultural
Technology Provider'' or ``ATP'') is the term frequently used to
describe a party external to the farm providing some service regarding
either crop production or management of the crop enterprise. Crop
production services could include fertilizer or chemical applicators,
custom operators, or harvest contractors whose equipment generate
agricultural data regarding the farm. Management services include
traditional services such as crop consulting and scouting, but
increasingly include services targeted specifically at data collection
and analysis.
2.2 Opportunities and Challenges arising from Small and Big Data use in
Agriculture
It is important to note this discussion would not occur were it not
for the tremendous potential the nascent farm data revolution promises.
Existing technologies such as real-time kinematics (RTK) and ``auto-
steer'' (sometimes referred to as ``GNSS-enabled navigation technology)
have already provided substantial economic returns to farmers.\13\
Improved sensing of soil conditions, crop health, and yields has led to
significantly improved management information for agricultural
producers. As mentioned above, this represents Small Data, with data
generated--and decisions made--at the farm level.
---------------------------------------------------------------------------
\13\ See, e.g., M. Darr, ``Big Data and Big Opportunities,'' paper
presented at PrecisionAg Big Data Conference, August 21, 2014 (Ames,
Iowa).
---------------------------------------------------------------------------
To date, much of the gains from improved sensing technologies and
their sharing with service providers have come from eliminating
inefficiencies in the utilization of agronomic and machinery inputs.
Put another way, we have seen significant increases in the use of Small
Data.
Small Data sees a variety of farm-level uses. Data kept isolated to
the originating farm has value, but the value of that data is limited
to just that farm or potentially to farms in relative proximity. The
primary uses of farm data are those for which the data were initially
generated such as documenting within-field site-specific yields with a
yield monitor.\14\ Typically, primary uses of data are restricted to
the field that the data originated; consider the analogy of using a
computer when that computer is not connected to the Internet. Primary
data uses are ``local'' to the field or operation from which they
originate and are not connected to data from other areas.
---------------------------------------------------------------------------
\14\ Note that secondary uses of data will be discussed later in
this testimony.
---------------------------------------------------------------------------
Considerable effort has been made by farmers, researchers, and
others from within and external to the agricultural industry to
profitably utilize data generated from precision agricultural
technologies. The majority of these efforts have historically focused
on one-field-at-a-time or maybe even at the whole farm level but for
only that one farm. The value of farm data when isolated to a specific
farm has been limited and only of value to that particular farm (or
some value for the next farmer of the land). At the very least, the
value of that data decays very quickly with distance from the field.
Indeed, it is possible that the site-specific value of farm data
might actually play a role in farmland values themselves. Griffin and
Taylor \15\ explored how big data could impact farmland values and
rental rates, stating ``It remains unclear whether the `data premium'
[for farmland conveyed with a significant farm-specific dataset] will
be a true premium (an amount added to the market price of land) or a
penalty (an amount deducted from the market price of land). In the
short-run, early movers who choose to provide data to land buyers may
see a premium. However, as the transfer of data with a land sale
becomes more common, a penalty to land parcels without data may become
more common.'' They also describe how biophysical data, such as
historical yield, soil test results, and other production data have
been included in farmland sales and/or rental agreements, but they
suggest these data have not substantially influenced farmland values
nor are sufficient to be considered ``big.'' These historical data
could be annual whole-field yield written on paper or site-specific
geospatial data including GPS yield monitor data or grid soil samples
in either electronic form or printed maps. Although the above mentioned
data may provide evidence of historical productivity and soil amendment
utilization, they do not impact farmland values directly. Farmland
values and rental rates will likely be a function of both quantity and
quality of geospatial metadata once the big data sector of the
agriculture industry matures.
---------------------------------------------------------------------------
\15\ T. Griffin., and Taylor, M.R. (2015). Precision Agriculture
Data Impact on Farmland Values: Big Data in Ag. K-State Department of
Agricultural Economics AgManagerInfo AM-TWG-PRAG-4.2015 Online: http://
www.agmanager.info/crops/prodecon/precision/PrecisionAgData
_FarmlandValues.pdf., last visited November 8, 2017.
---------------------------------------------------------------------------
Farmers have made use of precision agriculture technology and farm
data in a variety of ways, and oftentimes in ways that the
manufacturers had not anticipated. An early report on how farmers used
yield monitors indicated the primary uses of yield data include but not
limited to: 1) conduct on-farm experiments, 2) tile drainage decisions,
and 3) split crop share rents.\16\ To estimate the value of farm data
for each of these examples, the alternative decision making process
must be evaluated. However, back of the napkin extreme examples make
the point that the value of the above scenarios are finite and limited
to a single farm.
---------------------------------------------------------------------------
\16\ T. Griffin, ``Farmers' Use of Yield Monitors,'' University of
Arkansas Fact Sheet FSA36, available at https://www.uaex.edu/
publications/pdf/FSA-36.pdf (last visited November 8, 2017).
---------------------------------------------------------------------------
Perhaps the most dramatic gains lie ahead, though, as agriculture
puts the ``Big'' in Big Data by compiling datasets of sufficient size
to enable much more robust statistical analyses of multiple factors
influencing commodity production. Examples of how the aggregation of
farm data across large datasets can significantly increase value to
farmers are illustrated in Table 2 below.\17\
---------------------------------------------------------------------------
\17\ Table and scenarios taken from Terry Griffin, ``Big Data
Considerations for Agricultural Attorneys,'' paper presented at
American Agricultural Law Association Annual Symposium, October 23,
2015 (Charleston, South Carolina).
Table 2: Comparison of Primary and Secondary Agricultural Data Uses
----------------------------------------------------------------------------------------------------------------
Data Primary Use ``Small Data'' Secondary Use ``Big Data''
----------------------------------------------------------------------------------------------------------------
Yield monitor data Documenting yields; on-farm seed Genetic, environmental, management effect (G x E x
trials M) analyses
----------------------------------------------------------------------------------------------------------------
Soil sample data Fertilizer decisions Regional environmental compliance
----------------------------------------------------------------------------------------------------------------
Scouting Spray decisions Regional analytics of pest patterns
----------------------------------------------------------------------------------------------------------------
As an example of initial or primary use of farm data, yield monitor
data on one farm can help document the farm's productivity on a field-
by-field basis and can illustrate how a seed hybrid performed on that
farm in one year, given the environment of that farm for that year and
the management practices employed during that year. Interesting
opportunities arise when that data is ``re-used'' in Big Data
aggregation with similar data across hundreds or even thousands of
farms, and this aggregation creates the bridge linking Small and Big
Data.
Such aggregation allows for the evaluation of that cultivar across
tens of thousands of permutations of factors such as management
practices, soil type, and climate. This enables both seed companies and
agricultural producers to learn via observational data in one or two
years what would take decades of collections by use of traditional seed
trials via experimentation. Soil sample data coupled with yield data
can inform an agricultural producer about the nutrient uptake of the
crop on his or her farm, but Big Data could allow all the agricultural
producers in a region to effectively tackle nutrient loading to
impaired water bodies through voluntary management of non-point
pollution. Crop scouting can help an individual agricultural producer
make decisions about the application of a specific pesticide, but Big
Data could allow a crop industry to spot trends in plant pathogens that
could be used to head off the spread of potentially devastating plant
health threats. The true maturity of Big Data in agriculture may come
when the value of secondary uses realize greater aggregate economic
value than the primary uses of the data.\18\
---------------------------------------------------------------------------
\18\ V. Mayer-Schonberger, and K. Cukier, Big Data: A Revolution
That Will Transform How We Live, Work, and Think, Kindle Edition.
Houghton Mifflin Harcourt Publishing Company, New York, NY. 257 pp.
2014.
---------------------------------------------------------------------------
The integration of Small Data and Big Data at the farm level could
hold important implications for farms competitiveness.\19\ Early
adopters of big data in other industries (such as healthcare,
transportation, and retail) are shown to have gained a competitive
advantage within their industries and have realized significant
increases in operating margins.\20\ There is an emerging discussion in
the agribusiness industry and its literature about the potential of big
data and its capacity to change the basis of competition in
agriculture.\21\ This belief is based on the previous trends in the
history of innovations powering productivity and enhancing
competitiveness in the agri-food supply chain, enabled by information
and communication technology (ICT). Among such examples is precision
agriculture powered by GPS, remote sensing, and variable rate
technology (VRT) technologies in crop farming. While the adopters of
ICT-based applications in agricultural production were primarily
motivated by the efficiency gains, they also have laid the foundation
for the big data infrastructure within agriculture. As a result, modern
farms are generating, or have a capacity to generate, a substantial
amount of agricultural production data. This data becomes an important
intangible resource alongside the physical and human resources, which
if managed effectively, can produce substantial value for the farming
operation. The important question to ask is under which circumstances
the data, as an intangible resource, can become a source of competitive
advantage?
---------------------------------------------------------------------------
\19\ This discussion of agricultural data and competitive issues is
taken from Griffin, et al., supra note 1.
\20\ J. Manyika, Chui, M., Brown, B., Bughin, J., Dobbs, R.,
Roxburgh, C., & Byers, A. H. (2011). ``Big data: The next frontier for
innovation, competition, and productivity.'' McKinsey Global Group
report, available at https://www.mckinsey.com/business-functions/
digital-mckinsey/our-insights/big-data-the-next-frontier-for-
innovation, last visited November 8, 2017).
\21\ S. Sonka. (2014). Big Data and the Ag Sector: More than Lots
of Numbers. International Food and Agribusiness Management Review,
17(1), 1-20. Available online at http://www.ifama.org/files/IFAMR/
Vol%2017/Issue%201/(1)%2020130114.pdf, last visited November 8, 2017.
---------------------------------------------------------------------------
Beyond the benefits of Big Data to production agriculture, it also
presents the agricultural economics community with numerous
opportunities to enhance and expand the analysis of numerous
microeconomic, macroeconomic, and agricultural policy issues.\22\ For
example, microeconomic farm management issues could now be analyzed by
aggregating data across thousands of farms using management decisions
as variables instead of using a farm-by-farm case study approach. Food
program evaluations, regulatory impact analysis, and demand estimation
could be accomplished by rapid aggregation and analysis of grocery
store UPC scanner data. Geospatial analysis of crop yields could lead
to improved precision in the pricing of crop insurance products. Broad
environmental sensor networks coupled with farm data could
significantly enhance the ability to manage crop fertilizer
applications to minimize nutrient runoff impacts.
---------------------------------------------------------------------------
\22\ The following examples are taken from K. Coble, T. Griffin, A.
Misrha, and S. Ferrell, ``Big Data in Agriculture: A Challenge for the
Future,'' forthcoming in Applied Economics and Policy Perspectives
(accepted for publication October 20, 2017).
---------------------------------------------------------------------------
To understand the potential policy implications of Big Data's
growth in agriculture, one must recall that one of the defining
characteristics of agricultural Big Data is combining data from
multiple farms into a community. A leading reason for this is that each
farmer becomes a variable (rather than a constant) once a critical mass
of farms is in the community. When farm data were isolated to a single
farm, then there was no opportunity to evaluate the management
practices specific to that farmer, i.e. the management was held
constant.
Farm data must be aggregated to perform community analysis. A
leading example of community analysis is evaluating how a product (G
for ``genetics,'' from classic varietal or hybrid tests) in a given
location (E for ``environment,'' including soils, weather, and other
uncontrolled factors) under the farm's production practices (M for
``management,'' including controlled factors such as planting dates,
seeding rates, timing of operations, tillage practices and many
others). When farm data are not aggregated across numerous farms, then
the data remain `small' and the value is limited since the M in
analysis known as GxExM, is not a viable variable (only the traditional
GxE). When data are aggregated such that M is a variable to the
analysis GxExM, insights can be discovered for a majority of
participants. Examples of previously unknown discoveries may include
which products or bundle of products (seed, fungicides, planting dates)
maximize profitability for a given region under specific farm
production practices.
Each player (and each group of players) benefit differently with
respect to the big data system. One must consider how these different
players benefit to comprehend how the value of Big Data systems may be
captured relative to the data contributors (farmers) and aggregators
(ATPs). The economics of networks are important to fully understand the
value gained from the big data community. The value of the data
community depends not only on the quality of the data but on how many
others participate in the system. Data from numerous farms aggregated
into a community are more valuable than data from any one individual
farm. In the long run, the aggregator controlling the flow of data
enjoys the majority of the value. Other groups, such as those offering
analytic services of the aggregated data, enjoy their value capture
especially in the short run. Once a critical mass of farms are in the
data community, i.e. the long run, farmers' bargaining power with the
data aggregator likely will be greatly reduced.
In the long-run the majority of the value will be enjoyed by the
one controlling the data community, i.e. the data service provider.
Other players such as input manufacturers, retailers, and advisors may
enjoy their own levels of varying value capture. The important part to
be cognizant is that 1) the farmer is not the only player at the big
data table and 2) the farmer is not likely to receive the vast majority
of the value from participating in the big data system. However, that
is not to say that farmers will not still see potentially important
benefits from the analyses provided by Big Data systems. Such systems
pose the opportunity of providing potentially unprecedented insights to
inform farm management decisions, decreasing production risk, and
potentially reducing financial and market risks as well.
While there are countless potentially positive uses of Big Data
tools, any tool can also be misused. Farmers, ranchers, and other
participants in the agricultural industry have expressed concerns about
several potential misuses of agricultural data beyond the mere
disclosure of confidential information (discussed below). Some
producers worry that the ability of equipment manufacturers to access a
significant amount of data about their operations, giving the
manufacturers the ability to interpolate the farmer's financial
condition and use such information to an unfair advantage in
transactions with the farmer or to alter the balance of negotiating
power in the manufacturer's favor for any number of transactions.
Others worry about government agencies taking advantage of aggregated
datasets to acquire information that the producer could not be
compelled to produce without a formal legal process. Yet another
concern is that falsified data could be introduced into individual or
aggregated agricultural datasets to skew environmental assessments of
farm performance.
One additional Big Data challenge worries both producers and
economists. As stated in Coble, et. al: \23\
---------------------------------------------------------------------------
\23\ Supra, note 4.
The Holy Grail for market participants is to get perfect
information as soon as it is knowable, and preferably before it
is knowable to others. While Big Data has a long, long way to
go before achieving this, bigger steps toward that goal are
being taken faster than ever before. Thus, a significant
concern with aggregating agricultural data is whether--either
legitimately or not--a small number of market participants (or
a single actor) could get access to information sufficient to
move (or even manipulate) markets faster than, or to the
exclusion of, other market participants. While there are
numerous rules in place to deal with a broad range of market-
manipulating activities, none of these current rules
contemplate the type of actions that could take place with a
sufficiently large aggregated dataset. Currently, there are
various rules restricting insider trading (see 17 C.F.R.
Sec. 1.59(a), 17 C.F.R. Sec. 1.3(ee)), and government employees
are prohibited from using data for financial gain that has not
been disseminated to the public (7 U.S.C. Sec. 6c(a)(3)).
However, there are no rules governing ``very good market
information'' such as that which could be obtained through
completely legal means by aggregating sufficient telematics
data (as an example). As a result, research on the potential
market effects of growing market asymmetries that could be
triggered by growing Big Data aggregations and the implications
of policies restricting the use of aggregated data in commodity
market transactions could do much to inform the development of
---------------------------------------------------------------------------
law in the arena.
Only time and experience will tell whether these concerns are well-
founded, but the fact they exist may well impact producers' willingness
to participate in Big Data systems, and thus impact the future of the
industry. Most industry observers believe the benefits to individual
producers and the agricultural industry as a whole far outweigh the
potential risks. However, bringing about the full economic benefits of
Big Data in agriculture requires a robust system by which large numbers
of agricultural producers can share their data since the predictive
power of statistical analysis increases with the number of observations
available for each variable examined.\24\ The agricultural data
industry is working tirelessly to create those systems. The issue is
one of trust--farmers must feel they can trust Big Data systems before
they will participate. Thus, the issue of most concern to this hearing
may not be whether we will have systems that can accept and analyze
that data; it is perhaps how Congress can facilitate the development of
an environment in which farmers will share their data. Metcalfe's Law
states that the value of a network is proportionate to the number of
its members. Put another way, Facebook has little value if you are its
only member, but it has tremendous value when populated by millions of
members. Thus, agricultural producers can only harness the value of Big
Data if we can foster an environment in which they are comfortable
sharing their data. However, that participation might be inevitable
given the increasing prevalence of data-collection technologies. As
Griffin and Shanoyan observe, going ``off the grid'' with respect to
agricultural data may be possible in the near term, but eventually will
require farmers to use then-antiquated technology, placing them at
further competitive disadvantage.\25\
---------------------------------------------------------------------------
\24\ See generally George G. Judge, et al, Introduction to the
Theory and Practice of Econometrics (2nd ed, 1988), 96.
\25\ T. Griffin and A. Shanoyan, ``Is Going Off the Grid Possible
in the Age of Farm Data?'' Kansas State University, https://
www.agmanager.info/machinery/precision-agriculture/precision-ag-farm-
data-blog/going-grid-possible-age-farm-data (last visited November 7,
2017).
---------------------------------------------------------------------------
Given this potential inevitability of data sharing, one must turn
to questions of what rights farmers can retain in their shared data. Do
they retain ownership of their information? Is there any hope of
retaining their privacy in that information once it is shared?
2.3 Framing the legal issues surrounding data in agriculture
The issues involved in the discussion of data in agriculture are
almost innumerable, but many can be brought under the umbrella of two
over-arching concepts: agricultural data ownership, and protections
against its unauthorized disclosure. Although each of these issues is
discussed in greater detail later in this testimony, a brief framing of
each issue is provided here.
2.3.1. Ownership of agricultural data
As agricultural producers began to realize the information they
were generating (and, in some cases, sharing with service providers)
had potential economic value, questions began to arise regarding who
had the superior ``ownership'' right to that information, given that
multiple parties had a hand in its creation. Further, the realization
of agricultural data's value changed the relative negotiation power
between parties. This is an important concept; if their data is shared
by someone other than them with a third party, that sharing may cause
the farmer to lose negotiation power with vendors, landlords, and the
like as a result. Thus, farmers may wish to assert ``ownership'' of
data so as to exercise one of the rights of property ownership, namely,
to exclude others from its use. Thus, this issue might be framed as
``Who owns data generated about an agricultural producer's operation?''
2.3.2. Privacy rights for agricultural data
As discussed in more detail below, it is possible--and even likely
-the greatest economic value of agricultural data to the farm owner
comes not from his or her own analysis of the data but from its
aggregation with data from hundreds or even thousands of other farms
(in a true Big Data model) to provide management information and trend
identification that could not be derived from any smaller dataset. For
example, one of the most common analyses provided by ATPs to farmers
are ``comparative analytics'' (for example, benchmarking performance
relative to similarly-situated operations). While that might have some
economic value for the producer, much greater benefits await via
advanced analysis. The balance of negotiating power between the farmer
and the aggregator will eventually determine what proportion of the
analyses conducted benefit each party. While aggregation may in some
ways reduce the disclosure or discovery of information about any one
farm (through the anonymization of individual farm data by aggregation
with many other farms), it naturally also raises fears about the
release of that information (whether the result of intentional activity
such as database hacking or an accidental disclosure). This leads to
the second question: ``What protections prevent the disclosure of
agricultural data to outside parties?''
3. Current Legal Framework for Ownership of Agricultural Data
The United States has one of the most robust systems of property
rights in the world, empowered by a legal system making it easy
(relatively speaking) to enforce those rights. Thus, the first place
many look for a means of protecting one's data from misappropriation
and/or misuse is the property right system. This requires one to
examine who ``owns'' agricultural data. The answer to the question is
not simple, though, as traditional notions of property ownership find
challenge in their application to pure information.
The notion of property ownership typically involves some form of
six interests, including the right to possess (occupy or hold), use
(interact with, alter, or manipulate), enjoy (in this context, profit
from), exclude others from, transfer, and consume or destroy. Some of
these interests do not fit, or at least do not fit well, with data
ownership. Excluding others from data, for example, is difficult,
particularly when it is possible for many people to ``possess'' the
property without diminishing its value to the others, just as the value
of a book to one person may not be diminished by the fact other people
own the same book.\26\ Thus, the better question may be ``What are the
rights and responsibilities of the parties in a data disclosure
relationship with respect to that data?'' \27\
---------------------------------------------------------------------------
\26\ L. Smith. 2006. ``RFID and other embedded technologies: who
owns the data?'' Santa Clara Computer and High Technology Law Journal
\27\ R. Peterson. 2013. ``Can data governance address the conundrum
of who owns data?'' Educause blog, http://www.educause.edu/blogs/
rodney/can-data-governance-address-conundrum-who-owns-data, last
accessed November 8, 2017.
---------------------------------------------------------------------------
Data is difficult to define as a form of property, but it most
closely resembles intellectual property. As a result, the intellectual
property framework serves as a useful starting point to define what
rights a farmer might have to their agricultural data. Intellectual
property can be divided into four categories: (1) trademark, (2)
patent, (3) copyright, and (4) trade secret. The first three areas
compose the realm of Federal intellectual property law as they are
defined by the Constitution as areas in which Congress has legislative
authority.\28\ Since trademark is not relevant to a discussion about
data,\29\ the analysis will focus on patent, copyright, and trade
secret.
---------------------------------------------------------------------------
\28\ U.S. Constitution, Article I, Sec. 8, clause 8.
\29\ The Federal Trademark Act (sometimes called the Lanham Act)
defines trademark as ``any word, name, symbol, or device, or any
combination thereof . . . to identify and distinguish his or her goods,
including a unique product, from those manufactured or sold by others
and to indicate the source of the goods, even if that source is
unknown.'' 15 U.S.C. Sec. 1127.
---------------------------------------------------------------------------
3.1 Application of patent law to agricultural data
The U.S. Patent Act states ``whoever invents or discovers any new
and useful process, machine, manufacture, or composition of matter, or
any new and useful improvement thereof, may obtain a patent therefor''
(35 U.S.C. Sec. 101). Generally, for an invention to be patentable, it
must be useful (capable of performing its intended purpose), novel
(different from existing knowledge in the field), and non-obvious
(somewhat difficult to define, but as set forth in the Patent Act, ``a
patent may not be obtained. . .if the differences between the subject
matter sought to be patented and the prior art are such that the
subject matter as a whole would have been obvious at the time the
invention was made to a person having ordinary skill in the art to
which said subject matter pertains'').\30\ Patent serves as a poor fit
for a model of agricultural data ownership since it protects
``inventions.'' Raw data, such as agricultural data, would not satisfy
the definition of invention.
---------------------------------------------------------------------------
\30\ 35 U.S.C. Sec. Sec. 102, 103.
---------------------------------------------------------------------------
It should be noted patentable inventions could be derived from the
analysis of agricultural data. While this does not mean the data itself
is patentable, it does suggest that any agreement governing the
disclosure of agricultural data by the agricultural producer should
address who holds the rights to inventions so derived.
3.2 Application of copyright law to agricultural data
The Federal Copyright Act states the following:
Copyright protection subsists, in accordance with this title, in
original works of authorship fixed in any tangible medium of
expression, now known or later developed, from which they can be
perceived, reproduced, or otherwise communicated, either directly or
with the aid of a machine or device. Works of authorship include the
following categories:
literary works;
musical works, including any accompanying words;
dramatic works, including any accompanying music;
pantomimes and choreographic works;
pictorial, graphic, and sculptural works;
motion pictures and other audiovisual works;
sound recordings; and
architectural works.\31\
---------------------------------------------------------------------------
\31\ 17 U.S.C. Sec. 102(a).
More so than trademark and patent, the copyright model at least
resembles a model applicable to agricultural data. At the same time,
however, the model also has numerous problems in addressing
agricultural data. First, the list of ``works of authorship'' provided
in the statute strongly suggests a creative component is important to
the copyrightable material. Second, the term ``original works of
authorship'' also has been interpreted to require some element of
creative input by the author of the copyrighted material. This
requirement was highlighted in the case of Fiest Publications Inc. v.
Rural Telephone Service Company,\32\ where the U.S. Supreme Court held
the Copyright Act does not protect individual facts. In Fiest, the
question was whether a pure telephone directory (consisting solely of a
list of telephone numbers, organized alphabetically by the holder's
last name) was copyrightable. Since the directory consisted solely of
pure data and was organized in the only practical way to organize such
data, the Supreme Court held the work did not satisfy the creative
requirements of the Copyright Act.\33\ This ruling affirmed the
principle that raw facts and data, in and of themselves, are not
copyrightable. Put another way, the fact that ABC Plumbing's telephone
number is 555-1234 is not copyrightable. However, an author can add
creative components to facts and data such as illustrations,
commentary, or alternative organization systems and can copyright the
creative components even if they cannot copyright the underlying facts
and data. Continuing the analogy, ABC's phone number alone is not
copyrightable, but a Yellow Pages ad with ABC Plumbing's number
accompanied by a logo and a description of the company's services would
be copyrightable.
---------------------------------------------------------------------------
\32\ 499 U.S. 340 (1991).
\33\ See id.
---------------------------------------------------------------------------
Agricultural data in and of itself may not be copyrightable, but it
can lead to copyrightable works. For example, agricultural data may not
be copyrightable, but a report summarizing the data and adding
recommendations for action might be. Again, then, it is incumbent upon
those disclosing agricultural data to include language in their
agreements with the receiving party to define the rights to such works
derived from the data.
A separate issue regarding copyrights deriving from agricultural
data also continues to emerge. Increasingly, the original agricultural
data is never even disclosed to the agricultural producer; rather, the
data has been processed into a report or a new form through use of a
computer algorithm. Quite simply, agricultural producers may often
receive a completely computer-generated report with no human author.
This requires moving into the realm of copyrights in computer generated
works--an area that is far from settled.\34\ The evolution of
understanding who holds the rights to computer-generated works with
regard to agricultural data played out recently in the discussions
surrounding comments by Deere & Company on proposed exemptions to the
Digital Millennium Copyright Act \35\ regarding copyright protection
systems in vehicle software.\36\
---------------------------------------------------------------------------
\34\ See generally M. Leaffer, Understanding Copyright Law, 109-110
(5th ed. 2011).
\35\ 17 U.S.C. Sec. Sec. 512, 1201-1205, 1301-1332; 28 U.S.C.
Sec. 4001
\36\ See Deere & Company, ``Long Comment Regarding a Proposed
Exemption Under 17 U.S.C. 1201'' (2015). Available at http://
copyright.gov/1201/2015/comments-032715/class%2022/
John_Deere_Class22_1201_2014.pdf (last visited November 8, 2017).
Compare K. Weins, Wired (Business Blog Section, online edition)
(editorial) ``We Can't Let John Deere Destroy the Very Idea of
Ownership,'' April 21, 2015. http://www.wired.com/2015/04/dmca-
ownership-john-deere/ (last visited November 8, 2017).
---------------------------------------------------------------------------
3.3 Application of trade secret law to agricultural data
While trademark, patent, and copyright do not appear to fit as
models for farm data ownership, trade secret has the potential to serve
the agriculture industry's concerns regarding rights in data shared
with Big Data service providers. Importantly, trade secret is a
function of state law (unlike trademark, patent, and copyright, which
are all creatures of Federal law). At the time of this testimony, all
but three states have adopted the Uniform Trade Secrets Act, providing
a degree of consistency in trade secret law across most states.
