GPS and Business Intelligence: Rubenacker Farms Makes Hay With I.T.

 
 
By Kim S. Nash  |  Posted 2006-10-02
 
 
 

The rain may pour, the sun may pound, but Kelly Robertson doesn't mind. He's a southern Illinois farmer who's always drawn a living from nature.

Between Rubenacker Farms in Dahlgren, Ill.—the commercial crop farm he helps manage—and his own fields down the road, Robertson oversees more than 9,300 acres of corn, wheat and soybeans.

His crops travel more than he ever has, part of the $200 billion worth of agricultural products grown in the U.S. each year, and the

View the PDF -- Turn off pop-up blockers!

$68 billion to be exported this year to South Korea, Venezuela, Saudi Arabia and other points around the world.

Ten years ago, a grain farmer walked the fields and stooped to run his fingers through the topsoil for a measure of its moisture and depth. He'd slide a notebook from his shirt pocket to pencil in the date he seeded the fields and how much fertilizer he sprayed.

When he looked to the skies for the gray, cauliflower-shaped clouds that bring heavy rain, or the thin wisps that float over sunny days, he knew he could do only so much to coax corn planted in May to sprout enough sweet ears by September to make the sweaty, bent-back work profitable.

Robertson, who is in his mid-30s, used to work that way. But for the past 10 years, he's worked among the growing ranks of "precision farmers" who supplement their farm sense with software and statistical analysis to help decide what to plant, how to grow it and when to sell it.

The most advanced precision farmers plow and plant with auto-steer tractors equipped with global positioning systems that guide the machines over fields by themselves. They fertilize with sprayers loaded with geographic information systems maps and computerized instructions to vary the amounts of nitrogen applied to different spots on a field, down to four-inch patches.

And, as they harvest their crops, weight and force sensors attached to 15-ton combines record the volume of corn or soybeans pulled from each plant at two-second intervals, associating the data with specific locations on field maps.

Then, like the best Wall Street risk managers, precision farmers upload these data points, via Zip drives or wireless networks, into statistical analysis software on Microsoft Windows systems in their offices. There, the data is combined with business information, such as the price of seed, cost of fertilizer, weather records and current market pricing. Analysis tools from SAS Institute, SPSS and others match costs and income to specific physical locations in the field to let farmers calculate crop profitability, row by row.

Meanwhile, the U.S. Department of Agriculture and various universities offer data and analysis tools for free, for monitoring crop prices, predicting weather and comparing infestation treatments. And farmers buy and sell grain over the Internet, like day traders, at sites such as Farmstech.com by Farms Technology in Overland Park, Kan.

"Farming is the oldest known human activity. You'd think that after 10,000 years there'd be nothing left to improve. Not true," says Michael Swanson, agricultural economist at Wells Fargo, a bank whose $10 billion in annual agricultural loans make it the biggest lender to farms in the U.S.

Information technology has made farming an analytical activity, and though some farmers shy away because of costs, many embrace the science. A close look shows that the technology produces solid returns and is drastically changing the business of agriculture.

Education levels among farmers have increased since the early 1990s, the USDA says. In 1991, 35.8% of U.S. farmers had attended college. By 2003, the latest year for which data is available, it was 45.2%. Ninety-two percent of young farmers, age 18 to 35, use computers and 89% consider the Internet a farming tool, according to the American Farm Bureau Federation, a lobbying group in Washington, D.C.

A reliable count of farms that do precision farming doesn't exist. But Purdue University estimates that 20% of U.S. grain, for example, is planted using tractors with auto-guidance systems and custom-fertilized with computerized sprayers. And half of all U.S. grain is harvested by combines with yield monitors.

The effects of technology and business management techniques are evident to Swanson. In 1950, he notes, American farmers planted 83 million acres of corn, which produced 38 bushels per acre. In 2004, 81 million acres of corn were planted, with each acre yielding 160 bushels. Think about it: 2.5% fewer acres produced more than four times as much corn.

If today's farmers didn't use the technology they do, Swanson estimates, they would have had to plant 320 million acres of corn last year to meet demand: "We'd be planting parking lots and backyards."

NEXT PAGE: Tech Supplements Nature

Tech Supplements Nature

Several vendors offer hardware and software for precision farming.

Deere & Co. in Moline, Ill.; Trimble Navigation in Sunnyvale, Calif.; Caterpillar in Peoria, Ill.; Komatsu in Tokyo and CNH Global in Amsterdam build tractors, combines and other heavy machinery with global positioning systems, radar and scales.

For analytics, there is ArcView mapping and analysis software from ESRI in Redlands, Calif., and SMS Software specifically for the agriculture industry from Ag Leader Technology in Ames, Iowa. Also popular are statistical tools from business stalwarts SAS Institute in Cary, N.C., and SPSS in Chicago.

Robertson uses SMS Software, SPSS, a Microsoft Access database and Trimble GPS systems.

Information technology doesn't pervade agriculture the way it does, say, banking or manufacturing. There aren't many chief information officers on the farm. But in farming, good information management can mean the difference between making a go of it and going bankrupt, says Bruce Erickson, who teaches agronomy—scientific agriculture—at Purdue in West Lafayette, Ind. "You have sun, soil and water," he says. "You use technology to tweak those basic ingredients to eke out that bit of advantage to give you profit."