Under the Uniform Trade Secrets Act (``UTSA''), a ``trade secret''
is defined as:
information, including a formula, pattern, compilation,
program, device, method, technique, or process, that:
(i) derives independent economic value, actual or potential,
from not being generally known to, and not being readily
ascertainable by proper means by, other persons who can obtain
economic value from its disclosure or use, and
(ii) is the subject of efforts that are reasonable under the
circumstances to maintain its secrecy.
Importantly, this definition makes clear ``information . . .
pattern[s], [and] compilation[s]'' can be protected as trade secret.
This, at last, affords hope of a protective model for farm data. This
is not to say that trade secret is a perfect model for protecting farm
data, however. Note the two additional requirements of trade secret:
first, the information has actual or potential economic value from not
being known to other parties, and second, it is the subject of
reasonable efforts to maintain the secret.
The first provision requires that to be protected as a trade
secret, farm data such as planting rates, harvest yields, or outlines
of fields and machinery paths must have economic value because such
information is not generally known. While a farmer may (or may not)
have a privacy interest in this information, the question remains as to
whether the economic value of that information derives, at least in
part, from being a secret. The counterargument to that point is the
economic value of the information comes from the farmer's analysis of
that information and the application of that analysis to his or her own
operation--a value completely independent of what anyone else does with
the information--and that the information for that farm, standing
alone, has no economic value to anyone else since that information is
useless to anyone not farming that particular farm.\37\ One can see
this first element poses problems for the trade secret model. It should
be noted here there is a clear economic benefit to the collection of
farm data; otherwise companies would not be investing billions of
dollars to position themselves in the agricultural data industry.\38\
This represents a question yet to be answered clearly by the body of
trade secret law: whether one can have trade secret protection in
information that standing alone has no economic value to other parties,
but does have such value when aggregated with similar data from other
parties.
---------------------------------------------------------------------------
\37\ An agricultural producer could, hypothetically, use such data
to bid rented agricultural land away from another tenant if they could
somehow demonstrate they could provide the landowner with evidence they
could increase the landowner's returns. However, this seems a tenuous
argument for the economic value element of the UTSA test and has no
application at all in a scenario with owned agricultural land.
\38\ See B. Upbin, Forbes (Tech business blog), ``Monsanto Buys
Climate Corp for $930 Million,'' October 2, 2013. http://
www.forbes.com/sites/bruceupbin/2013/10/02/monsanto-buys-climate-corp-
for-930-million/.
---------------------------------------------------------------------------
The second provision--the data be subject to reasonable efforts to
maintain its secrecy--also finds problems in an environment where the
data is continuously uploaded to another party without the intervention
of the disclosing party. The fact that data are disclosed to another
party does not mean it cannot be protected as a trade secret; if that
were the case, there would be little need for much of trade secret law.
Rather, the question is how and to whom the information is disclosed.
As noted in the Restatement (Third) of Unfair Competition's comments on
the Uniform Trade Secret Act, ``. . . the owner is not required to go
to extraordinary lengths to maintain secrecy; all that is needed is
that he or she takes reasonable steps to ensure that the information
does not become generally known.'' \39\
---------------------------------------------------------------------------
\39\ Smith, supra note 4, citing Restatement of Unfair Competition
(Third) Sec. 757 (1995).
---------------------------------------------------------------------------
The question becomes what constitutes ``reasonable steps'' to keep
continuously uploaded data protected, or data that is voluntarily
shared with a Big Data ATP. Almost certainly this means there must be
some form of agreement in place between the disclosing party and the
receiving party regarding how the receiving party must treat the
received information, including to whom (if anyone) the receiving party
may disclose that information. Such agreements are discussed in greater
detail below. However, there is some question as to whether any
agreement could protect the trade secret claim for data that was
disclosed to an ATP. When one discusses farm data privacy, one often
consider the concept of remaining anonymous. However, in the Big Data
world anonymity is no longer achievable, at least in the same manner as
it once was. Mayer-Schonberger and Cukier describe how even sanitized
data can reveal the identity of individuals by combining additional
layers of (probably publicly available) data. Given the prevalence of
public geospatial data, data from USDA, and plat maps, it is possible
in many circumstances to use those data layers with a sanitized
community of farm data to reveal all the data that were intended to
remain anonymous. As a result, one could argue sharing data with an
aggregator essentially renders it ineligible as a trade secret
(regardless of a non-disclosure agreement with the aggregator) since
the receiver cannot make a reasonable guarantee that the data can be
kept secret.\40\ This concept has implications not only for the
potential application of trade secret principles to agricultural data,
but to broader privacy policy concerns as well.
---------------------------------------------------------------------------
\40\ Griffin and Shanoyan, supra note 20.
---------------------------------------------------------------------------
Assuming for the moment that trade secret protection can be
obtained for agricultural data, one should consider the use of a ``non-
disclosure agreement'' when sharing data with an ATP. While an explicit
written ``non-disclosure agreement'' (or ``NDA'') is not necessary to
claim trade secret protection, such an agreement is almost certainly a
good idea if an agricultural producer wishes to retain a protectable
ownership interest in their data if such an interest exists. Not only
can such an agreement clarify a number of issues unique to the
relationship between the disclosing and receiving parties, but also can
address numerous novel issues in the current information environment
that trade secret law have not yet reached.
The concept of NDAs as separate agreements may be practicable for
one-on-one relationships, such as those between agricultural producers
and smaller consulting firms, negotiating separate agreements with
multiple entities poses significant transaction costs. This problem is
particularly magnified when one considers larger corporate service
providers who would face the issue of negotiating tens of thousands of
NDAs. Unsurprisingly, such entities choose to create standard
agreements in their form contracts. While certainly understandable,
this in turn creates the ``opt-out problem'' wherein a farmer who
believes the form contract does not adequately protect his or her
interests is forced to either agree to the form or do without the
product or service--which may be the only product or service compatible
with a significant portion of the very expensive equipment he or she
already owns or uses. This then provokes the discussion of whether such
contracts are enforceable or are, instead, adhesion contracts. There is
yet to be found consistency among Federal courts as to the
enforceability of such software use agreements.\41\
---------------------------------------------------------------------------
\41\ The asymmetry of EULA's has led to allegations they represent
``adhesion contracts'' and should not be enforceable as a matter of
policy. However, some courts have found insufficient evidence of
adhesion and held such agreements enforceable. Compare cases finding
EULAs enforceable: Ariz. Cartridge Remanufacturers Ass'n v. Lexmark
Int'l, Inc., 421 F.3d 981 (9th Cir., 2005); ProCD, Inc. v. Zeidenberg,
86 F.3d 1447 (7th Cir. 1996); Microsoft v. Harmony Computers, 846 F.
Supp. 208 (E.D.N.Y. 1994); Novell v. Network Trade Center, 25 F. Supp.
2d. 1218 (D. Utah, 1997) with cases finding EULAs unenforceable: Step-
Saver Data Systems Inc. v. Wyse Technology, 939 F.2d 91 (3rd Cir.
1991); Vault Corp. v. Quaid Software Ltd. 847 F.2d 255 (5th Cir. 1988);
Klocek v. Gateway, Inc., 104 F. Supp. 2d 1332 (D. Kan. 2000).
---------------------------------------------------------------------------
To conclude the trade secret analysis, colorable arguments exist
both for and against the proposition farm data poses an ``ownable'' and
protectable trade secret. That said, this option provides the best
doctrinal fit among the traditional intellectual property forms, and
farmers wishing to preserve whatever rights they do indeed have in that
data seem best advised to use the trade secret model to inform the
their protective measures. Even so, use of trade secret doctrine as a
protective measure for agricultural data has drawbacks in the lack of
consistency among states in trade secret law (although the UTSA has
done much to add consistency to the field) and the fact it is often a
``backward looking'' and costly solution since trade secret must
frequently be used to seek damages (which are often difficult to both
prove and quantify) through litigation after a disclosure has already
been made.
4. Current Legal Framework for Privacy Rights in Agricultural Data
Those concerned about the disclosure of personal data can certainly
cite a number of damaging data breach examples. Recent history suggests
many of the real threats in data transfers come from insufficient
controls to prevent the disclosure of personally identifiable
information (``PII'') to outside parties and inadequate agreements on
the uses of data by parties to whom it is disclosed.
To the extent producers regard agricultural data as proprietary,
their concerns about its disclosure naturally invite a review of the
release or theft of proprietary information in other sectors. One need
not look far into the past to find numerous examples of the disclosure
of PII, whether merely inadvertent or the result of targeted hacks.
Attacks on companies' payment systems have resulted in the credit card
information of hundreds of millions of customers from Adobe Systems
(150 million customers), Heartland Payment Systems (130 million
customers), TJX (parent company of TJ Maxx and Marshalls, 94 million
customers), TRW Information Systems (credit reporting company, 90
million customers), Sony (70 million customers) each of which dwarf
breaches attracting more media attention such as Home Depot (56 million
customers) and Target (40 million customers).\42\ Perhaps the most
troubling data breach in recent history, though, was the 2017 Equifax
data breach, which exposed a large array of personal and financial data
for over 143 million.\43\ The Equifax breach is especially troubling
for many consumers, as Equifax was entrusted with the most sensitive
personal information consumers could provide, and was supposed to serve
as a secure repository for that information. It is reasonable to
surmise that particular breach was a significant setback for the trust
of agricultural producers in systems that could collect their financial
data.
---------------------------------------------------------------------------
\42\ J. Pepitone, ``5 of the Biggest-ever Credit Card Hacks,''
(2013) CNN Money, available at http://money.cnn.com/gallery/technology/
security/2013/12/19/biggest-credit-card-hacks/ (last visited November
8, 2017).
\43\ Federal Trade Commission, ``The Equifax Data Breach: What to
Do.'' https://www.consumer.ftc.gov/blog/2017/09/equifax-data-breach-
what-do (last visited November 8, 2017).
---------------------------------------------------------------------------
To some extent, there may be a very limited reasonable
``expectation of privacy'' in agricultural data since a significant
segment of such data is available from public sources or sources
obtainable from public vantage points (such as aerial or satellite
imagery). Nevertheless, there remains an also-significant segment of
data for which an argument could be made that a privacy interest
exists. The challenge may be figuring out who has the best ability to
protect that data from disclosure.
The greatest risk of data breaches for agricultural producers may
be attacks against aggregators, since attacks against individual farm
systems pose very high barriers relative to the amount of data such an
attack could obtain. Theoretically, a hacker could tap into the
tractor/implement network (also called the tractor/implement bus) using
a number of commercially-available technologies allow farmers to plug
into the network and access Controller Area Network (``CAN'') messages
directly; for example, one could purchase a CAN message reader (``CAN
sniffer'') to read machine diagnostic codes for repairs.\44\ Someone
wishing to ``steal'' data would likely not want to be present to
retrieve the data from the device, though, and would likely prefer to
use a CAN data logger coupled with a device to wirelessly transmit the
data. Many data loggers are available to the public as well; for
example, the ``Snapshot'' device used by Progressive Insurance for
some insurance programs is simply a CAN data logger plugged into a
vehicle's On-Board Diagnostic (OBD-II) port.\45\ Alternatively, of
course, if one wanted to steal large amounts of agricultural data at
once, one could attempt to hack a cellular network provider used by an
equipment manufacturer to carry their data signals. Further, it should
be noted the equipment manufacturer likely has no ability to specify or
enforce the security protocols used to safeguard such cellular
transmissions.
---------------------------------------------------------------------------
\44\ Interview with Dr. John Fulton, Ohio State University
Department of Food, Agricultural, and Biological Engineering, July 6,
2015.
\45\ See Progressive Corporation, ``Snapshot Terms and
Conditions,'' https://www.pro
gressive.com/auto/snapshot-terms-conditions/ (last visited November 8,
2017).
---------------------------------------------------------------------------
While such an approach would work for standard messages transmitted
over the bus, it would not work for proprietary messages. To decode
such messages, the prospective hacker would have to develop a system
for decoding the information being provided from the task controller
for the implement, and that task would take almost as much work (if not
more) than the work in developing the task controller system in the
first place.\46\ Note, that several companies now provide means for
reverse-engineering proprietary CAN messages (such as those related to
crop yield) so farmers can automatically transfer yield data to the
cloud. Such technology could also be used to decode other proprietary
information.\47\ Perhaps ironically, the growth of proprietary data
network protocols that lead to complaints about the lack of
interoperability of farm equipment systems could also provide greater
protection against data breaches.
---------------------------------------------------------------------------
\46\ See interview with Dr. Marvin Stone (June 10, 2015).
\47\ Interview with Dr. John Fulton, Ohio State University
Department of Food, Agricultural, and Biological Engineering, July 6,
2015.
---------------------------------------------------------------------------
Additionally, the Global Positioning System ``GPS'' receiver in
most systems connects directly to the implement's task controller. As a
result, a ``bug'' might receive information about the commands sent to
the implement but without the associated location data, rendering it
meaningless. The bug would require its own GPS receiver along with
implement data (the configuration and dimensions of the implement),
which today could be done for a modest equipment cost.\48\ Obtaining
agronomic data via a physical connection to an implement poses a task
manageable for someone knowledgeable in SAE J1939 and ISO 11783 \49\
technology.\50\ However, building and deploying such a device poses a
significant amount of effort (to say nothing of the potentially-
criminal trespass involved in deploying it) in relation to the prospect
of collecting data on only one farm.
---------------------------------------------------------------------------
\48\ A relatively quick search of Google will yield many GPS
receiver units for less than $50.
\49\ SAE International, ``The SAE J1939 Communications Network: An
Overview of the J 1939 Family of Standards and How they are Used,'' 5
(white paper), available at http://www.sae.org/misc/pdfs/J1939.pdf
(last visited November 8, 2017). See also International Organization
for Standardization, ISO Draft International Standard ISO/DIS 11783:
Tractors and machinery for agriculture and forestry--serial control and
communications data network (2012). The ISO 11783 standard is often
referred to as the ``ISOBUS standard'' and defines how the on-board
computer networks on most agricultural equipment works and how their
individual components work together. Combined, SAE J1939 and ISO 11783
govern much of how the data-collection network on any agricultural
equipment works.
\50\ M. Miettien, ``Implementation of ISO 11783 Compatible Task
Controller,'' XVI CIGR (International Commission of Agricultural and
Biosystems Engineering) World Congress, Bonn, Germany (2006), available
at http://users.aalto.fi/ttoksane/pub/2006_CIGR20062.pdf (last visited
November 8, 2017).
---------------------------------------------------------------------------
As illustrated from this discussion, a number of factors in the
configuration and operation of farm data networks limit the
opportunities for hackers to take agricultural data directly from the
agricultural producer. Admittedly, most producers put little thought
into their systems being physically hacked but worry instead about
their data being accessed through an intercepted cellular signal. They
might also worry about a bad actor hacking the system to implant false
data. First, virtually all cellular signals are encrypted when
transmitted and decrypted at the cellular tower;\51\ without the
decryption key, interpreting any data transmitted would be difficult
(although not impossible for a sophisticated hacker; recent news has
highlighted the ability of some groups to do so \52\). The use of data
encryption through a secure sockets layer (``SSL'') protocol by the
farmer and his or her service provider in data transfers adds another
difficult-to-break security barrier to interception of the data.\53\
---------------------------------------------------------------------------
\51\ For a primer on the process of encoding and decoding cellular
signals, see How Stuff Works, ``How Cell Phones Work,'' http://
electronics.howstuffworks.com/cell-phone.htm (last visited November 8,
2017).
\52\ See C. Timberg & A. Soltani, By Cracking Cellphone Code, NSA
Has Ability to Decode Private Conversations, The Washington Post,
December 13, 2013. Online edition, available at http://
www.washingtonpost.com/business/technology/by-cracking-cellphone-code-
nsa-has-capacity-for-decoding-private-conversations/2013/12/13/
e119b598-612f-11e3-bf45-61f69f54fc5f_story.html (last visited November
8, 2017).
\53\ See C. Heinrich, Secure Socket Layer (SSL), in Encylopedia of
Cryptography and Security 1135 (Henck C.A. van Tilborg, Sushil Jajodia,
eds., 2011)
---------------------------------------------------------------------------
Most agricultural data disclosed to a service provider is likely in
the form of telematics data, raw data regarding crop production, GIS
information about the farm, and other similar types. Conversely,
hackers frequently go after large concentrations of data with easily-
converted financial value, such as credit card information. Thus, it
may be difficult for hackers to make a ``quick buck'' from agricultural
data making it a less-appealing target of attack. Nevertheless, an
adage in computer security is ``where there is value, there will be a
hacker.'' \54\ As a result, systems storing agricultural data are less
likely to be directly attacked, but farmers are understandably
concerned that PII may be stolen if, for example, their vendor account
information is somehow linked to their agricultural data or if their
account information is stored with a third party that is a more
appealing target. Depending on the type of computer at issue and its
common use, the Federal Computer Fraud and Abuse Act (``CFAA'') \55\
may provide a means of prosecuting unauthorized access of the computer
in the event agricultural data linked to PII is compromised. Discussed
below, the Federal Electronic Communications Privacy Act (ECPA) \56\
could also be used as a potential prosecutorial tool for those
attempting to intercept agricultural data during the data transmission
process.
---------------------------------------------------------------------------
\54\ S. Sammataro, ``Cybersecurity for Small or Regional Law
Firms,'' paper presented at American Agricultural Law Association
Annual Symposium, Charleston, South Carolina (October 23, 2015).
\55\ 18 U.S.C. Sec. Sec. 1030 et seq.
\56\ 18 U.S.C. Sec. Sec. 2510 et seq.
---------------------------------------------------------------------------
The theft of PII by criminals is one threat posed by data
transfers, but so too is the inadvertent, or perhaps intentional but
misinformed, disclosure of data by the party receiving that data. Take,
for example, the disclosure of thousands of farmers' and ranchers'
names, home addresses, GPS coordinates and personal contact
information'' by EPA in response to a Freedom of Information Act (FOIA)
request regarding concentrated animal feeding operations (CAFOs) which
prompted a lawsuit from the American Farm Bureau Federation and
National Pork Producers Council alleging the agency overstepped its
authority in doing so.\57\ While this event represents the disclosure
of information by an enforcement agency, many farmers fear the
converse--that an enforcement agency could compel a data-receiving
party to disclose information even if such disclosure were not legally
required. Another concern is whether an adverse party in litigation (or
even a party contemplating litigation) could persuade a party holding a
farmer's data to disclose the data as an aid to their case, again even
if such disclosure was not legally required.
---------------------------------------------------------------------------
\57\ S. Wyant, ``Farm Groups File Lawsuit to Stop EPA Release of
Farmers' Personal Data.'' Agri-Pulse (2013), available at http://
www.agri-pulse.com/Farm-groups-file-lawsuit-to-stop-EPA-release-of-
farmers-personal-data-07082013.asp (last visited November 8, 2017).
---------------------------------------------------------------------------
Much work remains to be done on defining governmental safeguards
against disclosures, and even more work remains to be done in defining
how the government can obtain electronic data. Although laws such as
the ECPA (heavily modified by the USA Patriot Act) govern the
acquisition of information through intercepted communications, there is
little law to prevent a government agency from simply requesting data
from a service provider. Anecdotal evidence suggests service providers
and their legal counsel continue to struggle in defining parameters for
how to respond to non-subpoenaed requests for data by government
agencies.
All these issues surround restrictions on the taking of information
by some unauthorized (or at least questionable) means. While there are
at least some laws potentially applicable in these circumstances, there
are no laws defining an inherent privacy right in agricultural
data.\58\ For example, the Federal Health Insurance Portability and
Accountability Act (``HIPAA'') \59\ provides privacy rights and
restrictions against disclosure of health information; the Gramm-Leach
Bliley Act (also known as the Financial Modernization Act of 1999) \60\
and Fair Credit Reporting Act \61\ protect financial information from
disclosure; the Privacy Act of 1974 \62\ restricts disclosures of
personal information by held by the Federal government. As of now,
though, there are large categories of agricultural data that may fall
between the cracks of these laws with no Federal (and in most cases, no
state) protections against its disclosure.
---------------------------------------------------------------------------
\58\ T. Janzen, ``Legal Issues Surrounding Farm Data Ownership,
Transfer, and Control,'' paper presented at American Agricultural Law
Association Annual Symposium, Charleston, South Carolina (October 23,
2015).
\59\ 42 U.S.C. Sec. 300gg, 29 U.S.C. Sec. Sec. 1181 et seq. and 42
U.S.C. Sec. Sec. 1320d et seq.
\60\ 15 U.S.C Sec. 6803.
\61\ 15 U.S.C. Sec. Sec. 1681 et seq.
\62\ 5 U.S.C. Sec. 552a.
---------------------------------------------------------------------------
5. Potential Policy Responses to Address Agricultural Data Issues
Having reviewed the current legal environment surrounding the
ownership rights and privacy protections relevant to agricultural data,
what can this Committee and Congress do to enable U.S. farmers and
ranchers to take maximum economic advantage of Big Data tools? As
referenced above, Big Data cannot be Big Data without ``buy-in'' to the
system from large numbers of agricultural producers. In these beginning
years of agricultural data systems, there are many ATPs vying for
farmers and their acreages to enroll in their systems. As the system
matures, this relationship will likely shift, and there will be few (or
perhaps only one) ATP and the vast majority of farms may be
participating. Nevertheless, for the maturation process to begin,
agricultural producers must ``buy in'' to the system. At a fundamental
level, that buy-in requires trust in the system from those producers.
That trust, in turn, likely requires answers to the questions of
ownership and privacy in agricultural data.
None of the Federal intellectual property laws directly address who
holds a protectable intellectual property right in agricultural data.
Arguably, the most appropriate fit may be found in state law under the
UTSA, although the applicability of that law is questionable as well.
The UTSA may provide a useful map to any Congressional efforts to help
define ownership rights in agricultural data. Passage of statutory law
defining ownership of ``agricultural data'' may be a daunting task
given the complexity of the current Federal and state intellectual
property framework (which also draws from centuries of common law).
Thus, it may be advisable instead to use a consensus-driven approach
among agricultural producers and service providers to define
agricultural data rights. The coalition led by the American Farm Bureau
Federation and its ``Privacy and Security Principles for Farm Data''
\63\ represents a tremendous step forward on this issue. Other groups,
such as the Open Ag Data Alliance, continue to build coalitions on the
technical side of the Big Data issue to develop systems and standards
embodying the principles of interoperability, security and privacy.\64\
The next step is to see continued cooperation among groups such as
these in integrating their principles in legally-binding service
agreements.
---------------------------------------------------------------------------
\63\ American Farm Bureau Federation, ``Privacy and Security
Principles for Farm Data,'' November 13, 2014 (revised April 1, 2016).
Available at https://www.fb.org/issues/technology/data-privacy/privacy-
and-security-principles-for-farm-data (last visited November 8, 2017).
\64\ Open Ag Data Alliance, ``Principals and Use Cases,'' http://
openag.io/about-us/principals-use-cases/ (last visited November 8,
2017).
---------------------------------------------------------------------------
Another collaborative effort to help agricultural producers
evaluate the data policies and protections of data service providers
has been the Ag Data Transparency Evaluator, coordinated by the
American Farm Bureau Federation, which requires service providers to
undergo a ten-factor review (based in part on the Privacy and Security
Principles, with the review self-reported by the service provider) with
a satisfactory review resulting in the ``Ag Data Transparent''
seal.\65\ Congressional support of this and other efforts to equip
farmers and ranchers in evaluating the data tools available can help
foster trust, encourage Big Data participation, and drive many of the
potential advantages Big Data services have to offer.
---------------------------------------------------------------------------
\65\ See www.agdatatransparent.com (last visited November 8, 2017).
---------------------------------------------------------------------------
Modern agricultural producers are expected to be proficient in a
broad array of the disciplines of science and business, but few have a
background in intellectual property law. Support of educational
programs to help these producers understand the legal issues at play in
Big Data service agreements could do much to help increase trust,
advance the consensus process, and empower producers to make informed
decisions about the cost-benefit analysis of sharing their data under
those service agreements. The consensus process may also provide a
vehicle for developing an understanding among all stakeholders as to
the privacy protections necessary and appropriate to protect
agricultural data, which occupies a unique space between purely
personal and business information. Such information does not readily
fit into the existing framework of Federal privacy laws, and as
business information, may not belong in such a framework.
One matter in which Congressional action may be directly applied is
the development of clearer guidelines regarding the production of
agricultural data held by private data aggregators, more robust
safeguards against inadvertent disclosure or intentional hacking by
outside parties, and clear guidance on when disclosure of government-
held data is, and is not, required under the Freedom of Information Act
\66\ or other circumstances.
---------------------------------------------------------------------------
\66\ 5 U.S.C. Sec. 552.
---------------------------------------------------------------------------
Finally, although outside the direct scope of a discussion of legal
issues in agricultural use of agricultural data tools, rural access to
wireless broadband services is crucial to fully utilizing the potential
of agricultural data systems. Before the rapid adoption and usage of
agricultural data technologies will occur, the lack of this enabling
technology must be addressed. The expansion of connectivity across the
U.S. has been a priority, but access has grown slowly. This is
especially true in the major crop producing regions. The majority of
data transfer occurs over cellular systems, but there are worldwide
initiatives to provide wireless connectivity via satellite, balloons,
and other platforms. Regardless of platform, the agricultural industry
relies upon wireless connectivity to support big data systems.
Telematics allows data to be wirelessly uploaded and downloaded
between farm machinery and online servers. However, limited
connectivity is a barrier to adoption leading to potential economic
losses.\67\ Whitacre et al. addressed the current connectedness of
agricultural production areas.\68\ It was these areas that were
impacted by the United States Federal Communications Commission (FCC)
updated definition of connectivity that could be considered broadband
in January 2015. The definition changed from 4 Megabits per second
(Mbps) download and 1 Mbps upload to 25 Mbps download and 3 Mbps
upload. Although broadband speeds did not instantly change, the level
of connectivity that service providers could advertise as `broadband'
changed. The faster speeds required to be considered broadband brought
light to connectivity barriers, especially with respect to connectivity
gaps in rural areas where agricultural production occurs. Specifically,
the 25 Mbps download speed requirement negates the majority of United
States wireless connections from being classified as broadband.
---------------------------------------------------------------------------
\67\ Griffin, T.W., and Mark, T.B. (2014). ``Value of Connectivity
in Rural Areas: Case of Precision Agriculture Data.'' International
Conference on Precision Agriculture. July 20-23, 2014. Sacramento, CA.
\68\ Whitacre, B.E., Mark, T.B., and Griffin, T.W. (2014). How
Connected are Our Farms? Choices. Online: http://
www.choicesmagazine.org/choices-magazine/submitted-articles/how-
connected-are-our-farms.
---------------------------------------------------------------------------
However, the vast majority of data being passed between farm
equipment and online servers is uploaded rather than downloaded; and
upload speeds are typically only a fraction of download speeds. For
some types of data such as machine diagnostics and prescriptions,
current speeds may be adequate. However, yield data and specifically
imagery data may require connectivity speeds in excess of what is
currently available. In summary, a concerted national policy effort
must be made to expand broadband access in rural areas for a number of
important rural development purposes, not the least of which is to
facilitate the potential economic advantages to be gained by
integration of agricultural data technologies on farms and ranches.
Concluding Remarks
The application of Big Data to agricultural production holds the
potential to improve the profitability of U.S. agriculture and to
better prepare its farmers and ranchers to handle the inherent risks of
the industry. Additionally, Big Data could play a vital role in the
further development of tools and techniques necessary to feed an ever-
growing, hungry world. I commend this Subcommittee for its foresight in
addressing these issues, and sincerely thank the Subcommittee, Chairman
Moran, and Ranking Member Blumenthal for the opportunity to address you
today.
Senator Moran. Thank you very much, Doctor.
Mr. Janzen, welcome.
STATEMENT OF TODD J. JANZEN, PRESIDENT,
JANZEN AGRICULTURAL LAW LLC
Mr. Janzen. Thank you, Mr. Chairman, Ranking Member
Blumenthal, and members of the Subcommittee. My name is Todd
Janzen. I'm an attorney, a private practice attorney, in
Indianapolis, Indiana. And the firm that I work at, Janzen
Agricultural Law LLC, specializes in helping farmers, ag
technology providers, and also agribusinesses.
But I speak here today not only as an attorney, but also as
somebody who grew up on a farm in south-central Kansas, a grain
and livestock farm. And so I'm particularly attuned to the
issues that farmers are facing on a legal front.
Let me start, though, by talking about the types of data
that are being collected by the various ag data platforms that
are out there because I think that's helpful in understanding
the framework here.
So there are various streams of data I like to say that
come off of fields and farms, and this can include land data,
agronomic data, weather data, management data, machine data,
and also livestock data, which would be information about
genetics or feed consumption by animals.
But what's really changed in the last 5 years is not just
that farmers are generating this information, but now farmers
are taking this information and storing it in cloud-based
servers that are not located on the farm. And so this transfer
of information off the farm I think is a pretty monumental
transition in history for U.S. agriculture.
But with this transfer of information, there have also been
some groups that have started to raise concerns, you know,
about farmers losing out on the benefit of knowing all this and
having all this information on their farm.
A poll by American Farm Bureau in 2016 identified a number
of issues that farmers faced, but I classify them into three
big categories. One is a lack of trust with a lot of these ag
technology providers that are on the market today. Second would
be a loss of control to these companies. And then the third
would be something that some other panelists have already
mentioned, which is just the overall complexity of the
agreements that farmers are being asked to sign. I heard a
farmer yesterday describe this as the aggressive fine print.
So to address these problems, American Farm Bureau,
National Farmers Union, and the commodity groups from a number
of national organizations came together and said, ``Let's
create a set of core principles that we can all agree on.'' So
in 2014, they introduced a document called the ``Privacy and
Security Principles for Farm Data,'' which really outlined some
core principles that they wanted to see adopted by the
industry.
And 37 different companies and organizations signed onto
these core principles by pledging that they would incorporate
them into their contracts, but here we are today, and we don't
yet have all 37 of those who have really met the challenge that
they agreed to take on years ago.
So as a follow up to this effort, American Farm Bureau,
Farmers Union, and others came together and said, ``Let's
create some sort of certification process where we can
recognize companies that are adhering to these core principles
for agricultural data.'' And so out of that came this Ag Data
Transparency Evaluator process, which I've been fortunate
enough to be a part of. And I want to just briefly describe how
that works so that it's clear to the Subcommittee.
Companies that want to be certified as ``Ag Data
Transparent'' can submit their contracts and answer a 10-
question form about how they use farmers' ag data. That is then
reviewed by an independent third-party administrator, and
currently Janzen Agricultural Law is that administrator. So we
take that information and we review it, and we ultimately
determine if companies are being transparent with how they are
using agricultural data or if they're not being transparent, in
which case, we send them back and ask them to do it over.
But if they are transparent, then they are awarded the ``Ag
Data Transparent'' seal of approval. And you may see this on
some companies' marketing materials. And I'll mention
Farmobile, one of the very first companies to go through this
process and has been awarded the ``Ag Data Transparent'' seal
of approval.
So we review these questions, such as, ``What data is being
collected?'' Does the company obtain consent before the data is
transferred in or out of that platform? And then also, ``Can a
farmer delete their data if they're finished using that
platform?'' And then we post the results of this question-and-
answer to the AgData
Transparent.com website so that farmers can review this and
make informed decisions before they go down the road of sending
their data to one of these companies.
So here we are today. We've had nine companies go through
this evaluation process and be certified ``Ag Data
Transparent.'' There is still a lot of work to do because there
are a lot of companies that have said they want to do this
process or maybe they've committed to it, but yet they haven't
followed through and become certified.
So, Mr. Chairman and members of the Subcommittee, I'm
honored to be here today, and I look forward to your questions.
I think this is a very important issue for farmers, and I hope
that we can address them.
Thank you.
[The prepared statement of Mr. Janzen follows:]
Prepared Statement of Todd J. Janzen, President,
Janzen Agricultural Law LLC
Good afternoon Chairman Moran, Ranking Member Blumenthal, and
members of the Subcommittee. My name is Todd J. Janzen, I am the
president and attorney with Janzen Agricultural Law, LLC, a law firm
based in Indianapolis, Indiana that serves the needs of America's
farmers, ag technology providers, and agribusinesses.
One of the reasons we founded Janzen Ag Law in 2015 was that we
wanted to be at the forefront of the changes that have been occurring
on the farm for the past few years. Farms are becoming more digital
every day, and together with that digitalization is a movement of
agricultural data stored on computers in the farm office to cloud-based
data storage devices. Agricultural data (ag data) can be many things,
including yield data, soil data, planting information, weather data,
financial data, etc. This marks the first time in history that the
majority of the information that farmers generate and use on their
farms has been moved into the hands of companies outside the farm.
As a result, we are seeing a digital land-rush occurring across the
United States. The past few years have seen millions of dollars pour
into ag data startups from Silicon Valley, to Kansas City, to North
Carolina. Historic legacy agricultural companies, such as John Deere,
are also at the forefront of this movement by expanding their product
offerings to include cloud-based data storage platforms. All of these
companies are scrambling to get the most acres of data into their
platforms so that when consolidation of ag technology providers (ATPs)
begins, they are in the strongest position.
In the race to the cloud, we must also be cautious so that the
American farmer is not left behind. Today I will address the issues
facing farmers as digitalization occurs and how the industry has begun
to address these issues.
Issues Facing Farmers as Ag Data Moves into the Cloud
American Farm Bureau Federation (Farm Bureau) conducted a poll of
over 400 farmers in 2016 to understand their issues concerning ag data
privacy, security, and control. The poll highlighted what are
essentially three issues that continue to come up when asking farmers
about ag data concerns:
1. Lack of Trust
Seventy-seven percent (77 percent) of farmers expressed concern
about which entities can access their farm data after the data is
uploaded to cloud-based servers. The same percentage expressed concern
about whether uploading the data could cause it to be used for
regulatory purposes.
Sixty-seven percent (67 percent) of farmers said they consider how
outside parties will use their ag data when deciding whether to entrust
their data with a certain ATP.
A farmer's lack of trust can come from many sources, but I
speculate it originates in two places. Many ag data companies are new.
Ag data startups lack the goodwill that older agricultural companies
have spent years building. They have new sales associates who are
strangers to the farm, or in some instances, strangers to agriculture.
They are viewed as outsiders.
Older, long-established agricultural companies do not suffer from a
general lack of trust with the farmer, since they have spent years
building that relationship. But when a seed company, equipment
manufacturer, or ag retailer begins offering an ag data platform to
store the farmer's ag data, farmers often are skeptical about whether
the storage provider is trying to help the farmer raise a better crop
or using the ag data to sell the farmer more or higher-priced goods and
services. This skepticism may erode a farmer's trust.
2. Concern with Losing Control
Farmers are also concerned that uploading their ag data to cloud-
based platforms means they will lose control over downstream uses.
Sixty-six percent (66 percent) of respondents in the Farm Bureau poll
believe farmers should share in the potential financial benefits from
the use of their data beyond the direct value they may realize on their
farm.
Farmers raised concerns that ATPs could use their ag data to gain
an unfair advantage in the marketplace. Sixty-one percent (61 percent)
of farmers expressed worry that ATPs could use their data to influence
market decisions.
These concerns arise from a fundamental legal truth about ag data--
there are no laws that specifically protect farmers' privacy and
security concerns. Ag data is not typically ``personally identifiable
information,'' such that it would be protected by state laws which
prevent misuse of personal information like name, address, and phone
number. Nor does ag data fit into a class of data that Congress has
chosen to protect legally, such as medical information (HIPAA).
Finally, ag data does not neatly fit into existing legal protections
for intellectual property, such as patents, trademarks, or copyrights.
Ag data ultimately may be deemed a trade secret under existing state
and Federal trade secret laws, but that will depend upon whether courts
interpret existing statutes to include information such as agronomic
data.
These uncertainties mean that the contracts between farmers and ag
tech providers are very important. These contracts will determine
farmers' rights in the ag data their farms create.
3. Frustration with Complexity of Current Legal Agreements
Fifty-nine (59 percent) percent of farmers were confused about
whether current legal agreements allowed ATPs to use their ag data to
market other services, equipment, or inputs back to them. Zippy Duvall,
president of Farm Bureau, said: ``This indicates a higher level of
clarity and transparency is needed to secure grower confidence. One of
the topics I hear most about from farmers on the data issue is having a
clear understanding about the details of `Terms and Conditions' and
`Privacy Policy' documents we all sign when buying new electronics. You
should not have to hire an attorney before you are comfortable signing
a contract with an ag technology provider.''
Our experience as a law firm working in this area confirms that
this is a real problem for farmers and ATPs. There is no standard
agreement that governs ag data transfer, use, and control by ATPs.
Instead, technology companies have adapted other forms of legal
agreements to try to address the issues associated with moving ag data
into cloud-based platforms, but with limited success. A farmer seeking
to compare two similar products today might find that they are governed
by two very different sets of contracts.
This only adds to a farmer's confusion. If we want to make
technology easy to embrace and use--and we do--then we need to simplify
the contracts farmers sign when implementing new ag data technology on
the farm. Contracts that no one reads and understands set the stage for
problems down the road.
How the Ag Industry is Addressing Farmers' Concerns
1. The Privacy and Security Principles for Farm Data
Farm Bureau, National Farmer's Union, and national commodity
organizations for corn, soybeans, wheat, and sorghum, led an effort in
2014 to establish fundamental principles for companies working in the
ag data space. These organizations held a series of meetings where
roundtable discussions occurred among industry stakeholders, such as
John Deere, CNH Industrial, AGCO, Monsanto, DuPont Pioneer, Beck's
Hybrids, Dow Agrosciences, Farmobile, and other ag technology
providers. The culmination of these efforts was the drafting of the
``Privacy and Security Principles for Farm Data,'' also known today as
ag data's ``Core Principles.''
The Core Principles address thirteen key elements related to ag
data. These include:
Education
Ownership
Collection, Access and Control
Notice
Transparency and Consistency
Choice
Portability
Terms and Definitions
Disclosure, Use, and Sale Limitation
Data Retention and Availability
Contract Termination
Unlawful or Anti-Competitive Activities
Liability & Security Safeguards
After releasing the Core Principles in 2014, Farm Bureau asked
companies to voluntarily ``sign on'' to the document. As of October
2017, the following organizations and companies have agreed to
implement the Core Principles into their contracts with farmers.
AGCO
Ag Connections, Inc.
Agrible, Inc.*
AgSense
AgWorks
Ag Leader Technology
American Farm Bureau Fed.
American Soybean Assoc.
Beck's Hybrids*
CNH Industrial
Conservis*
Crop IMS
CropMetrics
Dow AgroSciences LLC
DuPont Pioneer
Farm Dog
Farmobile LLC*
Granular*
Grower Information Services Cooperative
GROWMARK, Inc.*
Independent Data Management LLC*
ohn Deere
Mapshots, Inc.
National Assoc. of Wheat Growers
National Barley Growers Assoc.
National Corn Growers Assoc.
National Cotton Council
National Farmers Union
National Potato Council
National Sorghum Producers
North American Equipment Dealers Assoc.
OnFarm
Raven Industries
Reinke Manufacturing Co., Inc.
Syngenta
The Climate Corporation--a division of Monsanto
USA Rice Federation
Valley Irrigation
ZedX Inc.
*Company certified to be Ag Data Transparent. For more information,
visit www.agdatatransparent.com
A copy of the Core Principles is attached as Exhibit A.
2. The Ag Data Transparent Effort
Having the Core Principles in place was a great starting point for
the ag data industry to address farmers' concerns with ag data privacy,
use, and control. However, the Core Principles are only guidelines, and
only valuable if companies incorporate the Core Principles into their
contracts with farmers. Therefore, following the release of the Core
Principles, several farm groups and industry stakeholders worked
together to create an independent verification tool that could help
farmers determine if ag tech providers are abiding by the Core
Principles. This tool is called the Ag Data Transparency Evaluator. It
is a simple three-step process:
Participating companies must answer 10 questions about how
they store, use, and transfer ag data.
The 10 question answer form is reviewed by an independent
third party for transparency and completeness.
If the evaluation is acceptable, the company is awarded the
``Ag Data Transparent'' seal of approval for use on its future
marketing materials.
Participation is voluntary, but all companies that signed onto the
Core Principles have been asked to participate in the Ag Data
Transparent effort as well.
a. The 10 Question Evaluation. Here is a list of the 10 questions
that each participant is asked to answer as part of the evaluation:
Answers to all questions except for question 1 are ``yes'' or
``no,'' but companies are also given space to explain their answer.
b. Reviewing the 10 Question Evaluation.
After an ag tech company completes the 10 question evaluation form,
the company submits its answers to an independent third party evaluator
to determine compliance. Janzen Agricultural Law LLC is the law firm
that has been selected to conduct the evaluations. After reviewing a
company's answers, we typically go back to that company with
suggestions for improving its contracts and policies to bring into
compliance with Core Principles. Companies then make those revisions to
their contracts and policies and resubmit their 10 question form. Once
a company's answers align with the Core Principles, we send an official
letter designating the company as ``Ag Data Transparent'' and
authorizing use of the seal of approval.
The final, approved 10 question answer forms are posted on the Ag
Data Transparent website at www.AgDataTransparent.com Farmers can
research and review companies' answers online. The website requires no
log in and is free to use. An example of the home page is attached as
Exhibit B.
c. The Ag Data Transparent Seal of Approval
Companies that undergo evaluation and are approved as ``Ag Data
Transparent'' may then use the seal of approval on their websites and
marketing materials. To date, nine companies have completed the
evaluation and been approved as ``Ag Data Transparent.'' These nine
companies are:
AgDNA
AgIntegrated, Inc.
Agrible, Inc.
Beck's Hybrids
Conservis Corporation
Farmobile
Granular
GROWMARK
Independent Data Management LLC
The participants are diverse, from a Silicon Valley ag tech
startup, to a Midwestern seed company, to one of the Nation's largest
farm cooperatives and ag retailers. These companies may use the Ag Data
Transparent seal on their websites, denoting their compliance with the
Core Principles. Farmers who see the seal of approval will know the
company went through the time and effort to certify its contract.
The Ag Data Transparent process addresses farmers' three main
concerns with ag data. First, the process instills trust. No company
submits its contracts to a voluntary evaluation unless the company is
willing to revise its contracts, as necessary, to bring them into
compliance with the Core Principles. Second, loss of control is
addressed by requiring tech providers to obtain farmer consent before
transferring data to third parties. Finally, farmers' complexity
frustration is addressed by condensing all of a tech provider's
contracts into a 10 question form that answers the questions farmers
want to know. The Ag Data Transparent process makes contracts better.
d. Who is behind the Ag Data Transparent effort?
The Ag Data Transparent effort is governed by a non-profit
corporation, the Ag Data Transparency Evaluator Inc. The corporate
bylaws create two classes of directors: (1) farm organizations that are
made up of farmer-member organizations; and (2) diverse ag technology
providers, referred to as ``industry partners.'' The farm organizations
are American Farm Bureau Federation, American Soybean Association,
National Corn Growers Association, National Farmers Union, National
Sorghum Producers, National Association of Wheat Growers and National
Potato Council. The industry partner board members are ag technology
providers ranging from large corporations, medium-sized companies, and
ag tech startup organizations.
Janzen Agricultural Law LLC, which serves as the administrator of
the program and conducts the evaluation reviews, is not a board member.
3. The Ag Data Use Policy
Our law firm also drafts terms of service, license agreements,
privacy policies, and other contracts for ag technology providers. This
work has confirmed many concerns facing farmers today when it comes to
ag data. We see how companies struggle to communicate clearly how they
intend to store, use, and transfer ag data.
For these reasons, we have encouraged companies to draft ``data use
policies'' or ``data use agreements'' for their farmers. In a typical
data use contract, the technology provider addresses all of the issues
raised by the 10 questions and the Core Principles. For example, a data
use policy will explain what information the provider collects and what
permission is required before the provider transfers that data to
another party.
From our standpoint, the Ag Data Transparent effort has helped
drive more technology providers into creating data use policies. Thus,
the effort has paid dividends even for some companies that have not
participated in evaluations because it has caused them to rethink how
they are contracting with farmers.
Conclusion
The Ag Data Transparent effort is great step towards bringing
transparency to ag data contracts between farmers and their technology
providers. Wider participation would certainly help the effort, but
that is up to the industry. Out of the dozens of ag tech providers with
cloud-based platforms on the market today, only nine have completed the
certification process. A few companies are in the process of
certifying, but uptake could be better.
Farmers should ask their technology providers why they have not
earned that Ag Data Transparent seal. This Subcommittee should ask
technology providers this question as well when they come before you to
testify in future hearings.
Thank you, Mr. Chairman and Ranking Member for your time and
attention to this important issue. I look forward to answering any
questions you may have for me.
______
Exhibit A
Privacy and Security Principles for Farm Data
(Ag Data's Core Principles)
November 2014
The recent evolution of precision agriculture and farm data is
providing farmers with tools, which can help to increase productivity
and profitability.
As that technology continues to evolve, the undersigned
organizations and companies believe the following data principles
should be adopted by each Agriculture Technology Provider (ATP).
It is imperative that an ATP's principles, policies and practices
be consistent with each company's contracts with farmers. The
undersigned organizations are committed to ongoing engagement and
dialogue regarding this rapidly developing technology.
Education: Grower education is valuable to ensure clarity between
all parties and stakeholders. Grower organizations and industry should
work to develop programs, which help to create educated customers who
understand their rights and responsibilities. ATPs should strive to
draft contracts using simple, easy to understand language.
Ownership: We believe farmers own information generated on their
farming operations. However, it is the responsibility of the farmer to
agree upon data use and sharing with the other stakeholders with an
economic interest, such as the tenant, landowner, cooperative, owner of
the precision agriculture system hardware, and/or ATP etc. The farmer
contracting with the ATP is responsible for ensuring that only the data
they own or have permission to use is included in the account with the
ATP.
Collection, Access and Control: An ATP's collection, access and use
of farm data should be granted only with the affirmative and explicit
consent of the farmer. This will be by contract agreements, whether
signed or digital.
Notice: Farmers must be notified that their data is being collected
and about how the farm data will be disclosed and used. This notice
must be provided in an easily located and readily accessible format.
Transparency and Consistency: ATPs shall notify farmers about the
purposes for which they collect and use farm data. They should provide
information about how farmers can contact the ATP with any inquiries or
complaints, the types of third parties to which they disclose the data
and the choices the ATP offers for limiting its use and disclosure.
An ATP's principles, policies and practices should be transparent
and fully consistent with the terms and conditions in their legal
contracts. An ATP will not change the customer's contract without his
or her agreement.
Choice: ATPs should explain the effects and abilities of a farmer's
decision to opt in, opt out or disable the availability of services and
features offered by the ATP. If multiple options are offered, farmers
should be able to choose some, all, or none of the options offered.
ATPs should provide farmers with a clear understanding of what services
and features may or may not be enabled when they make certain choices.
Portability: Within the context of the agreement and retention
policy, farmers should be able to retrieve their data for storage or
use in other systems, with the exception of the data that has been made
anonymous or aggregated and is no longer specifically identifiable.
Non-anonymized or non-aggregated data should be easy for farmers to
receive their data back at their discretion.
Terms and Definitions: Farmers should know with whom they are
contracting if the ATP contract involves sharing with third parties,
partners, business partners, ATP partners, or affiliates. ATPs should
clearly explain the following definitions in a consistent manner in all
of their respective agreements: (1) farm data; (2) third party; (3)
partner; (4) business partner; (5) ATP partners; (6) affiliate; (7)
data account holder; (8) original customer data. If these definitions
are not used, ATPs should define each alternative term in the contract
and privacy policy. ATPs should strive to use clear language for their
terms, conditions and agreements.
Disclosure, Use and Sale Limitation: An ATP will not sell and/or
disclose non-aggregated farm data to a third party without first
securing a legally binding commitment to be bound by the same terms and
conditions as the ATP has with the farmer. Farmers must be notified if
such a sale is going to take place and have the option to opt out or
have their data removed prior to that sale. An ATP will not share or
disclose original farm data with a third party in any manner that is
inconsistent with the contract with the farmer. If the agreement with
the third party is not the same as the agreement with the ATP, farmers
must be presented with the third party's terms for agreement or
rejection.
Data Retention and Availability: Each ATP should provide for the
removal, secure destruction and return of original farm data from the
farmer's account upon the request of the farmer or after a pre-agreed
period of time. The ATP should include a requirement that farmers have
access to the data that an ATP holds during that data retention period.
ATPs should document personally identifiable data retention and
availability policies and disposal procedures, and specify requirements
of data under policies and procedures.
Contract Termination: Farmers should be allowed to discontinue a
service or halt the collection of data at any time subject to
appropriate ongoing obligations. Procedures for termination of services
should be clearly defined in the contract.
Unlawful or Anti-Competitive Activities: ATPs should not use the
data for unlawful or anti-competitive activities, such as a prohibition
on the use of farm data by the ATP to speculate in commodity markets.
Liability & Security Safeguards: The ATP should clearly define
terms of liability. Farm data should be protected with reasonable
security safeguards against risks such as loss or unauthorized access,
destruction, use, modification or disclosure. Polices for notification
and response in the event of a breach should be established.
The undersigned organizations for the Privacy and Security
Principles of Farm Data as of April 1, 2016.
AGCO
Ag Connections, Inc.
Agrible, Inc.*
AgSense
AgWorks
Ag Leader Technology
American Farm Bureau Federation
American Soybean Association
Beck's Hybrids*
CNH Industrial
Conservis*
Crop IMS
CropMetrics
Dow AgroSciences LLC
DuPont Pioneer
Farm Dog
Farmobile LLC*
Granular*
Grower Information Services Cooperative
GROWMARK, Inc.*
Independent Data Management LLC*
John Deere
Mapshots, Inc.
National Association of Wheat Growers
National Barley Growers Association
National Corn Growers Association
National Cotton Council
National Farmers Union
National Potato Council
National Sorghum Producers
North American Equipment Dealers Association
OnFarm
Raven Industries
Reinke Manufacturing Co., INC.
Syngenta
The Climate Corporation--a division of Monsanto
USA Rice Federation
Valley Irrigation
ZedX Inc.
*Company that has also certified its policy is compliant with the Ag
Data Transparency Evaluator. For more information, visit
www.agdatatransparent.com
______
Exhibit B
Ag Data Transparent Homepage
www.AgDataTransparent.com
Senator Moran. Thank you very much.
Dr. Haman.
STATEMENT OF DR. DOROTA HAMAN, Ph.D., PROFESSOR
AND CHAIR, AGRICULTURAL AND BIOLOGICAL ENGINEERING,
INSTITUTE OF FOOD AND AGRICULTURAL SCIENCES,
UNIVERSITY OF FLORIDA (UF/IFAS)
Dr. Haman. Chairman Moran, Ranking Member Blumenthal, and
members of the Committee, thank you very much for this
invitation to talk about technology in agriculture and data-
driven farming.
My name is Dorota Haman, and I am Professor and Chair of
Agricultural Engineering at the University of Florida, and I
have been working there since 1985. Agriculture is a major
economic driver in Florida. Florida agriculture is very diverse
and focused on specialty crops with most farms smaller and
complex, and they are much more complex than large Midwest
farms dedicated to crops like soybean, corn, or wheat.
This diversity makes introduction of new technologies more
complicated, and data collection is also more complicated, and
analytics are more complicated. Many Agricultural Technology
Providers are focused on agronomic crops, not on specialty
crops, such as citrus, tomatoes, strawberries, blueberries, and
other fruit and vegetables produced in Florida.
Data-driven farming is the future of agriculture.
Furthermore, it is becoming clear that the future of
agricultural operation will embrace the concept of Internet of
Things, a system of interrelated computing devices, machines,
robots, sensors, actuators, and network connectivity. Many
technologies are needed to bring agricultural operation to this
new level, and many of them have been available for some time,
but they are now becoming economical to introduce in
agriculture. Farmers will need to accept these technologies to
be competitive.
New farm technologies and monitoring equipment are already
producing enormous amount of data at a wide variety of spatial
and temporal scales. Raw data are not very useful, but become
very valuable if appropriate algorithms are developed and
applied. Process data used by farmers are also valuable to
others, including insurance companies and commodity markets.
A level of data standardization will be necessary for
optimal sharing and utilization, including a common pool
infrastructure to facilitate transfer and integration of data
from different sources. Common pool infrastructure would also
facilitate collection of information for the USDA National
Statistics Service.
There is no doubt that standardization and openness of
platforms would accelerate solution development and innovation.
However, data ownership and privacy and security of information
in relation to agricultural Big Data analytics must be
addressed.
It can be argued that access to high-quality farm data
gives large agribusiness advantage over small farmers with
limited resources. Open source analytics developed by public
institutions, such as universities and Cooperative Extension
Service using public funds, for example, USDA funds, NSF funds,
would provide the solution for those farmers.
These are exciting times for agriculture. Scientists and
engineers have been focusing on research in the area of
robotics, remote sensing, machine vision, machine learning, for
many years. Progress has been made on early estimation of
yields for specialty crops, such as citrus, strawberries. This
was done using autonomous vehicles and machine vision. These
techniques need to be adapted for other specialty crops.
Extensive weather data and crop models help growers
evaluate climate change adaptation strategies. We have
developed technology to remotely diagnose citrus ``greening,''
a devastating citrus disease. Data-driven technologies can also
improve farm safety through use of alerts and wearable sensors,
and save lives.
I want to thank you for your time. I'm happy to answer any
questions you might have.