At Rubenacker Farms, that job goes to Robertson, who has a master's degree in plant and soil science, a passion for technology and a penchant for asking what-if questions such as the critical one that faced him last spring.

In recent years, he explains, Rubenacker has double-cropped on about 400 acres—first growing winter wheat from October to the following July, then, after harvest, planting soybeans on the same ground, for harvest in mid-September.

A look at average yields over 10 years shows wheat production has increased by 25 bushels per acre, but production of the double-cropped soybeans has dropped by six bushels per acre, he says.

Last spring, as his wheat plants sent up golden spikes, Robertson wondered if he shouldn't do something different. "We know we can get 30 bushels of double-crop beans per acre," he says, outlining how he analyzed his risks. "If we harvest the wheat earlier—first of June rather than first of July—we will still get a high yield of wheat. But can we plant corn that late and still get a profitable yield [on that]?"

A crop plan generated by SMS Software pointed toward corn. "You look at a whole lotta data plus real-time data—the price of corn, what can we sell it for versus beans," he says. When he talked about his decision-making process with Baseline, the corn crop looked good but wasn't ready to pick. Robertson was wishing for one more soaking rain. But until the corn is harvested and sold this fall, he won't know whether the change-up paid off.

NEXT PAGE: Big Bucks for Big BushelsBushels">

Big Bucks for Big Bushels

So, why hasn't everyone gone to precision farming?

Poor integration between different vendors' technologies has kept some farmers away, says Dusty Sonnenberg, who owns the 3,000-acre Sonnenberg Farms in Hamler, Ohio.

A lack of standards for sharing data between, say, Deere and Trimble equipment means farmers must choose one proprietary system or another, he says: "What can you do? You can't put Dodge parts on a Chevy."

But the bigger issue is cost. GPS systems and on-board computers for a tractor cost $20,000 to $25,000. A GPS base station with receivers to relay signals between it, an orbiting satellite and the farm machine costs another $15,000 to $20,000, Sonnenberg says. Tools from SPSS or Ag Leader go for about $1,500 to $2,000, depending on the module. "It's only innovators and early adopters now because of the price tag," he says.

Still, the financial returns do come, says Jess Lowenberg-DeBoer, a professor of agricultural economics at Purdue. When fertilizing, for example, a farmer pulling a sprayer behind his tractor will usually overlap up to 10% of the ground, to be sure he covers everything, Lowenberg-DeBoer says. But a tractor with auto-steer and a variable-rate sprayer overlaps just 1% to 2%, he says. More accurate spraying requires less labor and results in less wear on machinery. It also means spending less on expensive diesel fuel and fertilizer, both of which shot up more than 15% in price this year over last.

At Rubenacker Farms, Robertson likes to quantify the value he gets from agricultural consultants. Specialists come to the farm to scout for bug infestations or plant diseases, and to test soil for nutrients and moisture. Then, they recommend treatments.

The only way to know how effective their advice is, Robertson says, is to collect data and isolate different variables. Say aphids are eating the wheat crop and a consultant advises insecticide. The chemical plus labor might cost $12 per acre. To pay off, that investment must return at least as much in increased crop quality or increased yield.

Analysis of the situation would put Robertson in front of his Access database, querying for similar conditions at the farm in the past. Maybe there was similar insect trouble four years ago. Perhaps at that time, he sprayed some fields but not others because he didn't want to spend too much money not knowing how effective the treatment would be. No one wants to see a $12-per-acre activity result in returns of just $12.01. On 9,000 acres, one extra cent per acre—$90—isn't worth the trouble.

However, if past yield data shows that treated fields turned out several extra bushels per acre of higher quality—and more lucrative—corn than unsprayed ones, that's a different story. If, later on the open market, a farmer got 15 cents more a bushel for the corn because it's top-notch, and he harvested 148 bushels from each acre, that's $22 per acre in extra value from that $12-per-acre chemical spray.

"We would look back and see that not only did it pay, it paid in a big way," Robertson says.

His 400-acre double-crop corn gamble, however, is still an unknown. "Right now, it's looking like a good choice—if we get another rain in a couple of weeks," he says. "The database can't tell me when it might rain."

Like other precision farmers, Robertson may have data at his fingertips but he'll always have dirt under his nails.

At A Glance: Rubenacker Farms

Headquarters: Rural Route 1, Dahlgren, IL 62828
Phone: (618) 648-2257
Business: Crop farming.
Top Technologist: Kelly Robertson, agronomy manager
Size: 9,000 acres for wheat, corn and soybean crops; also runs irrigation tiling and trucking operations.

NEXT PAGE: Business Intelligence Then and NowNow">

Then and Now

Agriculture: Business Intelligence
10 Years Ago Today 5 years from now
Seed, fertilizer and machinery costs recorded on paper. Costs tracked in computer spreadsheets and business analysis software. Wireless networks in rural areas enable use of handheld computers for tracking and analysis.
Tractors, combines and sprayers driven by farm workers. Heavy machinery guided by global positioning systems, with drivers in the cab monitoring the activity and, usually, turning the steering wheel at the end of each row. Driverless, robotic machinery with onboard computers pre-programmed with 3D visualization maps of the fields do all plowing, planting and harvesting.
Planting decisions made on farmer experience and instinct. Experience and "farm sense" supplemented with statistical analysis of historical trend data. Computer analysis and farmer experience further refined to include variables such as near-real-time data on international market demand and pricing.