[The prepared statement of Dr. Haman follows:]
Prepared Statement of Dr. Dorota Haman, Ph.D., Professor and Chair,
Agricultural and Biological Engineering, Institute of Food and
Agricultural Sciences, University of Florida (UF/IFAS)
Chairman Moran, Ranking Member Blumenthal and Members of the
Committee, thank you for the invitation to talk about technology in
agriculture and data-driven farming. My name is Dorota Haman. I am a
Professor and Chair of Agricultural and Biological Engineering
Department (ABE) at the University of Florida (UF) in Gainesville,
Florida.
I was educated as an agricultural engineer at Michigan State
University and I have been living in Florida and working at the
University of Florida since 1985. My research, teaching and extension
work has been focused on water management in irrigated agriculture.
Since 2007, I have been in a leadership position in Agricultural and
Biological Engineering at UF and have been working with an
interdisciplinary group of scientists and engineers. I have also been
involved in the American Society of Agricultural and Biological
Engineers (ASABE) for over 30 years, serving on various technical
committees and on the Board of Trustees and I am Fellow of this
organization. This testimony represents my view on the emerging, and
rapidly growing area of data-driven agriculture.
Technology is rapidly changing the way we live, work and interact.
This is also true in agriculture. The way we farm and produce food and
fiber is experiencing a rapid change and it will only undergo more
dramatic change in the near future. Agricultural operations and
machinery are becoming more automated, computerized and data-driven.
The future of agricultural operations will include the Internet of
Things (IoT), defined as a system of interrelated computing devices,
machines, robots, sensors, actuators and network connectivity. The
transfer of data in this system does not require human-to-human or
human-to-computer interaction.
More and more sensors are introduced every day to agricultural
operations and monitoring systems. Increasingly, these sensors are
connected through wireless communications to the internet, making data
available in farm databases and mapping systems, and, more importantly,
for enabling analysis of what affects what, why, when and how.
Many technologies that are needed to bring agricultural operations
to this new level have been available for some time, but they are only
now becoming economical for introduction into agriculture. The
convergence of technologies in other fields is rapidly bringing down
the cost of devices and sensors, and therefore making agricultural
applications economically feasible. For example, sensors built as
components of costly medical devices are now economical for farmers.
The medical sensor may cost several hundred (if not thousands) of
dollars but a simplified, maybe less accurate form, but totally
adequate for the agricultural application, is becoming available for a
small fraction of the price. The use of unmanned aerial vehicles (UAVs
or Drones) in agricultural operations is another good example of
technology becoming affordable. A drone that costs a few hundred
dollars today, was sold for thousands of dollars a few years ago.
Agriculture is a major economic driver in Florida. Agricultural
research and extension at the University of Florida (UF) provides
knowledge, innovation and technology transfer that supports 2.2 million
agriculture-related jobs and direct industry output of $148 billion in
Florida (2014). Florida agriculture is very diverse and focused on
specialty crops with most farms smaller and more complex than large
Midwest farms dedicated to crops like soybean, corn or wheat. This
diversity of production often makes introduction of new technologies
more complicated for data collection and analytics. Many Agricultural
Technology Providers (ATP) are focusing on agronomic crops not on
specialty crops such as citrus, tomatoes, strawberries, blueberries and
many other fruits and vegetables produced in Florida.
Florida is also a significant milk producer. Many modern dairy
farms are highly computerized and data-driven to optimize their
operation. Last week, I visited one of the dairies in North Florida.
Each calf is tagged at birth and monitored throughout its entire life
on the farm. Calves are fed by robots that adjust the formula of their
feed based on need at their individual stage of life. The feed, health,
milk production, milk quality, location etc. of every cow are monitored
and available when needed. Solid waste is converted to high quality
compost sold to the ornamental industry and the liquid waste is
recycled through field irrigation to produce more feed for the cows.
This is just a glimpse where modern agriculture is now and where it is
going.
A successful peanut grower in Levy County, remotely controls
center-pivot irrigation systems in response to soil moisture probe
data, from anywhere in the world where there is internet, saving time
and labor. Remote video systems on center-pivots allow visual
monitoring for security and system operation. Enhanced GPS controlled
auto-steering systems on tractors on this farm are accurate to less
than an inch and virtually hands-off. They precisely map the location
of planting, for precise and efficient harvesting several months later.
Such precise operations eliminate over-application and overlap of
pesticides and fertilizers. Soil can be sampled automatically and
indexed to GPS coordinates to regulate a variable fertilizer spreader.
These efficiencies save water, fertilizer and money and reduce adverse
impacts on the environment.
Responding to Florida growers' needs, scientists and engineers have
been focusing on research in the area of robotics, remote sensing and
machine vision for many years. Early yield estimation is critical for
harvest planning and marketing. Great progress has been made on
estimating the yields of various specialty crops such as citrus and
strawberries using autonomous vehicles and machine vision. These
techniques need to be adapted to other specialty crops.
Huanglongbing (HLB)--commonly called ``greening'' is a devastating
disease that has had significant impact on citrus yield and quality in
Florida. Unfortunately, so far, no cure has been reported for HLB. The
first critical step for successful control of HLB is its detection and
diagnosis. It has been demonstrated that high-resolution aerial imaging
using a low-cost UAV or drones can detect the disease. Recently, a
vision sensor was introduced and successfully tested at UF for early
detection of HLB. Since the majority of Florida citrus is already
showing symptoms of the disease, research emphasis is shifting to
genetic solutions to the HLB problem.
Sensor-based management of water delivery through irrigation has
been implemented on many farms across the U.S. and in Florida.
Precision irrigation offers the potential for improving irrigation
efficiency through localized water delivery based on plant needs,
weather predictions and soil moisture sensors. Reported water savings
due to sensor control of irrigation and precise application of water
are on average 60 percent. This results in money savings to a farmer
and reduced use of precious water resources.
Data-driven technologies can also improve farm safety. Wearable
sensors, now under development, can save lives and reduce harm to farm
workers. Alert systems that can detect personal overheating, or inform
about approaching thunderstorms and lightening (typical in Florida),
will also increase safety in the fields. Nine out of the 10 hottest
years on record have occurred in the past decade and, according to the
Centers for Disease Control and Prevention, farmworkers are more than
20 times at risk for heat-related deaths compared to other occupational
groups. Wearable sensors can also alert workers to chemical exposure
through direct contact during application, contact with residue on
plants, or through drift from nearby applications. The wearable sensors
can be also used on various machines, drones and robots to provide
immediate safety intervention.
Weather data systems have been essential for efficient management
of agricultural systems. A system of automatic weather stations and
satellite data provide excellent management tools for Florida growers.
The cost of a sophisticated blue-tooth-enabled weather station is now
less than $1000. This monitoring will become more critical as climate
change leads to increased changes in seasonal temperatures, rainfall
patterns, and the frequency and severity of storms. For example, in the
last few winters, Florida growers have been affected by insufficient
chill-hours to optimize production of temperate fruits such as
blueberries, peaches, and strawberries. Adapting to these changes,
Florida blueberry growers are experimenting with ``evergreen''
production of blueberries that does not require chilling. Agriculture
in Florida is not only sensitive to the manifestations of climate
change mentioned above, but also to the salt water intrusions in
coastal irrigation wells as a result of sea-level rise.
Data quality has always been a key issue in farm management
information systems, and is more challenging in an era of Big real-time
data. Intelligent processing and analytics of Big Data is challenging
because of the large amount of often unstructured, heterogeneous data
which requires a smart interplay between skilled data-scientists and
domain experts. At present, new farm technologies and monitoring
equipment are producing enormous amounts of data at a wide variety of
spatial and temporal scales. Raw data sets are cumbersome and not
directly useful, but become very valuable if appropriate algorithms are
developed and applied. Data analytics, frequently provided by ATPs, are
being developed for agricultural applications and are a necessary step
to make these data valuable to growers. These processed data are also
very valuable to others including insurance companies and commodity
markets. A level of data standardization will be necessary for optimal
sharing and utilization, including a common pool infrastructure to
facilitate transfer and integration of data from different sources/
companies.
New technologies are introducing new problems and issues that need
careful consideration. There is no question that openness of platforms
would accelerate solution development and innovation. However, data
ownership, and related privacy and security issues are problems that
are frequently discussed in relation to Big Data and analytics. These
concerns need to be addressed, realizing that enforcement may slow down
innovation.
It can be argued that access to high quality data, that allows for
predictive business modeling of every aspect of farming, gives large
agribusinesses advantage over farmers, especially small farmers, who do
not have sufficient resources to pay for data analytics.
Leveling the playing field for smaller farmers, such as Florida
farmers who produce specialty crops, through the use of open source
analytics developed by public institutions such as universities and
Cooperative Extension Service using public funds (e.g., USDA, NSF) is
an option, providing data processing though utilities (apps),
interactive models and maps.
In summary:
Agricultural operations and machinery are becoming more
automated, computerized and data-driven. It is becoming clear
that the future of agricultural operations will embrace the
concept of Internet of Things (IoT).
Technology is becoming less expensive, and economical in
agricultural operations. Farmers will need to adapt to be
competitive.
Farmer-collected raw data become valuable if appropriate
algorithms are developed and applied. Data analytics,
frequently provided by ATPs, are necessary to make data
valuable to growers.
The processing and analytics of agricultural Big Data is in
its infancy, and is challenging because of the large amount of
often unstructured, heterogeneous data which requires a smart
interplay between skilled data-scientists and domain experts. A
common pool infrastructure should be developed to enhance
sharing and integration.
Land-grant universities and Cooperative Extension Service
may provide a public platform for data processing though
utilities (apps), interactive models, and maps.
Data ownership, and related privacy and security issues, are
problems that are frequently discussed in relation to Big Data
and analytics. These concerns need to be addressed.
I want to thank you for taking the time to focus on technological
innovations in agriculture. Thank you for inviting me to testify today.
I would be happy to answer any questions that you might have.
Senator Moran. I want to thank you for your time. Thank you
very much. And there will be questions.
I think I'll turn now to the Ranking Member for purposes of
his questions, and we'll then deal with the other Members of
the Senate who are here, and I'll go toward the end.
So, Senator Blumenthal.
Senator Blumenthal. Thanks, Mr. Chairman.
And thank all of you for your testimony. And I think that
you have all identified one of the key issues here, ``How do we
enable farmers to maintain control, either through some kind of
opt-in consent or some other means?''
And I just want to call your attention to the fact and ask
your comment on the vote that the Senate took not so long ago
on a resolution to overturn the Federal Communication broadband
privacy rules, which I found extraordinarily regrettable. These
rules were repealed, which constituted, in my view, an attack
on consumer rights to privacy, but also these rules underlay
the potential for protection of privacy in your industry as
well.
So I want to ask each of you whether you've considered that
all of these efforts may be undermined by the repeal of the
FCC's broadband privacy rules because they leave no baseline
protection, privacy protection, in place, and broadband
providers can sell and share farmer data just as they can other
consumer data, and whether you might recommend that we and
principally the FCC reinstate those rules? And I'll ask that
question of all of you.
You look like you're ready to answer, Dr. Ferrell.
Dr. Ferrell. Well, and the lens through which I've examined
a lot of the protections for agricultural data is interesting
because there is not really a specific existing statute that
would address the type of data that we're discussing here.
If my tractor uses a cellular signal and basically makes a
phone call to a service provider, there are Federal rules
governing telephone communications that would protect that
transmission. If I e-mail a file to my crop consultant, there
are protections for that as well.
But if we're just talking about the transmission of
telematics data, that really doesn't fall comfortably into any
realm, especially when we start to combine and package that
agricultural data with other things.
One example I sometimes give is that if my data service
provider has raw agronomic data about my farm, but they also
have a customer relationship management software program that
pairs with that data, well, now there might be protectable, you
know, personally identifiable information that is now coupled
with my farm, that's a very different privacy issue than if
we're just talking about agronomic data.
And so I think one thing that we may have to do is--I kind
of agree that we need more robust consumer protections for data
generally, and I think specifically when we talk about
agricultural data, we've got a lot of work to do definitionally
in defining what this kind of data looks like. It's not health
data under HIPAA, it's not financial data under the Fair Credit
Reporting Act, it's something unique unto itself that I think
may deserve some efforts to actually define individually and
define what those protections should be.
Senator Blumenthal. You are certainly correct that there
are lots of different kinds of data, even within farming, but
the principles I think have to be framed in a way that they're
applicable to all of these kinds of data. And you're right
also, there may not be a specific statute here, but the
principles have to be embodied at some point in regulations or
rules or statutes, and the authority is there for the FCC to
adopt protections.
So I would just invite you to think about what those
protections should be, because right now a provider could paint
an extraordinarily detailed picture of a farmer's business
practices, from the type of feed he may use, the type of crops
he grows, the type of fuel he uses, how many workers, how many
livestock.
And I recall well a lunch I had, referring back to my early
experience, with my grandfather, who had a farm--by the way,
Senator Fischer, just south of Omaha--and the lunch was at the
stockyards, and I was asking someone having lunch with us, one
of my grandfather's friends, how many cattle he was feeding.
And my grandfather nudged me and gave me that look of
disapproval that grandfathers do when they are not enamored of
the line of questioning, and he explained afterward that's like
asking somebody how many dollars he has in his bank account.
And that sense of privacy is broken, I think, by many of these
practices, and farmers are very, very respectful about their
privacy, with good reason.
Thank you.
Senator Moran. Senator Blumenthal, thank you.
The Senator from Nebraska, Senator Fischer.
STATEMENT OF HON. DEB FISCHER,
U.S. SENATOR FROM NEBRASKA
Senator Fischer. Thank you, Mr. Chairman.
And thank you, Ranking Member Blumenthal.
As a cattle rancher, I thank you for your respect to my
privacy.
I would like to thank both of you for calling the hearing
today so we can focus on managing that Big Data with new
technologies for our Nation's farmers and ranchers. Not only
will this help boost productivity, but advancements in digital
analytics can improve how we feed our communities and conserve
our natural resources. In fact, Tim Hassinger of Lindsay
Corporation, from Omaha, Nebraska, noted at last week's
Internet of Things hearing that the combined yield enhancement
and resource savings from new technologies can increase the
American farmers' profits by an average of $40 per acre.
And I would like to please submit for the record follow-up
materials that Lindsay has provided regarding its research on
data analytics and irrigation management, Mr. Chairman.
Senator Moran. Without objection.
Senator Fischer. Thank you.
[The information referred to follows:]
FOR IMMEDIATE RELEASE--November 8, 2017
Lindsay Corp President and CEO Speaks to U.S. Senate Committee
Tim Hassinger Addresses the Need for Improved Access to Rural Broadband
Service
(OMAHA, Neb.)--November 2017--Tim Hassinger, president and CEO of
Nebraska-based Lindsay Corporation, offered testimony during a hearing
yesterday conducted by the U.S. Senate Committee on Commerce, Science
and Transportation. The hearing, titled Advancing the Internet of
Things in Rural America, focused on the benefits of the Internet of
Things (IoT) in rural communities and the infrastructure needs
necessary to advance the IoT market to ensure rural America has access
to products and devices that are driving the digital economy.
``Like all business owners, farmers in rural communities need the
ability to go online,'' Hassinger said. ``The Internet fuels the
innovative, advanced technology that will help America's farmers meet
the food, fuel and fiber needs of our rapidly growing global
population.''
Hassinger's testimony contends that with reliable, high-speed
Internet access, farmers can take advantage of tools that deliver
hyper-local weather forecasts, real time data on soil moisture
conditions and GPS for planting and irrigation management. They can
also take advantage of a myriad of emerging technologies available from
Lindsay Corporation and other American manufacturers. The testimony
further explains that these innovations enable efficiencies through
remote data collection, transfer and analysis from connected devices
like soil moisture sensors, weather stations and cloud-based tools.
``At Lindsay Corporation, we are developing and deploying
technologies to help growers produce more with less. For example, our
FieldNET and FieldNET AdvisorTM remote irrigation
management and decision support tools help farmers decide precisely
when, where and how much to irrigate--to help them maximize yields
while reducing overwatering and related input costs and nutrient
losses,'' Hassinger said. ``But we know the technology is only as good
as the farmer's ability to access it.''
According to the Federal Communications Commission Broadband Access
Report, an estimated 39 percent of the rural population (23.4 million
Americans) lack access to broadband that meets today's benchmark speeds
of 25 Mbps for downloads and 3 Mbps for uploads. By contrast, only 4
percent of urban Americans lack access to 25/3 Mbps broadband.
``While cities and municipalities typically have access to several
high-speed Internet service providers, that access often ends at the
city limits. Those living in rural communities must depend on radio
networks, satellite or cell service--all of which typically operate at
lower speeds, limiting connectivity,'' Hassinger said. ``All farmers
are faced with the pressure to increase yields while conserving
resources. The lack of reliable broadband hinders their ability to
adopt the new technologies that will help them optimize their
operations and compete in the global marketplace.''
Hassinger's full testimony can viewed at https://commerce.senate.gov.
(https://commerce.senate.gov.)
Report
[GRAPHICS NOT AVAILABLE IN TIFF FORMAT]
Senator Fischer. Mr. Knopf, in your testimony, you describe
your farm management operations. And what benefit do you see on
your farm from implementing Big Data technology? How do you
analyze it? Do you have input from your nutritionists, your
equipment dealer, local co-op? How do you get it all put
together?
Mr. Knopf. That's a good question. There's a lot of
information to try to analyze. And so in our farm--that answer
is going to look different for each individual farm, of course,
on the strengths and talents and the passions of that
individual farm operator. Some farmers are very passionate
about their data, very passionate about precision agriculture,
and they want to be very hands-on in that process. And those
are also I think the guys that are very passionate about the
ownership and transparency of that data. They want it on a
server in their office.
There's the other spectrum of farmers--and we're kind of in
the middle, let me say that--there's the other spectrum of
farmers, and sometimes it changes based on age demographics as
well. I've noticed in my community, in my experience, and
thinking about our own farm, the older generation is much more
private about their data, very unwilling to share it. The
younger generation is--you know, they grew up with smartphones,
and Google is tracking them all the time and telling them
things about themselves that they didn't know already, and so
they're more used to that. So there's a generational change
there as well.
For us, we're kind of in the middle of--I have an agronomic
background from Kansas State. I have a young man that works--we
have one full-time employee that also is an agronomist. And we
do most of our data analysis in-house. But our data is stored
in a software system that is web-based, and it's subscription-
based. And that was a big change for us about 2 years ago, 3 to
4 years ago, is transitioning our data from being stored in-
house to web-based, which is somewhat of an uncomfortable step
because you feel like you're losing control, and that's why the
transparency is so important. But we are still processing a lot
of the data and analyzing a lot of it in-house. But there is a
growing amount of service providers dealing with that. We
collaborate a lot with our land grant with Kansas State
University, and a lot of their research we collaborate with and
host on our farm. So that's a very important collaboration. And
I appreciate two representatives here from universities as
well.
Senator Fischer. And as you noted in your testimony, the ag
economy is in a downward cycle right now, it's hurting. I also
know that there's a lack of access to broadband, and we have a
lot of slow downloading speeds. And how does that affect your
bottom line----
Mr. Knopf. That has a----
Senator Fischer.--when you have this slow access?
Mr. Knopf. Yes. Excuse me. That has a very significant
impact. In fact, the county that we live in is just now
receiving fiber optics ran to farms. In fact, it's not even
hooked up yet, it's just being ran right--right now, being
buried, because a lot of my peers that live in more--we have a
community of about 40- to 50,000 within our county, but yet
there is still a lot of rural part of the county. So a lot of
my colleagues that live in rural counties in Kansas qualified
for grants from the government to help build that initial
infrastructure.
Well, our county happened to not qualify because of the
urban center that we had within that county, and we have some
broadband antennas on top of elevators and towers and so forth,
but it is very unreliable and fairly slow. So that has limited
our utilization and access of data, particularly in web-based
software platforms, and been an economic disadvantage in
utilization of data because we're kind of caught in the middle
of that process.
Senator Fischer. Thank you.
Thank you, Mr. Chairman.
Senator Moran. You're welcome.
Senator Klobuchar.
STATEMENT OF HON. AMY KLOBUCHAR,
U.S. SENATOR FROM MINNESOTA
Senator Klobuchar. Thank you, Mr. Chairman. Thanks for
allowing me to go, too. And just to follow up with Senator
Fischer, and we've worked on some of these broadband issues, I,
like Senator Fischer, have heard many stories from my state
about the need for better broadband access. And I always
remember a precision ag company out of Willmar, Minnesota, that
actually told farmers to go to the restaurant parking lot
because they would get better Internet.
And when you think about it, you know, way back about 10
years ago people were just trying to get Internet so they could
write their grandkid an e-mail or something, and it has changed
so much that if they don't have that ability, they're just not
going to be able to do their work basically, and that's
everything from Jennie-O in Minnesota measuring the temperature
in barns for turkeys--I mention turkeys, we are number one in
turkeys in Minnesota, and Thanksgiving is upon us--and other
things that we've seen that it's just really a must-do for
these farmers.
Mr. Ferrell, most broadband connections provide faster
download than upload speed. Why is upload speed important for
ag data? And how are reliability and bandwidth needs of farmers
different than the average consumer?
Dr. Ferrell. Well, you're very right, Senator, in that
we've seen tremendous gains in terms of what rural broadband
access has been able to offer in terms of download speeds, and
that's been good for farmers acquiring information. But as the
witnesses have mentioned today, we're gathering tremendously
more data that has to be uploaded. And in certain cases, in
some applications, we need that data to be uploaded in real
time.
And so even if you're in a broadband-connected accessible
area, your farm equipment needs to be able to access wireless
broadband speeds as well to upload that data and receive real-
time updates about things like cropping prescriptions, like Mr.
Knopf was mentioning.
As you're doing intensive management of a specific field,
you need to have advanced broadband upload speeds to be
collecting that information, getting real-time recommendations
back, so you can make on-the-fly adjustments to things like
seed population, fertilizer application, pesticide application,
and things of that sort.
So the download speeds are good, but as the ag data
revolution continues, upload speeds I think will be even more
important because of the amount of data that we're pushing to
the cloud rather than receiving from it.
Senator Klobuchar. Thank you.
Dr. Haman, I just recently visited, in fact, just a week
ago, Aker Technologies in Winnebago, Minnesota. You may not
have been there before, but they're developing this innovative
drone technology to help farmers be more efficient when
applying fertilizer, pesticides, and water. It saves money. It
leads to more sustainable farming because they're going to be
able to see where they need the water, where they need the
pesticide, and not do one-size-fits-all for their whole fields,
nor do they have the resources, especially small farmers, to be
able to walk through every row.
The FAA recently introduced a pilot program to explore the
use of drones far beyond the line-of-sight operations in
limited cases. For farmers managing large tracts of land, this
could be a valuable option in the future. And in your
testimony, you mentioned the use of drones to combat citrus
diseases in Florida. Could additional drone monitoring help
prevent the spread of diseases?
Dr. Haman. Absolutely, yes. They absolutely. We already are
using drones, for example, in forestry. And in forestry, the
drones can detect the areas where the disease is likely, where
the plants are stressed. Using different spectra, you can
detect stress which is from water or stress which is due to
disease, and separate that. And we are right now doing research
to detect specific stresses due to specific deficiencies of
different nutrients and different--so possibilities are
enormous, and as the cameras are getting smaller, it makes also
big difference because they can be carried on small drones.
Senator Klobuchar. Exactly. And that's what this concept is
going on at this company.
Dr. Haman. Yes.
Senator Klobuchar. Last question, Mr. Knopf. The Minnesota
Corn Research & Promotion Council has been working with local
farmers to test innovative conservation approaches, pilot new
technology applications, by promoting the adoption of precision
ag technology, like drone monitoring, smart irrigation, that I
was just talking to the doctor about how we can help farmers
with conservation.
Could you talk about how Federal research initiatives or
land-grant universities partner with industry to further
advance innovation? Quickly.
Mr. Knopf. I really appreciate that question, and it is
invaluable and becoming increasingly invaluable to both
economic sustainability and environmental stewardship. We host
yearly on our farm between four to a dozen experiments in
collaboration with Kansas State. Their resources are limited
more as time goes on because of budgets, so farmer-private
university collaboration is increasingly important.
Senator Klobuchar. Thank you very much.
Thank you, Mr. Chairman and Senator Blumenthal.
Senator Moran. You're very welcome. Thank you, Senator
Klobuchar.
Senator Lee.
STATEMENT OF HON. MIKE LEE,
U.S. SENATOR FROM UTAH
Senator Lee. Thank you very much. Thanks to all of you for
being here. We're witnessing a technological revolution. It's
very significant. I think it's going to soon be the case we'll
look back at what we're experiencing today and realize that
this was just the tip of the iceberg. And it's one that I think
has the ability to allow farmers and ranchers enjoy cutting-
edge technology and make their operations more effective, more
productive, so that they can feed more people and do what they
do.
I want to focus today a little bit on the application of
unmanned aerial vehicle technology, drone technology, as we
sometimes refer to it, to agriculture.
Mr. Knopf, I want to follow up on some questions Senator
Klobuchar asked you about. Tell us what--how this has potential
application in your field?
Mr. Knopf. We happen to be 15 miles from the Polytechnic
campus of Kansas State University, which is in Salina, and they
are one of the leading institutions for UAS and UAV
programming. For the last couple years, they have been carrying
out a pilot project alongside the FAA with a grant they
received on some extended line-of-sight operations for drones,
which was already brought up, and looking at some of those
regulations and how that can work.
So specifically what that means to our farm, other
collaboration that we're doing with Kansas State from the
Agronomy Department is looking at things such as being able to
sense NDVI ratings, which is a plant health greenness index,
and across the field in real time to very quickly pick up any
problem areas in the field, to quickly address issues that need
to be addressed before they make more significant impact. And
it's just a better, more efficient, precise way to collect data
across a wide scale. So there are a lot of implications in
that, and I think satellite imagery is also an increasing area
of data that is going to be easily accessible as well that will
have a lot of applications.
Senator Lee. Thank you. And following up on that, Dr.
Ferrell, you mentioned in your testimony that drones are one
technology, one useful piece of technology, that can be used to
gather agronomic data. Can you explain how important this is
for farmers, why it's important?
Dr. Ferrell. Yes. I think it's incredibly important because
it's constantly being called upon for farmers to do more with
less, and that's not just in terms of input, sometimes it's in
terms of labor as well. You know, we've talked about the use of
UAS and drones to scout crop conditions, and certainly that
enables one farmer to cover potentially thousands of acres in a
day whereas if they were trying to do that on the ground by
scouting, that might take days out of their schedule, and they
just don't have the manpower to handle that.
But another important aspect of UAS use that we haven't
talked about much is actually in livestock operations,
especially if you're talking about large, you know, pasture-
based operations, sometimes using drones to check cattle can
save a tremendous amount of time and resources, and enable us
to identify cattle that are in poor health, get them treated
quickly, and get them back to health.
So I think there are unprecedented opportunities there for
us in the agricultural sector.
Senator Lee. It could help a lot of people in their
operations.
Now, Mr. Janzen, your law firm represents people in
regulatory compliance and registration. What can you tell us
about what the largest regulatory barriers are to farmers and
ranchers using drone technology, or for drone technology more
broadly being implemented in the United States?
Mr. Janzen. Thank you, Senator. I think right now probably
some of the largest barriers for farmers would be the cost of
the drones and also the ability to use the data and get real
results back that they can put into practice on their
particular farms. And I think that there was a lot of hype when
drones first started to become very popular, and, you know, now
here we are today, and we're trying to figure out how to best
use that information today. So I don't know currently that
there is an enormous regulatory hurdle for use of those drones
on the farm.
Senator Lee. What about the FAA's restriction on the use of
drone technology beyond line of sight?
Mr. Janzen. Oh, yes.
Senator Lee. Has that had an impact on people's adoption of
this technological platform?
Mr. Janzen. That's a very good point. Yes, it probably is
creating some restrictions on people wanting to use the
technology because that is a barrier to, you know, a single
farmer who may want to use this technology on their farm.
Senator Lee. OK. It's one of the reasons why I've been a
big believer in and an advocate for legislation that would help
provide clarity on drone regulation. This is a technology that
has tremendous promise, not just for the U.S. economy, but for
humanity, but because the technology is developing in the way
that it is here in the United States, it can offer some real
benefits to the United States specifically.
As we've seen in other areas, one of the things that can
hold back technological innovation is stifling government
regulation, and so that's why I've been an advocate on that.
Thank you, Mr. Chairman. I see my time has expired.
Senator Moran. Thank you, Senator Lee.
Let me start I think with Mr. Tatge, although I'm happy to
have a response from any of you. The promise, or a promise,
from data collection and its usage is its increasing value,
and, therefore, added profitability to production agriculture.
So in your testimony, Mr. Tatge, you estimated that a
billion dollars of annual revenue could be returned to rural
communities from multiple sales of that information.
Mr. Tatge. Yes, sir.
Senator Moran. So how do you envision Mr. Knopf earning
additional income as a result of collecting data on his farm?
Mr. Tatge. So we create what is called an electronic field
record. An electronic field record is a map basically of what
happened on what part of the field, whether it be seeding rates
that are being planted, whether it be application rates from
sprays that may have been applied to the crop, and then from
harvest as well. And once you get those--that data in a state
where it's in a standard, and we call this electronic field
record, we can make that available for sale.
So when we make that available for sale, companies come to
us, and we've already sold data for farmers, will come to us
and say, ``I would like to buy that layer of data, and this is
what I'm willing to pay.'' And then we go back to the farmers
and we tell them who the buyer is of that data and how much
they're willing to pay, and the farmer chooses if they want
to--let's think of it as licensing a copy of that data to that
company. And if they do, we split the revenue with the farmers.
So we've turned it into a profit center for basically
taking digital exhaust and packaging it to get it to go to the
people who have a desire to acquire it.
Senator Moran. So you're the market. You're the
intermediary between those who have something to sell and those
who have something they want to buy?
Mr. Tatge. Yes, sir. We're also helping with the collection
of the data in the first place to get it standardized in a way
that makes it easily portable or interoperable. Because our job
is not to do any analytics on data, our job is to help farmers
get the best dataset that they can get in one place and in a
format that it can move wherever they direct it to go.
So when our focus is exclusively on being able to focus on
how can we get the best dataset into a database, and then how
can we get that data to flow into whatever system the farmer
directs us to send it? If they're working with their local
agronomist, we've got several examples where a farmer is
automatically sending data to their agronomist without having
to do anything. The day they get done spraying the field, this
file is going directly to their agronomist, and there's no
charge for that.
The farmer can choose to share that data with anyone they
want. It's when you get enough of this data packaged together,
though, that's when you bring the opportunities for
monetization of the data, just like a brand-new commodity.
Senator Moran. That's not much value to individual data in
and of itself, an individual farmer's data, but if you can
combine it with broad data, bigger data----
Mr. Tatge. Yes, there are values to it on the farm itself
for your operations, making sure you're running your machines
as efficiently as you can, fuel usage, idle time. It's a
different value prop to the farmers. Then there's a value to
sharing that data with trusted advisers, and it adds increased
service opportunities and increased ability to react during the
in-season mode to make some changes to hopefully put down any
threats that come up or minimize their impact.
And then to the market, there's a greater opportunity to
really start to identify and make these datasets available for
us to understand best practices. Are we planning--how come when
the farmer rips open the bag, we are--we lose all this genetic
yield potential? Is it because they're planting the seeds too
deep? Are they planting them too shallow? Are they planting
them at the wrong time? Are they planting them too early? Are
they planting them too late?
There are all of these things that get opened up with the
data for us to identify what the best practices are and what
the best pieces of equipment are to help drive efficiencies.
It's very exciting. But that is how you can create a revenue
stream going back there because the major ag companies that you
see today, they could easily change their marketing budgets.
Instead of buying commercials and hats and jackets and boots--
right?--we could change that into let's be open and honest with
our customers, let's buy their data from them so that we have
the opportunity to service them better.
Senator Moran. Who are the purchasers of data today? And
who do you expect them to be in the future?
Mr. Tatge. They range from a wide range of data analytics
companies as well as insurance and reinsurance companies trying
to identify risk levels and some of that. They're also
equipment manufacturers. So there are a bunch of people that
are trying to figure out and benchmark data against each other,
and they're getting access to data that they've never had
before.
Senator Moran. Is this function that you're helping create,
is it similar to what USDA does as they collect data and then
make it public for purposes of farmer decisionmaking----
Mr. Tatge. Yes, it can be. You know, the USDA has got the--
I mean, they are the gold standard of data, and they are the
ones--when you see the markets move a lot, a lot of times it's
because a USDA report came out, and it wasn't what the market
was anticipating, and that causes a lot of the volatility
there.
Traders love the volatility, and I was one for several
years. But farmers traditionally, and the companies that use
that raw commodity as the base for their product, like a flour
mill, does not care for the volatility. So getting that
volatility out of the market and getting that information to
the market quicker is a huge opportunity to help reduce, I
think, overall volatility in the markets.
Senator Moran. In the billion dollars that you estimated,
does that include all three--the value of all three components
of that data that you described?
Mr. Tatge. That includes--no, that is the value of the
marketplace side of the data only.
Senator Moran. And you base that estimate on what?
Mr. Tatge. We base that estimate on 25 percent of the
acreage in the four major row crops--which would be corn,
beans, wheat, and cotton--and being able to monetize that data
four to five times on a turn. Once one person buys the data,
the others want to buy it as well.
Our problem right now is the fact that we don't have enough
data. And when you go into a new market like this, when you're
creating a new marketplace, supply brings demand. So that's one
of the things that we're working very hard with right now to
address.
Senator Moran. And this is a question for again maybe you,
Mr. Tatge. So when a farmer buys a piece of equipment, at what
point in time--what year, what model year, of equipment
collects the data? Where did we start? How recent of a
development is this?
Mr. Tatge. So I think in 2011 is when most of the new
equipment started coming with a modem in it. I think it was a
2G modem at that time. So I would say from 2011 and newer, most
of the equipment that is sold today has a modem in it. But we
have an aftermarket solution that allows you go to back to
roughly 2002 on most of the pieces of equipment. So as long as
there are sensors on it, we have a piece of hardware that can
allow us to make that older piece of equipment more modern in
its form of collecting data.
Senator Moran. So when you say there's a shortage of data,
is it because farmers are using equipment that's older than
that, the technology isn't utilized on their farm, or they
don't know or utilize the information, the data, that they
have?
Mr. Tatge. It's the second one more so. I would estimate
that 80 percent of the data generated in cabs of equipment
today does not make it out of the cab ever, and that's a
roadblock. The industry lacks--I'm going to use a navigation
app for me. So Google Maps is something that I use just about
everywhere. I used it to get here.
And when I went to buy my new pickup, I couldn't get it
without the electronics in it, without the navigation system. I
wanted to because it was expensive and I didn't need it. But I
couldn't buy that truck without it. My wife has a car as well
that we've had the same challenge on not buying the navigation
system. But we ended up buying two of them that we don't use.
And the reason we use Google Maps is because of the interface
and it is portable and it can move with me.
The ag industry lacks a common interface of that nature to
be able to get this data out of the cab. So when you only plant
once a year, I mean, you kind of forget how the technology
works until next year. And that's one of the challenges that we
have, and we have the opportunity with the technology available
today to make those tools smarter and easier for digital
immigrants.
Senator Moran. Thank you very much.
Do any of you have anything to add in regard to how this
use of data, the monetization of information, means money
returning to rural America?
Well said, Mr. Janzen?
Mr. Janzen. Yes. I would add a little bit to that. I always
tell people, because we get asked this question a lot, ``What
is the real value of the farm data?'' Right?
And I always say imagine that you had two 80-acre parcels
of land that you had grown corn and soybeans and wheat on for
the last 20 years, and, you know, one of those you have
accurate data records for how things were planted, when they
were fertilized, when you applied pesticides, and one you did
not. You know, how much more would somebody be willing to pay
for that information at the time of sale than the parcel that's
identical that doesn't have any of this that goes along with
it?
And so I think the answer is certainly something, right? I
mean, it's certainly worth something. Is it worth $20 an acre?
$50? $100? I don't think we really know that yet because it's
such a young industry. And I also think we haven't done a good
enough job of collecting that information year after year after
year to really see the value. But we will get there and there
will be a time when that is a valuable commodity and certainly
people are willing to pay more for it.
Senator Moran. Mr. Knopf?
Mr. Knopf. Mr. Chairman, I may quickly mention that I think
it's important to keep in mind that the value may not always be
just monetary. I think one of the things we're seeing in
private industry right now is there is tremendous competition
to grasp and collect that digital acre. Sometimes the value
going back to the farmer is in benchmarking against other farms
on our performance from a yield standpoint or a sustainability
standpoint. Other times, the value coming back to the farmer by
sharing his data is access to a model that that data can be
implemented to that has been designed by Big Data that a
private industry has paid to capture.
And so it may be access to a model or access to something
other than monetary value, it may also be the nugget given back
to farmers for sharing data.
Senator Moran. When Mr. Tatge says that 80 percent of the
information collected from a piece of farm equipment is not
being utilized, is that by the farmer? Is the information still
being utilized by the manufacturer, the seller of the farm
equipment?
Mr. Knopf. From my perspective, in most of our equipment,
it is not automatically, as far as I know, automatically being
sent somewhere even though there are modems in the machine.
And, yes, I think most of it is that it's not being utilized by
the farmer.
You know, we only have so much time. I think a lot of this
data utilization is going to have to be, with the exception of
very large farms, more of it's going to be outsourced to
service providers, which is a new economic opportunity for
rural America, and the drone industry as well.
Because the farmer just literally doesn't have the time to
manage all of this data to take it out of the cab and, you
know, I can't--my dad, we know the age of farmers, and this is
going to be an important transition, as that generational
change happens, there is going to be a lot of change as we see
that transition happen, but the current generation of farmers,
it's a big technological hurdle to cross for most farmers.
Senator Moran. Because of the access to this information,
farmers can be more efficient and, therefore, more profitable
is the theory, and I think the facts I think will demonstrate
that. Does that increase the chances that farms get larger, or
does it make it more likely that the medium-sized farm becomes
more efficient and, therefore, more sustainable economically?
Dr. Ferrell, you look interested.
Dr. Ferrell. We've actually speculated about that in some
of the academic research. And there's definitely the argument
to be made that says there are efficiencies to be had, and
those efficiencies might be magnified at the larger scale of
farms. And so from one perspective, that could actually
accelerate the growth and the scale of farming, and that's one
thing that's true.
It depends on what we talked about earlier, which was,
``How do we go about sharing the value and the gains from those
technological advances between the aggregators and the
individual farmers?'' It might actually be that this technology
might enable our medium and smaller producers to capture some
of the efficiencies that heretofore were only available to
those larger producers and might actually keep them more
viable. I think there are probably more macroeconomic trends
that are going to still push toward larger operations, but I
think this could be an important source of invigoration for
those small and medium producers as well.
Senator Moran. That's encouraging.
Senator Blumenthal has some additional questions. We'll
have more one more round. Senator Blumenthal will ask his
question, I'll ask mine, and we'll try to wrap this up.
Senator Blumenthal. Thank you.
I think, Dr. Ferrell, you were the one who used the term
``aggressive fine print.'' Or Mr. Janzen, sorry.
[Laughter.]
Senator Blumenthal. And let me address this question to the
panel, but first to you. What do you worry about in those
contracts, performers in the fine print, and is one of those
worries mandatory arbitration clauses, which, as you know,
deprive any potential party and litigant of the right to go to
court under some circumstances, often very broad circumstances,
and often contained in the fine print?
Mr. Janzen. Thank you. Yes. Well, as an attorney, I worry a
lot about this, and I get asked to draft these types of
contracts by ag technology providers that hire our services.
And so we think about this a lot. And, you know, what I always
tell my clients is you want the contract to be something that
the farmer can understand when they read it. And that's very
different than I think the traditional route that we've gone in
this country.
I mean, we all get the updates on our cell phones all the
time, and nobody reads those, you know, and that's a problem,
and it's a real problem when the information is not just, you
know, the stuff that you're clicking on your cell phone, but
it's actually proprietary information that belongs to a farmer.
So, you know, I worry a lot about the complexity of these
agreements, and I worry a lot that the industry adopts standard
form contracts for these new agricultural technologies when
it's like putting a square peg in a round hole and it really
doesn't work.
Regarding arbitration provisions, I haven't seen that as
widely used as probably in some other areas, but they are
definitely in there. And, yes, that certainly could be a
problem. You know, as somebody whose Equifax data was breached,
I can certainly relate to that issue.
I would say the other thing that I concern myself with is a
choice of forum provision in a contract because if anyone has
to go to New York City to adjudicate a dispute, and they live
in Wyoming, you already know the outcome of that, right? So
that's the other thing that I worry about in these standard
form contracts.
Senator Blumenthal. Any other responses?
Dr. Ferrell, you're nodding.
Dr. Ferrell. Oh, I would just agree. We've seen those in
lots of contracts. I've seen a lot of those in some of the ag
data contracts. And interestingly enough, I've seen those a lot
in oil and gas contracts, wind energy contracts, solar
contracts, and there is some limited empirical evidence that
shows that those arbitration clauses really work to disfavor
the landowner in those circumstances because not only is the
choice of forum usually a remote urban location as opposed to
the home court of the producer, but also the arbitration
process usually kind of tends to stack the deck in favor of the
more--and I don't want to call these unsophisticated, that's
not what I intend to say, but corporate entities tend to see
arbitration more, are very familiar with that process, and use
it to their advantage I would say generally.
Senator Blumenthal. They're more experienced in it.
Dr. Ferrell. Yes, sir.
Senator Blumenthal. Especially if it's mandatory.
Dr. Ferrell. Yes, sir.
Senator Blumenthal. Thank you.
Senator Moran. Mr. Knopf, does K-State Extension, for
example, have any ability to educate you and other farmers in
regard to data? Is that happening?
Mr. Knopf. K-State Research and Extension has data
specialists. I'm very impressed with Dr. Ferrell's knowledge.
Come north a little bit if you----
[Laughter.]
Mr. Knopf. So, yes, we have precision agricultural
specialists, but, again, it's--again, it's another--it's a
whole other realm that they have historically not dealt within,
and the amount of capital and things to just fund positions is
I think limited in how extensively they can take on that roll.
Senator Moran. Is there anyone who is vying for your data?
Does somebody want to buy it?
Mr. Knopf. Nobody has offered me any money for it yet.
[Laughter.]
Mr. Knopf. But, yes, certainly private companies are not as
much directly, but very indirectly competing for access to my
data, and they don't really know, I think in many instances,
yet what they're going to do with it because it's the amount
that they can collect. The bigger the amount is, the more
value. Whatever that value is, it's about scale. And so I think
my perception as a farmer is a lot of times they don't know or
aren't letting on what they're going to do with it, but once
they have a bigger scale of it, there's inherently going to be
an incredible amount of value there. And the competition for me
to just hand over my data card is very high.
Senator Moran. Is it possible that a monopoly could arise
owning the data, that the information is available only to a
select few? Could an exchange end up being the purchaser of
data and know more than others who are interested in the
markets?
Dr. Ferrell.
Dr. Ferrell. That's a perfect opportunity for me to loop in
the fact that Dr. Terry Griffin at K-State, their precision
cropping systems economist, actually taught me a lot about this
issue specifically. And if you look at the trends that we've
seen in other technology providers, take, for example,
operating systems for personal computers, you start with lots
of those systems out there, and over time, through
acquisitions, market share, and the fact that, as Mr. Tatge
said, we want to see interoperability, that's why so many of us
either have a Mac or a PC, is because eventually those two
operating systems kind of boil down, and I can create a
Microsoft Office document and share it with everybody who's on
the Committee with me here fairly easily.
And so lots of factors, economic and practical, tend to
drive us toward that, either monopoly or monopsony, that small
group of providers that are eventually going to take over.
There are lots of reasons to suspect you would probably see
that in the data space as well, and that's why I think that the
dialogue between the producers and the aggregators is so
important, is to make sure that as that consolidation occurs,
that we maintain some parity in the bargaining position between
the users and those aggregators.
Senator Moran. That could be a potential difference between
USDA information and data that's collected through this process
we're talking about. The access could be much more limited in
this setting?
Dr. Ferrell. I think so. And the interesting hypothetical
situation is, you know, we have USDA data trying to provide us
information about crop conditions, acres planted, things of
that sort, and, you know, as the other witnesses mentioned,
that's a huge driver of both market information and volatility
as well. What happens if someday USDA is, for lack of a better
word, outmoded by private aggregators of that information that
have much better, much more rapid market intelligence? How
would that influence markets?
Senator Moran. Right. I mean, that's not a question I
thought of until I was hearing the panel today. Information is
valuable, but there can be a corner on the market, which then
changes the nature of the market.
Mr. Tatge. So I talk about this and debate this a lot. I
traded commodities for several years. The interesting thing is
that if a farmer opts-in to sharing their data or selling their
data, let's say licensing their data, to a commodity brokerage
house, they're getting their profit, they're taking the risk
off the table right away because they're getting a piece of it,
regardless of where the market goes.
So I would argue that the data flowing to the market
quicker is going to reduce volatility and give the farmers
additional reasons and ways to monetize that data that they
haven't probably thought of in the past.
The farmer is the first consumer of that data, if you will.
When you're driving through your combine, you're the first
consumer of that data. OK? So you might think, ``Oh, my yields
are much better than I thought. I should hurry up and sell
some.'' Right? He--that's a good use of that first--that first,
you know, consumer of that data. The problem is you're not
quite certain enough if it's just that field or if it's just
your area. But if we can have a larger pool of that data,
suddenly that becomes really relevant for you in your marketing
decisions as well as the market at a whole.
I absolutely have no--I have no doubt that getting
information--once you get something faster, you never go back.
OK? And I believe that will be the same in the markets and the
information that we're collecting as well.
Mr. Knopf. Mr. Chairman, I may--you've been a--you
understand regulation and how that impacts farming systems, and
I think that's one important note that hasn't been perhaps
mentioned, is we also need to recognize the risk of--as access
of data increases, I think we need to be cognizant of the risk
that that--the risk that that presents in utilizing that data
down to the farm level perhaps to regulate a farmer on specific
practices even though we're working within a biological system.
There are going to be large swings across years based on
rainfall, on how much nitrogen we use per bushel of wheat per
se. So this year is going to look very different than last
year. And when folks don't understand a biological system and
weather variation and just look at a dataset, it can be an easy
thing to immediately think, well, how can we get them to always
just use this much nitrogen per bushel of wheat? So I think
that's an important thing to think about as well.
Senator Moran. These markets, of course, are international.
And is the market for data international? Is the same thing
we're talking about here happening in Brazil and Argentina and
Europe?
Mr. Tatge. I think hypothetically, absolutely. There is
going to be a lot of different regulations and rules, and it's
going to take a while to understand how each country wants to
deal with their data that's collected. But I absolutely think
this is a global opportunity for sure.
Senator Moran. Let me just ask a few questions that I think
can be quick. Is my understanding correct that there really is
not an agency of the Federal Government that protects farmers
or regulates the collection of data? Is that the conclusion of
what I heard you all say?
Dr. Ferrell. That's correct, sir.
Senator Moran. And I guess I should ask the next question:
Should there be?
Dr. Haman?
Dr. Haman. I think that it is very important, one thing is
very important, that the data is standardized because the
closer the formats of the data, the possibility of comparing
that is extremely important. Right now, we are getting a lot of
data which is really not very good. And even farmers would
admit that the data, raw data, is something which they cannot
process.
I have mentioned that we are moving toward Internet of
Things or Internet of agriculture. At that point, we remove the
factor of human being in a lot of situations because machines
talk to each other and the computers talk to each other.
Sensors are talking directly to the computers.
So some kind of standardization will become reality anyway.
That's also true of the data which is coming from the
satellites because all of that has to be downscaled because
it's on a scale which is not really practical for farmers. But
there's a lot of work going on, on downscaling the data which
is coming from satellites.
But all these databases have to talk to each other, and
that's very difficult to do if there are a lot of small--it
seems to me if there are a lot of small entities, small
companies, which are organizing the data differently, then
there is a problem. Of course, sooner or later, the Microsoft
or an Apple is going to emerge and just take over. This is
probably what's going to happen, but the sooner we make this
data comparable, the easier it will be for everybody.
That also creates kind of a positive situation for very
small farmers and for farmers with specialty crops because the
big companies are not interested in data on specialty crops,
and this is where farmers don't have such large farms, but
there is a lot of very good data going on there.
And I just recently heard that in Brazil they are tagging
individual trees in the orchard and collect data on individual
trees. We are not doing that right now, but maybe somebody else
is doing. So the interest is all over the world. Germany is
investing a lot of money into data as well.
Senator Moran. Thank you very much.
Dr. Haman. Thank you.
Senator Moran. I can see USDA is the agency or the
department that would be involved in the standardization of
data.
Dr. Haman. Probably.
Senator Moran. That's a slightly different question than
where I was going, but an important one. And ultimately that
could be of great value to FSA and to RMA, risk management and
to farm programs, if the Federal Government actually had the
information in which they're basing payments, crop losses, or
crop production, standardized. My question, slightly different
than that, is no one is out there protecting the farmer.
And, Mr. Janzen, my understanding of what you do is related
to transparency, not related to privacy, not related to
standardization. You've got a--I mean, ``niche'' makes it sound
less important, but you have an area that you're dealing with
related to transparency, which is important, but nothing
broader than that, and no one else is doing what you're doing
in those other areas, right?
Mr. Janzen. That's essentially right. I would say we do
deal with privacy issues because the Transparency Evaluator
does ask questions that really get at privacy as well and
protecting that.
Senator Moran. OK.
Mr. Janzen. And going back to your question about
regulation, I think that the Transparency Evaluator really is
sort of an attempt at the industry to police itself so that
Congress doesn't have to step in. Now, for that to work, we
need full buy-in from the industry, and right now I would say
we have a good first step, but we need wider buy-in from other
players.
Senator Moran. When you indicated earlier that there is
uncertainty as to who owns the data, whose job, whose
responsibility is it to clear up this issue of data ownership?
Dr. Ferrell. Well, it's interesting, we've actually tried
to explore this. There is not really a way to trademark it,
patent it. You really can't copyright data because copyright
requires an element of creativity, and data is raw facts, which
we have a Supreme Court and President that says you cannot
copyright true facts.
So that leaves us in the realm of trade secret, and this is
something that Mr. Janzen and I have talked about quite a bit.
We think there's a colorable argument to be made that farm data
is a trade secret. If it is a trade secret, there are a lot of
strings that come with keeping it a trade secret. And trade
secret is also a function a state law, not Federal law. So you
run into some problems with that being consistent across
states.
Now, we do have the Uniform Trade Secret Act, which the
last time I had checked, 48 states had adopted, and the
remaining 2 states were considering it. But there might be an
advantage toward a Federal protection there to keep that
consistent across states and to allow, in a way, more
interoperability as well because we would have a consistent
application of those principles no matter where you were.
Senator Moran. Senator Blumenthal and witnesses and
audience, this is I think my last question, to give you hope.
[Laughter.]
Senator Moran. So my question was going to be about, what
are the risks and consequences of data breach, data security,
insecurity? And this question of about, ``Who owns data?''
causes me to ask, so if someone steals your data, is there a
question about whether that's a theft because who owns it? So
is it that uncertain?
Dr. Ferrell. There's a very valid question with that
because if you are claiming that your ag data is a trade
secret, number one, that means it has to derive independent
economic value from not being known. In other words, the data
actually has to be more valuable because people don't know it
rather than people are--you get value from it, but so does
someone else as well.
And the argument there is to say, you know, if I have data
about Mr. Knopf's farm, that doesn't do me any good on my farm
because I don't have his farm, I have mine. And so the one side
of that argument is that, well, so what if ag data gets
disclosed because that's data that's incredibly location-
specific to your farm, and it gets out there?
But the other argument is, well, if you know more about my
farm, you might be able to bid resources away from me. You
might go to my landlord and say, ``Wow, Ferrell is doing a
terrible job of farming your place, I could do a much better
job, and I've got the data to back it up.''
So there's an argument there that disclosing the data might
actually reduce its economic value. But the point of that is
it's really nebulous, and you would spend a lot of time trying
to make a very legal argument about whether or not you even had
that ownership in the absence of something more carefully
defining what ag data ownership really meant.
Senator Moran. So in today's world of cyber breaches, cyber
attacks, and data breaches, is there something we should worry
about in the realm of agricultural production data?
Mr. Janzen. It seems to me--oh, go ahead. I was going to
say it seems to me like the most likely candidate for hacking
into and obtaining disclosure of agricultural data from someone
that shouldn't have it would probably be some form of corporate
espionage or, you know, a competitor of one company thinking
that if they could get their hands on that information, it may
give them a leg-up in the marketplace. And, you know, I don't
think we would see that from a United States company, but it's
very possible that it could be from a foreign company that
wants to get their hands on information to give them a leg-up.
Senator Moran. I can see some production facilities who
would love to keep their information private.
Mr. Tatge. I think one of the big questions that's yet
unanswered is the companies that have data, and they've had it,
collected it for years, what happens to that data when those
companies are sold? That's the question that is looming out
there as far as if we've collected data for a long time, and we
go to sell our company, part of the value of that company is
based upon the data that you've collected. And did you obtain
it in ways that were described and really covered under some
form of agreement? And I would say for the most part today that
doesn't exist.
Senator Moran. Justin, Mr. Knopf, if you sold your farm in
the future, I don't imagine you have enough data at the moment
to make the data itself valuable, does the data go with the
farm, or that's a separate commodity?
Mr. Knopf. Well, that's a fair question. I haven't really
conceived that question.
Senator Blumenthal. Why is it not like any other property--
--
Mr. Knopf. I think----
Senator Blumenthal.--the grain that remains in the silo or
the gas in your tractor or----
Mr. Knopf. Yes.
Senator Blumenthal. Now, the interesting question, though,
is--and we haven't really talked about it--mentioning the word
``copyright,'' what about the software in your tractor? When
you sell your farm and you sell that equipment, you're also
selling the tractor, but the software belongs to GM or John
Deere or whoever made the machine, and you can get in trouble
if you start tinkering with the software even on your car, let
alone your tractor. So this is a common problem in terms of who
owns what when we're talking about Big Data.
Mr. Knopf. Yes, two questions there. Chairman Moran's
question, I think a lot of it is the value is still very much
being defined, and it's rapidly changing every day in the
marketplace because I think as we're--right now, farmers are
still struggling to know what the value is in that data and how
to use it, and there's a lot of personal intuition about
looking at that data from your farm that has to--that another
person from outside your farm that doesn't have the
understanding that you do of your farm is not going to see the
same value that you do in your data. But as models develop and
software develops to better automate that process, that data
will become more valuable and an independent value proposition.
And, Mr. Blumenthal, that's a very appropriate question,
and perhaps you all talked about this as the right--the right
to repair, and that's a whole other realm of, you know, when
our tractors break down with some--a sensor or a piece of
software that's running something mechanical, we get a code,
the tractor stops moving. We have to call out the service
repairman from the implement company that has the only access
to that software to be able to reset that code. There is no
third-party competition within the access of that right to
repair that specific code.
Now, I understand protecting intellectual property rights
and innovation and being able to collect and based on your--
recoup value and innovation and protecting that, but there is
some--there is also a balance there between at some point in
that machinery allowing third parties to enter that software
and to be able to access it for repairs as well.
Mr. Tatge. One example, just to add on to that, that I've
heard quite a bit is if you were to buy a piece of land in your
general area, how many years does it take you to get it to
perform the way it was performing for the individual before? On
average, I hear it's probably 3 or 4 years for you to really
get to know that piece of dirt.
When you think about that, when you go to resell or when
you go to market a piece of land, having that history and
giving the formula as to what was used before I think could
greatly allow the new acquirer to be able to bring that up to
the same production standards much faster than it would be on
their own.
Mr. Knopf. And we've always had that, you know, we've
always had soil samples and more kind of raw non-digital data
on our farms in the past, and that typically, depending on the
transaction, if it's a friendly transaction, sometimes it will
just be given over, but as that value increases, that becomes
more of a very appropriate question.
Senator Moran. Senator Blumenthal, I take no offense at you
indicating now this is the more interesting question, I guess
because truth is a defense.
[Laughter.]
Senator Moran. Thank you for an interesting question.
Senator Blumenthal. Thank you all for your participation
today. And I want to thank the Chairman for having this
hearing.
Senator Moran. Mr. Blumenthal, thank you very much. Thanks
for your participation.
The hearing record will remain open--let me--I always ask
this question, Does anyone, any witness, have anything they
want to say before we close today's hearing? Anything we
missed?
[No audible response.]
Senator Moran. Great. The hearing record will remain open
for 2 weeks. During this time, Senators are asked to submit any
questions for the record. Upon receipt, the witnesses are
requested to submit their written answers to the Committee as
soon as possible.
That concludes today's hearing. I thank you all for your
presence.
[Whereupon, at 4:07 p.m., the hearing was adjourned.]
A P P E N D I X
Prepared Statement of Hon. Bill Nelson, U.S. Senator from Florida
Thank you, Mr. Chairman, and thank you for holding this hearing to
discuss a topic that is certainly of paramount importance to the State
of Florida.
Mr. Chairman, agriculture is one of the most important pillars of
Florida's economy, with an annual economic impact of over $120 billion
dollars. We have over 47,000 farms and ranches, which employ over 2
million people. Many people know about Florida Oranges, which by the
way make up 60 percent of the total U.S. value, but in addition to
oranges, we are also ranked #1 in value of production for snap beans,
cucumbers, grapefruit, sugarcane, fresh market tomatoes, and
watermelons.
So I am very pleased to be holding a hearing that discusses how we
can make farming more productive, more efficient, and more sustainable
through the use of data. Human activity is drastically changing our
world's climate, and not for the better, and we will have to use data
and analysis to adapt agricultural practices and protect workers.
In Florida, farmers use precision agriculture to monitor the health
and development of beef cattle and dairy cows. They use it to assess
soil health in peanut and cotton fields.
And if a fascinating experiment continues to show promise, citrus
growers might soon be using lasers to inject pesticides directly into
the leaves of citrus trees. Last January I visited with researchers
from the University of Florida Citrus Research and Education Center in
Lake Alfred, to see this incredible technology in person. A laser makes
a tiny incision in the leaf of an infected citrus tree, and then a
machine sprays a bactericide directly into the laser cut to get it into
the tree's system more effectively.
Research like this couldn't come at a more crucial moment for the
citrus industry. Last week, the U.S. Department of Agriculture updated
its estimate for the 2017-2018 season, lowering the expected harvest to
50 million boxes. The initial prediction for this season was already a
decades-low 54 million boxes.
And that was before Hurricane Irma swept through the state in early
September. Farmers and ranchers are still trying to calculate total
losses and figure out how to salvage the season. The initial reports
are devastating: $760 million in losses for citrus growers, and $2.5
billion in total agriculture losses in Florida. I'm working very
closely with Senator Rubio and others to make sure that the next
supplemental package includes disaster aid for these farmers so that
they don't go out of business.
In the meantime, this hearing can explore how precision agriculture
is playing a role in our fight to find a cure for citrus greening,
canker, laurel wilt, and other pests and diseases.
Lastly, this committee has a long history of examining the issues
of ``big data'' and how it affects consumers and our economy. Whether
it's a consumer shopping online, or a farmer tending to his or her
fields, companies are collecting information and monetizing that
information for their purposes. Just as I have long advocated for
consumers when they are using their smartphone or shopping at the mall,
they should have a say in if and how their information is collected and
used.
Lastly, I'd like to welcome our distinguished witness panel today,
but I want to particularly recognize Dr. Dorota Haman (Door-OH-tah HAH-
man). Dr. Haman is Professor and Chair of the Department of Agriculture
and Biological Engineering at the University of Florida. She is a
renowned expert as an agricultural engineer in advanced irrigation
management and technologies. I look forward to hearing Dr. Haman and
the testimony from all of our witnesses today.
Mr. Chairman, I am excited to see the new ways in which
technological advances can increase agricultural productivity and
efficiency in my state and in the rest of the country. Thank you for
holding this important hearing.
______
Prepared Statement of Timothy Hassinger, President and CEO,
Lindsay Corporation
Thank you Subcommittee Chairman Moran, Ranking Member Blumenthal,
Senator Deb Fischer and all of the members of the subcommittee for this
opportunity to submit this written testimony.
My name is Tim Hassinger. I am the president and chief executive
officer of Nebraska-based Lindsay Corporation--a leading manufacturer
of center pivot and lateral move agricultural irrigation systems. For
more than 50 years, Lindsay Corporation has been at the forefront of
research and the development of products and services designed to meet
the world's rapidly growing agriculture and transportation needs.
As you may know, it's estimated that by 2050, the global demand for
food will be 60 percent higher than it is today. To meet this daunting
challenge, it's imperative that we develop and deploy technologies that
will help growers produce more with less, while preserving water and
other natural resources.
Data-driven farming is the key to allowing growers to meet the
food, fuel and fiber needs of the rapidly growing global population.
With the touch of a button or swipe of a finger, farmers who have
broadband access can:
Receive commodity price information;
Monitor and respond to changing weather conditions;
Use GPS for planting and irrigation management;
Get real time data on soil and moisture conditions;
Connect with other farmers and agriculture experts, and
Store and analyze data to increase sustainability and
productivity.
They can also take advantage of a myriad of new technologies now
available from Lindsay Corporation and other American manufacturers.
Among other things, these innovations enable remote data collection,
transfer and analysis from connected devices like soil moisture
sensors, weather stations and cloud-based support tools. Farmers are
using this information to streamline their operations, maximize
efficiency and increase productivity.
We work with farmers every day, so we know the power that comes
with the ability to leverage big data. We now offer technology that
helps farmers decide precisely when, where and how much to irrigate--
maximizing yields while reducing overwatering and related input costs
and nutrient losses.
In recently conducted field studies, our researchers found that
remote telemetry streamlined growing operations in several key ways,
including:
3 percent increase in corn yield (driving profit of $25 per
acre);
17 percent reduction in water usage (saving more than 9.25
million gallons on a 130 acre field);
$10/acre reduction in energy costs; and
75 percent reduction in time spent going back and forth to
the fields (another $5/acre saved).
This combination of yield enhancement and resource savings can
increase American farmers' profits by an average of $40 per acre--
profits that can be reinvested in their operation and in their local
economy.
Data-driven technologies allow growers to increase yields while
conserving water and other natural resources. These technologies are no
longer luxuries. Rather, they are critical tools needed to increase the
overall operational efficiency and productivity needed to complete in
the global marketplace.
______
Prepared Statement of Deere & Company
Deere & Company (``John Deere'') respectfully submits these
comments for the record as part of the Subcommittee's November 14, 2017
hearing on the subject of ``Technology in Agriculture: Data-Driven
Farming.''
John Deere is a global leader in the manufacture of agricultural,
construction, turf and forestry equipment. Deere provides advanced
agricultural and other equipment and services to customers that
cultivate, harvest, transform, enrich and build upon the land to meet
the world's dramatically increasing need for food, fuel, fiber and
infrastructure. Deere has been providing innovative equipment and
services to customers since 1837, and today is pioneering state-of-the-
art data and information solutions designed to greatly enhance
productivity and sustainability.
The Value of Data-Enabled Agriculture
John Deere believes that the growth of data-enabled agriculture is
as transformational today as was the introduction of mechanization to
the farm almost one hundred years ago. Insights generated from producer
data will be critical to meeting the goal to produce enough food and
build the infrastructure required to sustain a growing global
population. Properly used, producer data has the potential to greatly
improve precision, productivity, profitability, and sustainability on
the farm.
American farmers face constant pressure to improve efficiency,
environmental stewardship, and output. For this purpose, farmers look
to advanced smart farming technology solutions, including solutions
that take advantage of mobile and fixed broadband access. Today,
producers are able to farm to within a few centimeters of accuracy
thanks to innovative GPS-enabled positioning systems that are now
standard on virtually all modern farming equipment, as supplemented
with data available from satellite signals. Using these high precision
techniques, advanced agricultural equipment and services now include
technology that provides real-time agronomic data that can be analyzed
to optimize the precise amount of seed, fertilizer and pesticides
needed, reduce costs for fuel, labor, water, and identify best
practices for fields in a given location. (Deere's Precision Ag
Technologies, for instance, gives farmers access to detailed agronomic
information in the field essential for improved decision-making with
respect to managing costs and recourses.)
Where possible, producers use data and communication technologies
to interact with customers and vendors, follow commodity markets,
obtain real-time information on field conditions, weather and other
environmental factors, and manage fleets and regulatory compliance.
Farmers can also employ innovative machine-to-machine (``M2M'')
operations in the field and machine-to-farm (``M2F'') from the field
that enable producers to make significant improvements in real-time
productivity and cost management.
Today, these technologies are making an enormous contribution to
improved use of limited resources, regulatory compliance and ag
sustainability. Precision technologies are enabling more efficient,
prescriptive use of soils, water, fertilizer, herbicides and fuel by
allowing producers to tailor farming practices and applications to the
specific conditions of an individual field.
For example, when the farmer leaves his field in the fall, he is
able to share harvest yields directly and immediately with trusted
agronomist advisors. This helps the advisor to prescribe the
appropriate amount of nutrients to be added back to the soil, based
only on what the farmer took off at harvest, and ensure those nutrients
are added and incorporated before winter. The farmer can also make
decisions on which seeds to buy for next year, taking advantage of
early order price discounts. By reducing inputs, improving resource
management, minimizing land impacts and lowering costs, these
technologies are delivering the promise of sustainability on the farm.
The economic impact of these technologies is significant. According
to recent reports, data-driven decisions about irrigation,
fertilization and harvesting can increase corn farm profitability by $5
to $100 per acre, and a recent 6-month pilot study found precision
agriculture improved overall crop productivity by 15 percent.\1\
---------------------------------------------------------------------------
\1\ See Kurt Marko, Forbes, Precision Agriculture Eats Data, CPUC
Cycles: It's a Perfect Fit for Cloud Services (Aug. 25, 2015),
available at: http://www.forbes.com/sites/kurtmarko/2015/08/25/
precision-ag-cloud/.
---------------------------------------------------------------------------
The Importance of Data Privacy
In addition to offering a full line of innovative, high-quality
agricultural equipment to producer customers worldwide, John Deere
provides data and data application services that support customer
business needs and the optimal utilization of Deere machines. These
services are provided through Deere's proprietary data management
platform, the John Deere Operations Center.
John Deere believes that all involved in the generation and use of
data and data services should have effective processes in place to
ensure privacy, security and ultimate control for the producer. Deere
has been actively engaged with individual customers, grower
organizations, ag service providers, agronomists and many others to
develop practices and processes that ensure producer privacy and
control, while making data processing, analysis, and use as seamless as
possible. Deere believes that the market participants across this value
chain--through collaboration, private agreement and mutual trust--are
best able to develop and implement the necessary practices and
protocols that protect producers and serve commercial needs. To this
end, Deere has developed a set of business data principles that govern
its use of production data, machine data and administrative, and are
incorporated into every customer's John Deere Operations Center
services contract. These principles are designed to give the customer
the ability to control whether and how his/her data can be used, by
whom, and for how long.
These principles are:
1. Deere provides data and end-user application services for one
reason: to support the business needs of its producer customers
and improve the use of Deere equipment and technologies.
2. The producer's data should be differentiated into machine,
production, and administrative data, and each data subset
should be managed in accord with these important distinctions.
3. Deere utilizes customer data only with the customer's consent, in
order to improve grower productivity and profitability, and to
optimize the utilization of John Deere products and services in
the customer's farming operations.
4. The producer customer retains control of his business data
including whether, what and how his data is used and shared.
The customer may withdraw this consent or request that data be
deleted from his account at any time.
5. Any disclosure of explicit customer business data is determined
solely by the customer's designated account preferences and
through contractual agreements with John Deere.
Farming is a complex, dynamic industry and this makes the notion of
farm data ownership complex as well. Farmers use Deere's tools and
offerings in many different ways, which further complicates the issue
of ownership. Expectations, relationships, contracts and laws regarding
data control and ownership vary from place to place, operation to
operation and even on a single farm. Companies that assert that farmers
own their data are not being transparent--it is not as simple as that,
which is why John Deere is focused on efforts to ensure that farmers
control their data.
Real-world circumstances that make data ownership complicated and
uncertain include:
Custom harvesters or equipment operators who may have the
right to share production data.
Landlord and/or tenants who may have the right to share some
or all production data from a farm.
Agronomists and other consultants who may have the right to
share data.
A farmer who may buy licenses to use commercial prescription
files, other technologies, or seed hybrids that the farmer does
not own.
This is why Deere believes that customer control of the data is the
most important issue. Deere's data management services and applications
are designed to ensure customer control of their farm's data.
There are important distinctions between the types of data that are
generated through integrated ag technologies, and Deere and its
customers agree to manage these differentiated data sets accordingly.
John Deere segregates customer data into three subsets--Machine Data,
Production Data, and Administrative Data.
Machine Data are data that generally relate to how equipment
is functioning (fuel consumption, vehicle diagnostic, engine
performance). This data may be utilized, with the customer's
consent, in original or aggregated, anonymized form to
proactively address equipment issues and improve the customer's
experience with the machine.
Production Data relate to the work being performed by the
customer, and enable Deere to administer services the customer
has opted into, such as field tasks, location history or
wireless data transfer. By using our systems, customers agree
to allow Deere to create aggregated and anonymized Production
Data sets. These anonymized data sets are proprietary to John
Deere. John Deere is free to use and disclose the anonymized
data, and John Deere may promote information and services
derived from anonymized data. Anonymized data is never
traceable back to a specific customer.
Administrative Data is information that helps Deere support
a customer's account and activities in our system. Examples of
administrative data are: data sharing permissions, account
users, machines and licenses connected to accounts, acres or
file sizes.
These distinctions are a critical part of the data management
process. They preserve customer control while distinguishing the
sensitivities associated with certain data sets. They are reflected in
the contractual agreements between John Deere and its customers.
It should be noted that the marketplace for technology around data
collection, transmission, storage and use is evolving rapidly and will
continue to evolve in the years to come. Producers will continue to be
presented with new options and product offerings that can deliver even
greater value, while rewarding the most innovative technology and
service providers at the same time. This can best happen through the
collaborative private sector efforts of market participants, without
the specter of more rigid standards or codes imposed from outside that
could stifle innovation. The Ag Data Transparency Evaluator, created
and managed in conjunction with the American Farm Bureau Federation, is
a good example of this private sector effort. John Deere played an
integral part in the creation and implementation of this tool. We are
actively working with this group to achieve its ``seal of approval.''
Initially, discussions around the requirement to recognize farmers'
ownership, rather than control, of data slowed our progress. After much
discussion, Deere believes we have greater alignment with this group
and look forward to adding our name to the list of companies that have
gone through the transparency process.
Finally, it should also be noted that, without essential broadband
connectivity in croplands, many of the potential benefits of data-
driven agriculture can never be realized. Real-time ag services using
data generated on the farm are dependent on reliable, high-speed wired
and wireless connections to the Internet--connections that in turn
depend on a robust rural broadband infrastructure that is currently
lacking in many parts of the country. More attention must be given at
the Federal level to ensure that the build-out of wireless broadband
infrastructure, including connectivity in the fields where farmers and
equipment operate, is achieved.
At the heart of John Deere's efforts and principles around data-
enabled agriculture lies our history of, and commitment to, helping
those linked to the land. Since 1837, John Deere has been building
lasting relationships with agricultural producers based on our core
values of integrity, quality, innovation, and commitment. Deere
believes that the trust we have established with producers, built up
over these 180 years, is exponentially more important than the value we
might derive from producer data.
Deere & Company appreciates this opportunity to share its views,
and looks forward to working with the Subcommittee on these important
issues.
______
Response to Written Questions Submitted by Hon. Jerry Moran to
Justin Knopf
Question 1. Your testimony described the usage of data as a
``driver of economical sustainability and environmental stewardship.''
As a Kansas farmer focused on both being a good steward of the land and
making a living to provide for your family, could you please further
describe your efforts to balance sustainable farming practices with
improving efficiencies to increase profits? Can the two goals go hand
in hand?
Answer. Yes, often the two goals can go hand in hand. I believe
most ``sustainable farming practices'' will have positive impacts on
profit in the long term. The challenge is ``the long term'' might be
ten years, or twenty, or a lifetime. There are numerous examples of
sustainable farming practices which evidence would point to having a
long term positive impact on profit, yet be a possible net cost in the
short term. These types of practices will tend to have slower adoption
curves. An example of one of these practices on our farm would be cover
crops. They are crops planted between our main grain crops solely
intended to provide environmental benefits to soil protection and
health. The seed and investment in time and machinery to plant them is
a significant cost and we have yet to document a yield or economic
benefit to the subsequent grain crop. However, evidence and agronomic
principles predict that across time, the environmental and biological
benefits from cover crops in our climate and soils will improve the
resiliency of our farm and perhaps the productivity, and therefore
profit. In the meantime, we utilize data from both on farm research
trials and field scale to evaluate what cover crops in which part of
our cropping sequence will have the most impact with the least amount
of cost. We then start with limited acreage and hopefully scale the
practice to broader implementation as we learn and begin to reap some
benefits over time.
There are other examples of sustainable farming practices that will
also improve efficiencies to increase profit as well in the short term.
An example from our farm is zone management. We utilize various sources
of data, typically multi-year yield maps, satellite imagery, and soil
maps, to divide a field into zones based on productivity. Then, the
more productive zones of the field receive the right amount of
fertilizer to sustain that productivity while fertilizer rates are cut
on less productive zones. Right away, we have improved the efficiency
of our fertilizer which increases our return on that investment, plus
reduces fertilizer carryover and loss into the environment.
Question 2. How does data collection and sharing specifically
assist farmers in striking the appropriate balance, including
innovations in live-time monitoring of crops and measurements of
surrounding conditions?
Answer. Data collection and sharing improves our understanding of
how crops are impacted by certain factors such as weather events,
management decisions, soil types, etc. Obviously some of these factors
are outside of the farmer's control, but access to improved live-time
monitoring of crops and measurement of surrounding conditions can help
farmers be more proactive in predicting crop response to these factors
or events and lead to more timely and improved decision making. One
quick example, this past summer we had a summer hail storm that
significantly reduced soybean stands in its path. We utilized satellite
images of the impacted fields that were available several days after
the storm to fine tune our scouting and decision making about where to
replant and where we could salvage the stand of soybeans.
Question 3. Your testimony divides the data that your farm
specifically uses into three categories: microdata, service provider
data, and macrodata. Will you please describe how farmers and their
operations benefit from each category of data?
Answer. Microdata-this is data a farmer collects from his own
operation and is specific to his operation. This likely helps better
characterize specific aspects or management factors unique to his farm,
leading to improved decision making.
Service provider data-this is data that is produced by a service
provider outside of the farm. Likely, there will be data from the farm
shared with or collected by the service provider, but then the service
provider will utilize data from that farm and perhaps integrate it with
data from other farms and/or a proprietary algorithm or internal data
set to provide analyzed data back to the farm in order to help with
improved decision making by the farmer.
Macrodata-this is ``big data'' collected from many farms likely
across a broad geography. The farmer may or may not have contributed
data from his farm, however, there are insights gained from the sheer
volume of data that may not be possible if the data set was not so
large. The insights may be more universal in nature yet still applied
by many farmers to improve decision making.
Question 4. In a 2016 poll conducted by the American Farm Bureau
Federation, regarding the loss of control over downstream uses of data,
sixty-six percent of the farmers polled expressed concern about not
being compensated for the potential benefits from the use of their data
beyond the direct value they may realize on their farm. Meanwhile,
sixty-one percent of the farmers were concerned that agricultural
technology providers (ATPs) could use their data to influence market
decisions. Which of the two concerns do you believe is the greatest
threat to farmer profitability and well-being, and what should be done
to alleviate these concerns?
Answer. I don't know which of these two concerns is the greatest
threat to the farmer. As the statistics indicate, both are of
significant concern to many farmers. Farmers are accustomed to dealing
with concrete things we can put our hands on-tractors, soil, grain.
Data is very abstract and therefore more difficult for farmers to
quantify the value of, although most of us certainly recognize it does
have value. As was mentioned in the hearing, I believe one of the most
important steps to reducing the threat of non-compensation is
transparency and understandable communication up front before data
transactions and agreements take place. It is important the farmer can
quickly and easily understand what is happening with his data and the
parameters of any data agreement he is considering. I also believe it's
important to recognize that farmers may be compensated for their data
in forms other than money. Compensation may be access to insights
gleaned from the larger data set they are contributing to, or access to
a proprietary decision making tool.
I believe most would agree that there is power in data and
recognize the consolidation in agricultural companies, which is likely
why farmers feel concern about their data being used to influence
market decisions. Consolidation and concentration of data is perhaps
something that should be monitored.
Question 5. With connectivity being crucial to the successful
implementation of the technology we have discussed today and almost 30
percent of farms not receiving adequate broadband connection according
to the USDA's Farm Computer Usage and Ownership August 2017 report,
what role can this Committee play in closing the gap to make sure all
of our farms are able to benefit from broadband and innovative
technologies? Do you see a role for advanced wireless networks in
achieving that goal?
Answer. Our farm is one of the 30 percent not receiving adequate
broadband connection. We do have a broadband connection, but to date it
is slow, not very reliable, and with only one provider choice, there is
no competition to drive improvement. Nearly all data management
software has now become web based so as the amount of data on the farm
needed to be uploaded and downloaded exponentially increases, effective
utilization of data and implementation of technology becomes impossible
without reliable and high speed data transfer technology. My time as a
farmer is very limited as it is, especially during the growing season,
so I cannot afford to sit and wait on a slow data connection. There
have been numerous times I've had to abandon a project because of slow
data transfer. However, I am hopeful as a local communications company
has undertaken the project of running new fiber optics to rural
residences and farms in our community. This will help with effective
data transfer from our farm office, but it will take advanced wireless
networks to achieve this goal in the field from mobile devices. It is
my hope this committee would have a renewed commitment to learning
where these gaps still exist and assisting small local companies, such
as Home Communications, Inc. in Galva, Kansas, along with wireless
network providers in closing those access gaps to fast and reliable
data networks.
______
Response to Written Questions Submitted by Hon. Catherine Cortez Masto
to Justin Knopf
Question 1. Obviously in the west we have some greater agriculture
challenges than other regions, and wildfires are one of them that we've
seen have detrimental effects over and over again. In many cases, these
fires have a multiplier effect on production and for things like future
flooding. And, in general, water is also another constant challenge.
Are there technologies that can help measure and account for drought
conditions, or measure the volatility of fires within parched forests
or grasslands before we have fires that potentially get out of control?
Are there any associations with improving broader weather prediction or
forecasts? And are there other specifics where you foresee this data
and technologies helping us get a better handle on our climate change
crisis?
Answer. It has and continues to be heart breaking to see and hear
reports from the west on the devastation caused by these wildfires.
We've had devastation in Kansas as well from wildfires recently.
Thousands upon thousands of acres of grassland have been burned,
countless cattle and miles of fencing lost, and homes and barns that
have been passed down through families on the prairie for generations
destroyed. There are certainly technologies that can provide data to
help measure and account for drought and predict risk of fire. While
I'm certainly not the best person to speak to these technologies, I can
give you an example from Kansas that is helping. Since 1986, Kansas
State University has managed a network of high tech weather stations,
called the Kansas Mesonet. Through recent efforts to grow the number of
stations in the network, they now have coverage across the state. These
stations are not only monitoring and recording typical weather data,
but also soil moisture, which is of course helpful in characterizing
drought, and when used in conjunction with relative humidity and wind
speed, has been a helpful proactive warning for high risk of grassland
fires. I believe increased number of weather stations such as these
that are available for farmers, ranchers, emergency preparedness
personnel, and the public to access will hopefully help quantify and
characterize more extreme weather, in turn leading to increased
preparedness.
There are technologies being implemented on farms that are scalable
enough to help mitigate climate change. Two specific examples are
improved fertilizer management through using the right source, rate,
time, and place for plant nutrient applications, which significantly
reduces Nitrous Oxide emissions, and also no-till farming practices
which sequesters Carbon Dioxide from the atmosphere in the soil. With
the help of Dr. Charles Rice, soil microbiologist at Kansas State
University and world-renowned soil Carbon expert, I recently calculated
that through building soil organic matter by implementing a no-till
based cropping system, our family farm has offset the average Carbon
emissions for roughly 4000 Americans. The soil is one of the biggest
Carbon sinks on the planet and practices such as no-till that increase
the organic matter of the soil not only offsets Carbon Dioxide, but
also improves the ability of the soil to capture and hold water,
reduces erosion, and allows soil biology to thrive, all leading to a
more productive and resilient system.
Question 2. There were many references to the environmental
benefits of agricultural data. Are we in a position yet where we can
authoritatively quantify the environmental benefits experienced by the
use and attention to these technologies and data analysis? For example,
is this science proven to the point that we should be creating
incentives in the farm bill conservation title for their utilization to
keep pristine watersheds like Lake Tahoe, or water quantity in drought
areas, solidified for the decades to come? Or is there a place for this
use in connection with the Federal crop insurance program?
Answer. Yes, we are now able to much better quantify environmental
benefits experienced through utilization of technology, data, improved
decision making, and effective conservation practices. However, the
challenge becomes as geography, climate, soils, and numerous other
things change from state to state, region to region, and even farm to
farm, the outcomes and impacts of practices will vary widely.
Therefore, the ``right'' decision for meaningful and long lasting
environmental impact on my farm may be the ``wrong'' decision for a
farm 100 miles west of me. Farmers are operating within an incredibly
complex biological system with an infinite number of relationships all
impacting each other. This is why it becomes so difficult to legislate
effective change to a biological system. That being said, I do believe
incentives can be incredibly effective at driving long lasting and
meaningful change if they are flexible and tailored to a local level by
local experts and advisors whom farmers trust. I have personally
utilized the Conservation Stewardship Program in the farm bill to help
offset some of the risk and short term costs in utilizing cover crops,
which has allowed me to implement the practice on more acres. And I do
believe crop insurance could be an effective avenue to offer incentives
for implementing conservation practices, but they must be practices
that are proven to be effective, economic, and increase reliance at a
local level. As you know, the devil is always in the details.
______
Response to Written Questions Submitted by Hon. Jerry Moran to
Dr. Shannon Ferrell
Question 1. In your testimony, you state that as participation in
the data community increases to a critical mass, farmers' bargaining
power with the data service providers likely will be greatly reduced
and a majority of the value will be enjoyed by the providers. You then
state that for farmers to take maximum economic advantage of Big Data
tools, large numbers of farmers must ``buy-in'' and participate in the
data community. Where is most of this value enjoyed by the data service
providers derived from? What can be done to mitigate the disparate
levels of value received, especially for producers?
Answer. Witness response to Question 1, part A: Where is most of
this value enjoyed by the data service providers derived from?
Before discussing the source of the value enjoyed by data service
providers through data compilation and deployment of Big Data tools, it
is important to discuss the value of data at the farm level in the form
of Small Data. At the individual farm level, farmers may take advantage
of today's data collection and analysis tools to run on-farm
experiments with respect to seed varieties, fertilizer applications,
moisture management, and so on. They can also use data to calculate
crop shares for rent, look for improved efficiencies in equipment,
management, conservation practices, and so on. In these uses, farmers
capture all of the data's value on-farm in the form of increased
returns or reduced costs.
While these ``micro'' level benefits can be considerable for the
individual producer, there can also be significant value derived
through the use of Big Data systems and analytical techniques at the
``macro'' level when data from hundreds or even thousands of operations
are aggregated. As mentioned in the written testimony, Big Data
analytics can be used to much more rapidly develop hybrids by running
trials across multiple soils and environmental conditions, developing
more accurate and robust models for predicting risk factors such as
weather patterns and production numbers, and development of improved
agricultural equipment. All of these items can be derived from the
aggregation of data from farmers and provide value back to those
farmers.
When one examines the potential economic values of data for the
data service providers, there are five primary sources to consider:
(1) Data service providers can derive revenue from the services they
provide directly to the farmer that provided the data. This
could be in the form of service fees for things like data
collection and validation, creating a repository of the
farmer's data that can be easily shared with other parties to
whom the farmer would like to provide access (such as a crop
consultant, landlord, etc.), or in the case of a data service
provider that also serves another role such as a crop
consultant, providing reports, prescriptions, or
recommendations to the farmer based upon the data.
(2) The data service provider could derive revenue from using the
data to market goods or services to the farmer. For example, if
the data service provider is a subsidiary of a seed company,
the farmer's data could help the seed company make seed
recommendations that are a good fit for the farmer's operation.
This form of focused marketing can sometimes benefit the farmer
as well. For example, consider purchasing a mixing bowl for
your kitchen on Amazon. Based on that purchase, Amazon suggests
a whisk and baking sheet that are commonly purchased with your
mixing bowl that have gotten favorable reviews from people who
purchased all three items simultaneously. If you also make the
suggested purchases and enjoy them, it is a ``win-win''
transaction for both you and Amazon. Similarly, purchasing a
seed variety that performs well for your farm and increases
profits creates a win-win for you and the seed company.
However, companies can also use a farmer's data to provide
recommendations while extracting more profit from the producer.
As another example, seed companies already use information
about producers to know which seed varieties are better suited
to the farmer's land and charge them more for that variety than
they would charge another producer. With increasing access to
producer data, input suppliers could continue to derive more
precise information about a producers willingness to pay (or
ability to pay) for their inputs and adjust their pricing
accordingly.
(3) Data service providers might provide data products or services
based on farmers' data to other companies. For example, data
service providers might sell reports or predictive models to
insurance underwriters might help them price crop insurance
products. These reports or models might be derived from
farmers' data but their sale would not necessarily involve the
transfer of the farmers' data.
(4) Data service providers can function as data aggregators and then
sell the farmer's data to third parties, deriving revenue from
those sales. In some cases, such data service providers may pay
the farmer for their data, but in other cases they may charge
the farmer a fee for their data collection services (while
providing some analytical services or reports back to the
farmer) and also derive a fee from the sale of the data to
other entities.
(5) Eventually, if and when enough farms join data networks, a fifth
use could come of that data--use of that information for
significant transactions in commodities markets. A hypothetical
example would be a data service provider who had access to a
sufficiently large sample of farms to make accurate predictions
of eventual U.S. crop yield who then takes positions in the
commodities markets well before anyone else would be able to
access that information.
A potential sixth value source is from the data service provider
positioning itself as an acquisition target with the purchasing company
getting either the data it holds and/or the subscription relationships
with the farmers it serves. Economic theory and historical precedent
both suggest that we will see an evolution in the agricultural data
industry starting with a large number of service providers vying to
engage farmers because, as Metcalfe's law suggests, the value of their
data networks will increase as a square of the number of their network
participants. Better-capitalized firms or firms with another
competitive advantage will acquire other firms until eventually only a
handful of dominant service providers--or even a singular monopolistic
provider--emerges. In the course of this evolution, the more farmers
and data a company can acquire, the more attractive they become as an
acquisition target. While some firms are certainly pursing the strategy
of becoming one of the dominant providers, it is equally certain that
other firms are seeking simply to be acquired.
Witness response to Question 1, part B: What can be done to
mitigate the disparate levels of value received, especially for
producers?
Answer. Research continues to determine both the value of
agricultural data in the aggregate and what proportion of that value is
captured by the farmer relative to others in the value chain. As with
the current USDA estimate that farmers capture 15.6 cents of the food
dollar, it is likely farmers will not capture a large proportion of
data values since they are relatively small, ``atomistic'' players in
the market with little bargaining power and face significant barriers
to the kind of collective action necessary to increase that bargaining
power.
Having said that, farmers' bargaining power may be at a maximum
right now. As mentioned in the response to Part A of Question 1, most
data service providers recognize they are in a race to acquire access
to the data of as many farmers and their acres as they can, as quickly
as possible. Some are approaching this with the strategy of a
telemarketer telling a prospect ``sign up today, because this offer
will be gone tomorrow!'' However, farmers should be thinking like
economists, and carefully weighing the benefits presented by any
particular service provider with the value they can receive from that
provider's services (or, indeed, the payment the provider is offering
to secure the farmer's data). To that end, farmers should ask five
questions of any prospective data service provider:
(1) How many growers/farms/fields/acres are in the data service
provider's data community? The higher the number, the greater
the value it can potentially provide to the individual farmer.
(2) What analytics conducted on the community will benefit my farm?
This aims at the direct ability of the data service provider to
increase the farmer's ability to make profitable decisions,
regardless of any external benefits.
(3) What data quality control standards are being used? If the data
service provider is not taking strong measures to ensure the
quality of the data in their community, it cannot provide
reliable insights to the producer. To quote the age-old
computer principle: ``garbage in, garbage out.''
(4) What uses will be made of my data? This question has a number of
implications discussed in the responses to other questions
below, but here its purpose is to help the producer gauge what
potential values the data service provider may be trying to
capture that will not directly benefit the producer (or may
actually increase costs for the producer).
(5) What assurances can the data service provider make that the
farmer's data will not be provided to third parties without the
ability of the producer to share in the revenues from that
transaction? Consider the analogy of a farmer giving an
unrestricted easement to an oil pipeline company. The farmer
may get a payment for the easement, but the oil pipeline
company may be able to sell co-located easements to a natural
gas company, a telephone company, a fiber optic company, and so
on, without the farmer having any ability to capture the value
realized from those transactions. Similarly, if a farmer does
not have an agreement restricting ``downstream'' uses of his or
her data, they only have one opportunity to capture the value
of that data.
While farmer's negotiating power may be at a peak, it may also be a
matter of timing to make sure farmers do not completely miss out on the
ability to capture the value of their data. As the industry continues
to evolve, we will likely see a progression going from a) farmers
paying data service providers for their services, b) data service
providers realizing they need more and more farmers in their network
and thus reducing their costs, potentially to nothing, to join, c)
companies actually paying farmers for their data, but then d) companies
securing a critical mass of farms or acres to have sufficiently robust
data networks that they no longer need additional farmers to join.
In the end, the most effective means of helping farmers secure the
maximum amount of their data's value may be educational efforts to help
them determine the value of their data, evaluate data service providers
and their agreements, and make informed choices about their data
sharing relationships.
Question 2. You note the irony in the growth of proprietary data
network protocols that lead to complaints about the lack of
interoperability of farm equipment systems also providing greater
protection against data breaches. What measures can be taken to
continue improving the interoperability of data collection without
sacrificing security?
Answer. Witness response to Question 2: There are two primary
reasons computer viruses are far more problematic for computers using a
Windows operating system than an Apple OS system: first, far more
machines use Windows, and second, the nature of Windows architecture
permits more access points to its code than Apple OS. At the same time,
one could also make the argument that all computer users have
benefitted from consolidation in the computer industry in that so many
programs are now interoperable and data can be shared among what is now
billions of other users with relatively little friction. By analogy,
farmers and data service providers may both benefit from consolidation
in the industry and the increased interoperability it provides.
Such consolidation may also ``consolidate'' security concerns as
more eggs are kept in fewer baskets. This is the Windows problem
again--since there are far more Windows-based computers in the world,
virus writers devote more resources to viruses that target them. Thus,
as consolidation continues in the agricultural data sector, increased
research and development efforts will be needed to make sure that these
``fewer baskets'' are guarded by increasingly robust security tools.
Congressional support of these research efforts would benefit not only
agricultural producers but a vast number of Americans who rely on data
security to keep their personal information secure.
The increasing automation of agricultural data collection and
transmission tasks might actually serve as a means of increasing data
security. The second reason Windows computers suffer from so many
security issues is the Windows operating system was not built with
security as one of its primary concerns, since its foundations were
built before networked computing was a primary use of PCs. As a result,
there are far more access points to its code that virus writers can use
to insert malicious instructions. Conversely, there are fewer access
points in Apple OS--a system built from the ground up for networked,
multi-user applications that thus requires explicit user permission for
code to be activated. The analogy in agricultural data is that the
tighter integration of data networks across agricultural equipment
creates fewer intervention opportunities for third parties. The market
will likely continue to drive this integration. With that said, the
issue of user's data security needs to be continuously brought to the
attention of hardware and software developers so they can keep security
as a foundational principle in their designs.
Another tool that might aid the security of farm data networks may
not seem to be directly connected to the security issue, but certainly
is: the expanded development of broadband cellular networks in rural
areas. Cellular transmissions are encrypted by the nature of the
processes that make cellular communication work; this has the most
beneficial side effect of adding to the security of the transmission.
However, in many rural areas, sufficient cellular signal or bandwidth
is not available to make use of many current agricultural data
technologies, requiring farmers to manually download and transmit their
data using much less secure methods. Expansion of rural broadband
cellular access would have the dual benefit of making agricultural data
tools more accessible and profitable for farmers while also reducing
the security risks associated with collecting and transmitting
agricultural data.
Question 3. In your testimony, you note that the current legal
framework for ownership of agricultural data is inadequate for the
transfer and aggregation of agricultural data. Is the agricultural data
space unique enough to require specific legislation regarding ownership
and property rights, or is a novel combination of existing property
ownership laws more appropriate and adequate?
Answer. Witness response to Question 3: One of the first questions
farmers always ask about their data is ``do I own it?'' It is a natural
question to ask, since farmers depend on access to land whether it is
owned or leased, and thus are closely attuned to property rights.
However, traditional notions of ownership break down to some extent
with agricultural data. Thus, the better question for the farmer to ask
may be ``what rights do I have with respect to my data, and what rights
to others have with respect to it. The most critical element of this
may be what rights (and abilities) does the farmer have to exclude
others from access to the data.
If one thinks about it, the notion of ``privacy'' is really a
function of one's ability to exclude others from access to information.
For example, HIPAA's provisions regarding health information naturally
couples with the notion of privacy in one's health matters, as they
pertain to matters of one's own body. Thus, HIPAA provides a legal
right to exclude others from access to health information without
explicit consent for the disclosure of that data. The Fair Credit
Reporting Act (FCRA) covers matters of financial information, which
also couples with traditional notions of privacy and the right to
exclude others from access to one's financial records.
However, the barriers start to blur a bit with the FCRA in that
financial transactions mean reaching out to and communicating with
another party; we recognize that others may have some limited right to
access our financial information if that information is relevant to
their financial risk in some respect, such as loaning us money.
Therefore, we allow credit bureaus to collect financial information and
to disclose that information to others so they may make credit
decisions about us.
The credit reporting analogy is important to understanding whether
we need a specific legal framework for agricultural data. Whenever
someone requests credit, they are required to ask if the potential
borrower will give consent to the lender to access a credit report.
Thus, for a third party to make use of the financial data, they must
have the consent of the person about which the data is collected.
Conversely, though, in many farm data agreements, the farmer may not
have the right to approve or deny access to their data to
``downstream'' users.
One could argue that agricultural data is significantly different
from HIPAA-protected health information (about the workings and
condition of one's own body) or financial data, but one should also
consider the fact that farmers lack the ability to protect their
information from disclosure. Put another way, it would be impossible
for farmers to try and keep confidential much of their production
information and practices because they can literally be seen from
aerial or satellite imagery. With relatively little effort one can tell
how many acres a particular farmer owns and what proportion of it is in
what crop for a given year (compare this to a bicycle maker or a coffee
shop--it would be extraordinarily difficult to determine the volume or
nature of their business, or to even tell one business from the other
without continuous monitoring of the business at high image
resolutions, simply because their businesses have roofs and farms
cannot). It could be argued then that perhaps farmers should be given
more protections than other businesses because one could derive a
significant amount of financial information about them from publically
available resources, to say nothing of the improved ability to do so if
one coupled data sources from a data service provider that transferred
that information without the farmer's consent.
However, if one did desire to provide enhanced protections for
agricultural data and allow farmers to exclude others from their data
without explicit consent, two significant barriers loom. First, one
would have to define the type of agricultural data subject to the
protection. This would be challenging, given the broad diversity of
data that can be collected and transmitted on farms today. Second, it
would be difficult to define when such consent would be needed. Would
the farmer have to give consent to any data transfer to another party?
There would be significant transactional costs in such an approach.
Further, there are doubtlessly data uses that will be available in the
near future that might not even be conceivable today; it would be quite
challenging to give informed consent in an up-front data use agreement
when one doesn't know what data uses might be possible in the future.
The current legal framework might be serviceable as an interim tool
to help provide farmers some grounds for the excludability of
agricultural data, and enhancements to that framework may be possible.
In the near term, though, perhaps one way to help farmers maintain
control of their data is additional research into encryption algorithms
that give farmers a key that would be required to access the
information--this would put more control over downstream uses back in
the hands of the farmers, and also give them an increased ability to
participate in the value received for data transactions.
Question 4. As more and more firms enter the agriculture-technology
space and interact with data used by and/or generated by farmers, the
need for clarity and consistency on privacy principles is growing. For
these new entrants, can you suggest any best practices these firms
should engage upon to ensure their data privacy procedures properly
convey the data's expected use?
Answer. Witness response to Question 4: Much use is made in the
agricultural data industry of the word ``transparency'' but there can
often be much ambiguity in what that term means. The greatest value of
that term, in the witness' opinion, is to err on the side of disclosure
to the farmer when discussing the internal and external uses that the
data service provider will be making of the farmer's data. Those uses
should be disclosed clearly in language that is understandable by
farmers with a wide range of experiences and educational backgrounds.
One such example may be a Truth in Lending Act (TILA) disclosure.
Though an agricultural data use agreement might not bear a clear
analogy to a lending transaction, TILA makes clear the potential
impacts of the lending transaction and the borrower's rights and
remedies. Data service providers could benefit from making sure their
data use agreements have similar levels of clarity.
Another principle beyond the clarity of the disclosure is its
frequency. Using another financial analogy, individuals can use credit
monitoring services to receive notifications when someone makes a
credit inquiry about them. Data service providers could also provide
notices when an external entity has made a request to access the
farmer's data, or when a new internal use is made of the data. Robust
notification procedures can also help farmers take protective actions
in the event of a data breach.
As mentioned in the response to Question 3, there arises the issue
of informed consent when a new data use arises that was not
contemplated by the original data use agreement. Though it might
increase transactional costs, the simple answer to this problem is to
require disclosure of a potential new use and secure the farmer's
consent to the use before it is implemented. The counterpoint to this
approach, however, is that its increased transactional costs might make
companies implementing it less competitive than those who do not.
Finally, as new companies enter the agricultural data sector, they
would do well to avail themselves of the efforts of farm groups,
existing data service providers, and equipment manufacturers to develop
consensus on the principles that should govern agricultural data
management. The Privacy and Security Principles Farm Data developed by
the American Farm Bureau Federation and the Ag Data Transparency
Evaluator are two good starts for companies to use in developing their
operating policies and procedures. Both of these tools continue to
develop, and the dialogue can provide greater benefits to the
agricultural industry with increased participation from more farmers
and data firms.
Question 5. While much of the data we discussed in the hearing is
generated on farm and captured by farmers or their equipment,
significant quantities of data is publicly available and critically
important to inform risk modeling, yield prediction, etc. in both the
public and private sector. How can we encourage the continued use of
this type of data, and even grow our sources, while ensuring that
farmers understand their role in this process?
Answer. Witness response to Question 5: Perhaps the best steps that
can be made toward this goal are to continue funding of research and
extension efforts through our Land Grant universities to help producers
understand the value of the data resources to their decision making
processes. For example, the witness is currently the principal
investigator on a Southern Risk Management Education Center grant
funded through USDA-NIFA to develop a handbook and decision tools that
can help producers understand the value of agricultural data tools and
help them make informed choices about their uses (SRMEC Agreement
21667-19).
Additional research on how agricultural data systems can be made
more robust, reliable, and accurate can also add to the volume and
quality of publically-available data. For example, the most commonly-
logged seed variety on planter data systems is the variety that comes
first alphabetically on the system's drop down list. This means that
producers sometimes inadvertently (although potentially carelessly or
intentionally) select data inputs that are inaccurate, which in turn
affects all downstream uses of their data. Research of tools to help
improve data accuracy will not only increase profitability for
producers, but will also improve the data and decision tools available
to the industry.
______
Response to Written Questions Submitted by Hon. Catherine Cortez Masto
to Dr. Shannon Ferrell
Question 1. Ranching is a part of the fabric of rural Nevada's
frontier. We have operations under generations of family management.
While we heard many of the virtues of row crop farming and big data
during the hearing, can you outline the advancements and prospects for
animal agriculture? Is there also an advantage of agricultural data in
ensuring a higher bar of food safety as well, for farmers, ranchers and
consumers?
Answer. While most of the discussion around agricultural data and
Big Data in agriculture has revolved around applications for crop
systems, the data revolution holds tremendous implications for
livestock producers and the consumers of their products. Space does not
permit a full exploration of all the potential avenues by which new
data technologies could impact the livestock industry, but below are a
few examples.
Animal traceability: One example of how data technologies are being
used right now in other countries is to provide robust animal
traceability from farm to fork. For example, the bovine spongiform
encephalopathy (BSE, sometimes called ``mad cow disease'') outbreak in
the U.K. during the late 1990s spurred the implementation of an animal
identification system that can trace an animal throughout its life from
birth to the retail packaging in which its cuts are sold. Put another
way, a steak could be traced back to the retailer, wholesaler, meat
processor, feedlot, and cow-calf operation at which the calf was born.
That level of traceability has the potential to prevent losses that
could reach billions of dollars in the event of a BSE discovery in the
United States by quickly isolating every other animal with which a
diseased animal came into contact and drastically reducing the number
of animals that would have to be quarantined or destroyed to prevent
spread of the disease. It could simultaneously determine the wholesaler
and retailer location of any potentially dangerous food products. Quick
and accurate disease traceability thus has important implications for
both livestock producers and consumers.
Disease traceability is an important example of the loss-prevention
capability of data technologies, but there are also important value-
added applications as well. The same technology can enhance the ability
of livestock producers to provide age and source verification of beef
products, enhancing their ability to market their beef as a branded
product rather than a commodity. Traceability also allows ranchers to
show compliance with certifications and standards to demonstrate to
consumers that given affinity traits have been maintained throughout
the beef supply chain.
Keeping of such traceability information is possible with manual
technologies, but is made faster, easier, and much more accurate and
reliable by automation technologies such as radio frequency
identification (RFID) tags for animals coupled with automated readers
and loggers. Those technologies, working together with an already
robust supply chain data system in the wholesaler and retailer sector,
have significant potential to rapidly improve traceability within the
U.S. beef system.
Managing production inputs: One of the reasons the use of data
technologies has received so much attention in crop production is that
it is relatively easier (though one hates to apply the term ``easy'' to
the significant work in technological research and application that has
occurred there) to deploy sensors and data collection/transmission
systems on tractors, planters, sprayers, and combines than it is to
cattle. However, advancements in sensors and computers continue to make
it easier to use these advanced data technologies in livestock
applications.
In addition to the traceability systems mentioned above, we
continue to see advances in sensor systems allowing ranchers to learn
more and more about their production input use. The corn farmer cares
about seed variety, water uptake, fertilizer inputs, and disease
pressure as variables in determining their crop yield. A robust genetic
database and traceability system coupled with wearable sensors could
help a rancher track a cow's water and feed intake and health. We
already have technologies allowing for feed and water tracking in
closed environments like a feed yard or dairy, but advancements
continue toward wearable sensor technologies that would allow similar
tracking in open pasture environments.
Beyond tracking these factors, such sensors could also be coupled
with other data systems to provide significantly enhanced management
information for ranchers to see what production variables yield the
greatest financial returns. An example of this is a program here in
Oklahoma called the Oklahoma Quality Beef Network (OQBN). Through a
number of manual protocols, producers participating in the OQBN have
the information needed to show compliance with a number of value-added
programs giving them access to premiums in the marketplace for their
cattle. Participation in the program also gives ranchers the ability to
analyze the costs and returns of their production practices. OQBN
represents another system that could be enhanced by the deployment of
data technologies and a strong animal traceability system. For example,
as an animal is given the OQBN vaccine regimen, each vaccine could be
scanned as administered to the cow after the cow's RFID scan; those
vaccines would then be part of the cow's OQBN record to show compliance
with its vaccination requirements, and that information could be
displayed at the auction as the cow comes into the ring.
Managing range cattle production: At the risk of stating the
obvious, managing range cattle production is hard, and those challenges
are multiplied in Nevada by the significant size and/or rugged
topography of ranches. Many of the sensor technologies referenced
herein require cattle to be close to a collection point such as a
watering point or feed area, but those points are difficult to come by
in range production. Even if a rancher could put a transceiver near
common watering points or feeding areas on a range, those transceivers
depend on cellular data networks to collect and transmit data. The
expansion of rural broadband cellular coverage could increase the
opportunities for producers to take advantage of these technologies,
however. Broader and stronger networks would allow many more
opportunities for range cattle producers to take advantage of these
data technologies.
Another technology that could be of great benefit to range cattle
producers is unmanned aerial systems (UAS or ``drones''). In their most
straightforward application, UAS could be used to help ranchers check
on cattle much more quickly than they could by ground vehicle or
horseback. This could include everything from taking pictures or video
to count cattle, inspect fences, or check on water availability. This
is particularly true in rugged, mountainous terrain that is difficult
to access.
If the UAS were equipped with data transceivers backed by a strong
rural broadband network, the UAS could even collect and relay the
sensor data mentioned above, or could alert a rancher to a cow that was
ill and in need of immediate treatment. While some current UAS
technologies could already help in this regard, current FAA regulations
only allow most UAS to be operated by ``direct line of sight'' which
means the pilot must have direct visual contact with the UAS. That rule
makes some of these applications impractical, particularly in
mountainous terrain. Continued improvements in camera and transmitter
technologies could soon make a ``point of view'' camera mounted on the
UAS an acceptable substitute for direct visual contact, though,
permitting amendment of the FAA regulations to allow longer flight
ranges.
Consumer-side data: Much has been discussed in terms of how Big
Data technologies can create models to help crop production, but there
is also significant potential for those technologies to impact beef
production, and indeed, consumption. We already have large volumes of
universal product code (UPC) scanner data from retail outlets that can
tell us much about where and when certain food products are consumed.
However, if the beef industry were to couple that data and supply chain
information with the traceability elements mentioned above, the amount
of data available for analysis would grow exponentially, and tremendous
insights could be derived about beef demand, consumption patterns, and
other factors. Big Data could also provide significant insights into
the effectiveness of food policy such as the use of SNAP EBT benefits
for certain types of foods, the nutrient profile of those purchases,
and consumer incentives.
Question 2. There were many references to the environmental
benefits of agricultural data. Are we in a position yet where we can
authoritatively quantify the environmental benefits experienced by the
use and attention to these technologies and data analysis? For example,
is this science proven to the point that we should be creating
incentives in the farm bill conservation title for their utilization to
keep pristine watersheds like Lake Tahoe, or water quantity in drought
areas, solidified for the decades to come? Or is there a place for this
use in connection with the Federal crop insurance program?
Answer. The most concise answer is ``we're not there yet, but we're
getting closer rapidly.''
At the ``micro'' or farm scale, we may indeed be getting close to
having the technology in place to help producers demonstrate compliance
with environmental requirements. For example, say a producer has a
feedlot that meets the definition of a Concentrated Animal Feeding
Operation (CAFO) under the Clean Water Act and wants to show their
compliance with the CAFO's nutrient management plan (NMP). A grid-
sampled soil test of the field to receive animal waste could create a
precise prescription for how much of each macronutrient should be
applied to that field. Sensors on the tractor used to inject liquefied
animal waste to crop ground, coupled with a nutrient analysis of the
waste, could show the amount of nitrogen, phosphorous, and potassium
applied using variable-rate technologies to comply with the
prescription. Later, the crop harvested from the ground where the
wastes were applied is reported using a combine's yield monitor data,
which can show the nutrient uptake of the crop and thus the nutrient
balance for the soil.
This scenario demonstrates the technologies available to manage
agricultural environmental factors on the input side, but if we want to
reliably automate environmental compliance efforts, we must also have
strong sensor networks in and around environmental receptors such as
streams and lakes, particularly at points where watersheds connect.
This will require continued development of environmental sensors and
again, strong rural broadband networks over which to transmit the data.
The discussion to this point has focused on farm-level compliance,
but what about compliance at a ``macro'' or regional level, such as
demonstrating compliance of a watershed with an agriculturally-based
total maximum daily load (TMDL)? We may have a bit to go before we can
reliability show compliance on a regional level, simply because doing
so would require 1) the continued penetration of sensor and
transmission technology into more agricultural implements and 2) the
interoperability of those systems to ``speak the same language'' so
insights could be derived across multiple farms, rather than on a farm-
by-farm basis. However, with that said, these factors continue to
advance rapidly, and we may cross that threshold soon. Such
technologies could soon facilitate the ability of farmers to manage
nutrients in impaired watersheds. For example, in a TMDL-limited
watershed with a ``cap and trade'' system, producers could validate how
much of each nutrient they have applied to show their compliance with
the nutrient limitations they received or traded.
In any of these scenarios, it is important to note that the
accuracy of any compliance system depends on the proper calibration and
operation of the sensors used. While automation may make compliance
much easier, if automation is ever used to demonstrate compliance with
a regulatory program, it should also be paired with the opportunity for
a producer to review their submissions and explain anomalies. If a
sensor suffers a malfunction, it could show a producer applied far more
than their allocation of a nutrient, even though the proper amount was
actually applied. Conversely, a third party might try to ``hack'' a
tractor system or the data collection to implant false data. Any
compliance system created in the future should make allowances for how
to handle these anomalies.
With respect to the crop insurance program, there are tremendous
opportunities to hone the actuarial models for the system and to
facilitate produce's demonstration of compliance with insurance program
requirements afforded by both Small Data and Big Data systems. Big Data
analytics can continue to drive gains in accuracy for models in setting
premiums and cost management, while Small Data can help producers
accurately report yields (again, presupposing proper calibration and
operation of the equipment sensors).
Question 3. Obviously in the west we have some greater agriculture
challenges than other regions, and wildfires are one of them that we've
seen have detrimental effects over and over again. In many cases, these
fires have a multiplier effect on production and for things like future
flooding. And, in general, water is also another constant challenge.
Are there technologies that can help measure and account for drought
conditions, or measure the volatility of fires within parched forests
or grasslands before we have fires that potentially get out of control?
Are there any associations with improving broader weather prediction or
forecasts? And are there other specifics where you foresee this data
and technologies helping us get a better handle on our climate change
crisis?
Answer. There are indeed a number of technologies that can help
address these concerns. One present example of a weather and climate
monitoring system that has provided tremendous benefits for my home
state is the Oklahoma Mesonet. The Mesonet is a system of weather
stations spread throughout the state, with at least one (if not more)
such stations placed in every county. These stations fill in the
significant geographic gaps between NOAA weather monitoring stations,
and allow us to collect data on dozens of weather and climate
parameters. These observations are fed into a number of models that
help us keep close tabs on a number of factors for everything from fire
weather hazards to drought monitoring and evaporative losses. The
Oklahoma Mesonet has already helped our meteorologists and
climatologists refine their predictive models and led to a tremendous
output of research into Oklahoma weather and climate issues. The
application of similar technologies in Nevada could provide significant
improvements to prediction of fire weather conditions with the
potential to issue advisories and reduce the risk of ignition sources.
Similarly, the evaporation, soil moisture, and mesoscale models could
help refine the use of scarce water resources. To help illustrate the
applications of the Mesonet and our models, such as the fire-weather
model, I have attached a summary prepared for you by Mr. Al Sutherland,
coordinator of Mesonet Agricultural Data and Products, with the
assistance of Dr. J.D. Carlson, lead investigator on the OK-FIRE model.
Further, in the future, the reduction in cost of reliable weather
sensors could mean individual landowners could have their own weather
stations (as an increasing number of Oklahoma farmers and ranchers do)
that could be interoperable with Mesonet stations. This would have the
effect of filling in even more gaps in the network, increasing the
precision of its observations and providing even better information for
predictive models. Indeed, weather monitoring is an arena in which
there is tremendous opportunity for improved sensor systems and Big
Data to have a positive impact in climate management.
______
Response to Written Questions Submitted by Hon. Jerry Moran to
Todd J. Janzen
Question 1. As administrator of the Ag Data Transparency Evaluator,
you are familiar with the lack of trust and confusion that many farmers
experience in identifying what exactly is done with data collected from
their land. Can you please describe considerations among industry
stakeholders that led to publishing the ``Core Principles'' that are
incorporated into the Evaluator?
Answer. The main drivers behind the ``Core Principles'' for ag data
were the concerns from farmer-members of national farm organizations,
such as American Farm Bureau Federation, National Farmers Union,
National Corn Growers, National Association of Wheat Growers, American
Soybean Association, and National Sorghum Producers. These organization
spearheaded the effort to develop the Core Principles because their
members wanted a basic framework around how ag data is collected and
shared.
Question 2. Your testimony states that only nine companies
(including Farmobile) are currently approved as ``Ag Data Transparent''
according to the Ag Data Transparency Evaluator's formal process. Why
have not more companies voluntarily completed the evaluation,
especially given the fact that nearly 40 companies participated in
drafting the ``Core Principles?''
Answer. This is a question best addressed to those companies that
have not participated. I can only speculate as to their delay in
participation. My belief is that these companies want more control over
farmers' data than they are willing to publicly admit. Therefore, it is
easier to remain quiet and say nothing than subject themselves to the
Ag Data Transparent process.
Question 3. How can we incentivize more active participation by
industry stakeholders to complete this evaluation?
Answer. I think the fear that Congress might step in and regulate
the privacy and collection of ag data is something that will drive more
companies to participate. As the value of the Ag Data Transparent brand
increases over time, that will drive more participation as well.
Question 4. As agricultural data becomes more valuable to entities
outside of the farmers that collect it, data security concerns are
likely to grow exponentially while criminals with all types of motives
seek to illegally gain access to and capture privately-owned data. How
do you foresee data security practices in the agricultural industry
evolving as a result?
Answer. Data security in the ag data space must progress at the
same rate as data security in the non-agricultural space. Ag tech
companies should not think that they are immune to security challenges.
Question 5. Are there any specific security traits to agricultural
data that need to be accounted for steps going forward?
Answer. Ag data can contain proprietary information, which makes it
different than other types of consumer-type data that may not be
proprietary.
Question 6. In a 2016 poll conducted by the American Farm Bureau
Federation, regarding the loss of control over downstream uses of data,
sixty-six percent of the farmers polled expressed concern about not
being compensated for the potential benefits from the use of their data
beyond the direct value they may realize on their farm. Meanwhile,
sixty-one percent of the farmers were concerned that agricultural
technology providers (ATPs) could use their data to influence market
decisions. Which of the two concerns do you believe is the greatest
threat to farmer profitability and well-being, and what should be done
to alleviate these concerns?
Answer. I believe the greatest threat to the farmer is that ATPs
will be able to influence the ag markets by using ag data, but without
making that same g data available to farmers. That would put certain
holders of information in a superior position to the average farmer.
Question 7. As more and more firms enter the agriculture-technology
space and interact with data used by and/or generated by farmers, the
need for clarity and consistency on privacy principles is growing. For
these new entrants, can you suggest any best practices these firms
should engage upon to ensure their data privacy procedures properly
convey the data's expected use?
Answer. New firms in the ag data space should do two things when
they begin to collect data. First, they should determine their guiding
principles for how they intent to treat ag data. Second, they should
develop easy to understand data use policies that they can share with
farmers that explain how the firm intends to use the farmer's data.
Question 8. As more and more firms enter the agriculture-technology
space and interact with data used by and/or generated by farmers, the
need for clarity and consistency on privacy principles is growing. For
these new entrants, can you suggest any best practices these firms
should engage upon to ensure their data privacy procedures properly
convey the data's expected use?
Answer. New firms in the ag data space should do two things when
they begin to collect data. First, they should determine their guiding
principles for how they intent to treat ag data. Second, they should
develop easy to understand data use policies that they can share with
farmers that explain how the firm intends to use the farmer's data.
Question 9. While much of the data we discussed in the hearing is
generated on farm and captured by farmers or their equipment,
significant quantities of data is publicly available and critically
important to inform risk modeling, yield prediction, etc. in both the
public and private sector. How can we encourage the continued use of
this type of data, and even grow our sources, while ensuring that
farmers understand their role in this process?
Answer. Witness did not respond.
______
Response to Written Questions Submitted by Hon. Catherine Cortez Masto
to Todd J. Janzen
Question 1. Given the dramatic benefits of agricultural data that
were outlined in the hearing, are we addressing these advancement in
our high school agricultural education and training efforts? Or at the
technical or community college education level? And much of what we
heard about, and what we think about in the innovation sector, requires
more students and a workforce who can work with the technology, data
analytics, and various other computer science skill sets. Is that
accurate from your perspectives? Are there ways in your mind we can
better incentivize and be developing the workforce we'll need to see
the great promise of what we're talking about today? Including the
cyber security needed to protect agricultural and ag-related business
in this sector?
Answer. I am not involved with high school or secondary education
and cannot therefore directly speak to this question. From my personal
experience, more could be done to drive more students into science and
technology focused careers, including agriculture. National and state
FFA organizations do a lot to help foster growth in this area, but
these are non-profit organizations that rely on donations and
volunteers.
Question 2. There were many references to the environmental
benefits of agricultural data. Are we in a position yet where we can
authoritatively quantify the environmental benefits experienced by the
use and attention to these technologies and data analysis? For example,
is this science proven to the point that we should be creating
incentives in the farm bill conservation title for their utilization to
keep pristine watersheds like Lake Tahoe, or water quantity in drought
areas, solidified for the decades to come? Or is there a place for this
use in connection with the Federal crop insurance program?
Answer. The data collection and analytic tools today are already
smart enough to make meaningful differences in environmental
protection. For example, we could measure fertilizer use on farmland
and fertilizer run-off from field tiles and other pathways to determine
proper fertilizer application rates, assuring that as little as
possible is lost. Likewise, we can compare fertilizer usage and yield
across fields to determine proper application rates and timing to
maximize fertilizer and soil resources.
There could certainly be more use of data in the Federal crop
insurance program. Deb Casurella at Independent Data Management LLC is
an expert on this subject and would be the right person to testify on
this topic.
Question 3. So if we acknowledge the virtues of agricultural data,
and that small data can even be used to market the sale of land or an
operation, what safeguards in place to verify this information so a
potential land buyer isn't defrauded by skewed land performance data or
analytics that inflated the profitability of the land?
Answer. The best way to verify that ag data is not fraudulent would
be to insist on receiving the original raw data from a seller, or
obtain a copy of the same data from a third-party source. Even then,
there would still be a question whether the machine that produced the
raw data was properly calibrated. One way to verify that would be
benchmark the seller's data with other data in the same area. Any
dramatic difference could be due to data manipulation.
Question 4. Have agribusinesses been utilizing varying strategies
to collect big data in agriculture? What was the old typical procedure
for companies to obtain information from farmers, or their customers,
on the progress or performance of their product? And are there
situations where the producers are forced into a data collection
program by seed, fertilizer or equipment companies? Or can they opt-
out?
Answer. Many of the larger agribusiness companies have been
collecting agricultural data from customers for years. Often, they
collected this information because they were providing services to
farmers, such as fertilizer of pesticide application. They collected
data during such activities and then retained it for their own records.
I am not aware of any specific situations where companies require
ag data submission as a condition of using a machine or device,
although I think that will inevitably happen.
______
Response to Written Questions Submitted by Hon. Jerry Moran to
Dr. Dorota Haman
Question 1. One of the aspects of your background I found
interesting is the work you've done with farmers in developing nations,
specifically as it relates to irrigation. I chair the Senate Hunger
Caucus and have worked for many years on the issue of reducing food
insecurity in the world. I believe that agriculture development
initiatives that help countries feed themselves is a key part of the
long-term strategy to end global hunger. Can you elaborate on your work
with farmers in developing countries, specifically as it relates to
using water more efficiently, and how that work has reduced food
insecurity?
Answer. I would like to take this opportunity to thank Senator
Moran for asking this question and for his bipartisan leadership on the
Senate Hunger Caucus. Food assistance, and other support provided by
the U.S. all over the world, leads to reduction of global food
insecurity.
My expertise is in agricultural engineering with a focus on water
management and irrigation. As reported by the Food and Agriculture
Organization (FAO) of the United Nations and by the World Bank,
approximately 70 percent of fresh water usage is for agriculture. Most
of agricultural irrigations systems are poorly designed and poorly
managed. Even the best irrigation systems, if not maintained and
carefully managed, are inefficient. Most of the irrigation systems
over-apply water and there is a potential to improve efficiencies
through technology and education.
My major effort has been focused on Florida growers and specialty
crop production in Florida. I have also been a university teacher
working with the next generation of farmers, academics and irrigation
specialists. I teach people how to design, manage and maintain
irrigation systems. My work has been focusing on efficient systems such
as microirrigation and sprinklers. These systems are usually used for
higher value, specialty crops such as fruits and vegetables but can
also be adapted for small farmers in developing countries.
My international work has been largely in education. I have worked
for FAO in Zimbabwe designing curriculum and lab experiments for a six-
months intensive course focused on planning, design, maintenance and
management of irrigation for smallholder farmers. I have taught two 2-
week courses in Egypt. I have spent 3 months in Mexico and 4 months in
Chile investigating and teaching efficient methods of irrigation. In
addition, my students have worked with farmers in India, Ecuador,
Columbia and Poland.
As an example, one of my students, working in Jamaica on his
Masters project, implemented a simple drip system for calaloo (Jamaican
spinach) and cucumbers. The increases in yield and reduction in water
usage were significant. After the experiment was finished, the farmer
adopted the system on the entire farm.
Question 2. What opportunities exist, if any, to take the
technology being used today by large-scale U.S. farmers and use it to
help smallholder farmers in developing nations be more productive and
sustainable?
Answer. New technologies can eliminate many maintenance mistakes
through automation and sensor control. New technologies can provide
inexpensive alerts, and in the future, automatic intervention.
Technology leapfrogging is likely in agriculture in developing
countries. Use of cell phones in developing countries is often cited as
an example of leapfrogging. Apps and advisory programs, built and
available from the open sources, can be made available on the
smartphones. Solar phone chargers are becoming available even in very
remote locations without an electrical grid. Access to quality data is
critical for development of Apps and tools that can be available to
poor farmers. One of the examples of an open platform is the BioSense
Institute in Serbia. This project was funded from the European program
Horizon 2020.
______
Response to Written Questions Submitted by Hon. Bill Nelson to
Dr. Dorota Haman
Question 1. In addition to improving crop yields, how can big-data
be used to protect farm workers from heat-related injuries? Is the
farming industry using technology and data to improve worker safety?
Answer. In collaboration with the Farmworker Association of
Florida, Dr. Linda McCauley, Emory University in Atlanta (GA), has
established the Los Girasoles heat stress study. Los Girasoles, funded
by the National Institute for Occupational Safety and Health (NIOSH),
is a four-year project (now in its third year) that aims to better
understand how agricultural workers respond to heat stress and collect
better data on the magnitude of heat-related illnesses like heat stroke
in agricultural work.
The data have been used to develop the Heat Related Prevention
Illness training for the PISCA (FSU-FWAF) project and the Farm Labor
Supervisor Training (FLST) in IFAS/UF.
An APHA policy statement have been drafted and approved. A local
representative is sponsoring local legislation to promote a regulation
of heat stress. A pilot study has been conducted with FSU looking at
alternative clothing. Data on kidney functions as related to heat
stress have been collected by researchers last year and will be
continued next year.
Large amounts of physiological data collected by Los Girasoles was
presented at several conferences and meetings. This is a list of
current publications on the topic:
1. Flocks, J., Mac, V., Runkle, J., Tovar-Aguilar, A., Economos, J.,
& McCauley, L. (2013). Female farmworkers' perceptions of heat-
related illness and pregnancy health. Journal of agromedicine,
18(4), 350-358.
2. Mutic, A., Mix, M., Elon, L., Mutic, J., Economos, J., Flocks, J.
Tovar-Aguilar, A., & McCauley, L. (2017). Classification of
Heat-Related Illness Symptoms among Florida Farmworkers.
Journal of nursing scholarship.doi:10.1111/jnu.12355
3. Mac, V., Tovar-Aguilar, A., Flocks, J., Economos, E., Hertzberg,
V., & McCauley, L. (2017). Heat exposure in Central Florida
fernery workers: results of a feasibility study. Journal of
agromedicine, 22(2), 89-99.
4. Hertzberg, V., Mac, V., Elon, L., Mutic, N., Mutic, A., Peterman,
K., Tovar-Aguilar, A., Economos, E., Flocks, J., & McCauley, L.
(2017). Novel Analytic Methods Needed for Real-Time Continuous
Core Body Temperature Data. Western journal of nursing
research, 39(1), 95-111.
Wearable sensors similar to Fitbit can measure temperature and
heart beat (or other biometrics) to alert workers to dangerous
conditions. Sensors that are built into the clothing are also under
development. CorTemp http://www.hqinc.net/ has been used for body core
temperature, which also allows capture of heart rate signals, in
addition to button temperature sensors in and outside workers'
clothing, including ActiGraphs (medical-grade wearable activity and
sleep monitoring solutions for the research community).
In addition to Florida, there are teams in California, Tennessee,
and Sinaloa (Mexico) who are using similar equipment. To improve
diagnostics, researchers are collecting blood, urine, BMI, HBP, body
fat percentage, and temperature at workers' homes.
Question 2. Florida has faced both record droughts and record
rainfall in the past few years. Can you describe how technology can
help farmers prepare for and adapt to wild swings in weather?
Answer. Technology can help farmers prepare and adapt to wild
swings of weather in several ways. First it can help inform farmers of
upcoming weather conditions in a timely and user-friendly way. It can
also help farmers monitor conditions on a farm to better track the risk
of pests and diseases, soil moisture conditions, and other potential
yield-reduction factors that may require control measures. However,
collecting data in an efficient and cost-effective way is just the
first step in the process. Data must be translated into information
through the use of mathematical models, machine learning, artificial
intelligence and other data analytics to infer information from this
new ``data rich'' environment. Data and information only have value if
used to drive a farmers' decisions. That is why we need to provide
farmers with tools that are customized to their farming environment and
conditions.
It is also important to highlight that we need to develop and
promote technologies and management practices that help farmers become
more resilient to climate variability and change: http://
agroclimate.org/fact-sheets/climate/
To adapt to these extremes, systems should be designed to a higher
standard than is currently normal. For example, drainage systems should
be designed to handle extreme rainfall events. At the same time, to
cope with extreme drought events, efficient irrigation systems and
additional water supplies should be identified.
Question 3. How can big data help us prepare for and adapt to the
effects of climate change?
Answer. Large-scale geo-reference data can be very helpful in
identification of the long-term trends in weather patterns and climate
change. With new ways of data collection through satellites, drones,
robots and cheap sensors, there will be improved data collection and
analysis.
Reliable and early prediction of both drought and rainfall based on
climate models and global and local weather data (that are becoming
more dense and precise) are critical for dealing with extreme weather
events. Forecasting weather for the next few days or a week, the next
rain front, the next heat wave, the path of the next hurricane and
trends in climate change over the next decades use the same climate
models which are build, tested and updated with billions of weather/
climate data from tens of thousands of weather stations and sensors
from land, ships, airplanes, ocean floats at various depths, research
stations, and satellites. These are examples of big data already used
and applied in daily life.
Increasing accuracy of weather event predictions is critical to the
timeliness of preparations, decisions and interventions. For example,
expectation of unusual rainfall may require lowering of the water table
in the field or draining of water reservoir to create more space for
upcoming rain. Timing of this operation is critical. For example,
draining water storage in preparation for extreme rain cannot be done
too early or the plants may be stressed to the point of yield
reduction. The accuracy of prediction is critical to successful
preparation and big data from numerous sensors, as well as remote
sensing, can increase the accuracy of prediction.
Technology can also be helpful in managing scarce resources during
drought and regulating water application to optimize incomes under
drought conditions. Sufficient lead-time is important for
implementation of water storage to minimize the impact of drought on
production.
It is important to remember that improved short-term forecasts (on
the scale of days up to the entire season) can be very useful for
management decisions and can mitigate the impact of an extreme event.
Big data may be able to help in analyzing reports of the impacts of
extreme events. Currently, state climatologists and the National
Drought Mitigation Center rely on somewhat subjective reports of
drought impacts. Big data can help in analyzing diverse reports of
impacts in a more objective manner.
Question 4. Can big data help us reduce greenhouse gas emissions
from agriculture?
Answer. Big data allow us to see the patterns and fine-tune systems
to make them more efficient. Well-known examples of the agricultural
sources of greenhouse gas emissions are paddy-rice production and
animal production. Another is oxidation of the organic material in
cultivated soils.
To shift from the paddy-rice, scientists are working on increasing
production of upland rice that is not flooded and this growing system
contributes much less to atmospheric greenhouse gasses. Increased
efficiency in animal production and waste management may be beneficial
to reduction of methane emissions. Efficient management of these
agricultural systems is critical and the information provided by big
data can help us optimize the systems through mathematical modeling.
For example, by increasing the land-surface area covered with plants
and reducing deforestation we can increase removal of
CO2 from the atmosphere. New technologies, including
robotics and low-energy LED lights for protected (indoor) plant
production, are opening possibilities for vertical growth (in
multistore buildings) of high value vegetables and fruits in urban
areas. This production may benefit from CO2 enrichment
(using industrially produced CO2) at the same time
preventing its release into the atmosphere.
To reduce CO2 in the atmosphere, numerous management
strategies have been discussed including using plants to draw carbon
out of the atmosphere and developing techniques to hold it in the soil.
These strategies vary in effectiveness across different climates, soils
and geographies. Sequestration of carbon is achieved through
transferring atmospheric carbon into the soil via plant photosynthesis.
Soil carbon must then be protected as effectively as possible from
microbial activity that will release the carbon back to the air. Most,
if not all, of the management techniques (operations) that promote
carbon sequestration also improve soil aggregation, water retention,
soil fertility, and food security.
Question 5. Should we be thinking about ``big data'' and how
farmers control their information the same way we think about consumers
using the Internet or conducting financial transactions?
Answer. I believe that these factors need to be carefully analyzed
and a set of acceptable rules should be developed and established. I
believe that it is important to make sure that sufficient high-quality
data are available as an open source to allow for creating free access
to critical information (or at very reasonable cost) especially for
small, family farms and for farmers in developing countries. For
example, in Europe, all the data created under the Horizon 2020 program
are open access and the program includes a lot of agricultural big
data. Another example of this type of platform is the BioSense
Institute in Serbia.
Question 6. Can you describe how extension services like the one at
the University of Florida help famers--particularly small farmers, new
farmers, and specialty crop farmers--access the same type of
information that corporate farmers have?
Answer. The University of Florida Institute of Food and
Agricultural Sciences (IFAS) has been engaged in developing services
and decision aids to help Florida farmers, including small farmers,
improve resource use efficiency and reduce risk associated with climate
variability and change. AgroClimate.org (http://www.agroclimate.org)
for example provides climate related information and dynamic
application tools that interact with a climate database system for the
Southeastern U.S.A. Information includes climate forecasts combined
with risk management tools for a range of crops, forestry, pasture, and
livestock. It has been quite successful and adopted by farmers in
Florida to decide when to apply fungicide to strawberry fields based on
weather conditions or to track the accumulation of chill hours during
the winter in farms growing temperate fruits such as blueberries and
peaches. AgroClimate integrates weather and climate data from public
sources such as the Florida Automated Weather Network (FAWN) and
gridded weather data from NOAA that covers the entire U.S.A. and the
globe. Several mobile phone apps have been developed based on the
weather and climate database services provided by AgroClimate including
the Smart Irrigation Apps http://www.smartirriga
tionapps.org) that help farmers in Florida and Georgia schedule
irrigation of crops such as citrus, strawberries, cotton and turf;
thereby saving water and reducing leaching of nutrients into the
groundwater.
At UF, we are planning to create a similar portal for safety
information and safety practices in agriculture. We are in the process
of hiring a new faculty member to work on this project.
Question 7. Do you think that consumer concerns about genetic
engineering and GMO crops might create a backlash against precision
technology, particularly as it relates to our food supply?
Answer. I believe that this is unlikely. Precision agriculture is
not directly linked to genetic engineering or GMO. In fact, due to
precise management of resources (water, nutrients, pesticides and other
chemicals, etc.), available technologies benefit the environment and
reduce input costs. Precision technologies have been used for years in
agriculture. Data availability, especially Big Data, makes these
systems more precise, allows for faster and better decisions, and
increases operation and production efficiency. These technologies do
not introduce anything new or foreign into food chain. New genetic
technologies, such as CRISPR (gene editing) seems to be less
controversial and more likely to be accepted by the general public.
______
Response to Written Question Submitted by Hon. Catherine Cortez Masto
to Dr. Dorota Haman
Question. There were many references to the environmental benefits
of agricultural data. Are we in a position yet where we can
authoritatively quantify the environmental benefits experienced by the
use and attention to these technologies and data analysis? For example,
is this science proven to the point that we should be creating
incentives in the farm bill conservation title for their utilization to
keep pristine watersheds like Lake Tahoe, or water quantity in drought
areas, solidified for the decades to come? Or is there a place for this
use in connection with the Federal crop insurance program?
Answer. I do not believe that we are yet ready for inclusion of
specific incentives, policies and regulations. Scientists are working
on information that can improve crop insurance program but we are not
ready to implement it yet. The new technologies and sensors are
providing abundant and potentially high quality agricultural data that
can be very beneficial (in the future) to evaluate the environmental
impact of agriculture and hopefully optimize the entire system to
reduce the impact on the environment. Massive data require the
development of adequate models, and these are not yet mature but are
rapidly evolving. At this point, model development is one of the major
challenges in big data analytics. The concept of analyzing big data
relies excessively on ``blind'' machine learning, where ``black boxes''
of data are ``mined'' in the hope that the process will ``tell us what
it contains''. The problem is that these large data sets are of highly
dimensional (complex) and there are many possible combinations of the
drivers that could potentially produce similar results. Identifying the
correct relationships requires experts with mature and relevant
conceptual models to drive the search process. In addition, ``dimension
reduction'' (selecting a smaller set of really important factors
suitable for management) is a critical step.
Interdisciplinary approach is necessary to assure that big data are
appropriately analyzed. Teams of subject-matter experts need to team up
with machine-learning experts (from statistics and informatics) to
identify the correct solutions to the problems. The challenge today is
that because of high potential economic gains, sensors, data, and
machine learning has progressed rapidly but has not teamed up with
content experts for specific problems. This is particularly true in
Agriculture and Environmental Sciences, which are particularly complex
as they are based on open and largely uncontrolled systems (as opposed
to other artificial or human-controlled settings).
[all]
This page intentionally left blank.