Ag technology is becoming more seamless. According to Jess Lowenberg-DeBoer, agricultural economist at Purdue, in order for any ag technology to be economically viable, and in turn become widely adopted, it must be easy to use. In technical terms, it must move from being an “information-intensive technology” to an “embodied-knowledge technology.”

“So what do I mean by information-intensive technology?” says Lowenberg-DeBoer, who has studied the economics of precision farming for 22 years. “I mean that using field-level data to make decisions typically requires you to collect data, analyze it, and put it back in a format that is usable.” Examples of this kind of technology include variable-rate technology (VRT) and Integrated Pest Management. He says technologies like these have high associated costs and relatively slow adoption rates.

Embodied-knowledge technology, in contrast, is purchased in the form of an input and can be used with little or no knowledge or effort. In other words, you don't need to know much about the technology in order to benefit from it. Examples of it are widespread.

“Most of the ag technology of the 20th century was embodied technology,” Lowenberg-DeBoer says. “Examples include hybrid seed, chemical weed control and mechanization. All of those are classic embodied technologies, and they have been very productive.”

Lowenberg-DeBoer says embodied-knowledge technology is the most economically efficient solution for using scientific knowledge to solve agricultural problems because it compensates for skilled labor. “People with the skills to interpret precision ag data and implement the spatial management plans are expensive and hard to find,” he says. “And most farmers did not become farmers because they wanted to sit behind a computer and analyze data. Most became farmers at least in part because they valued an outdoor, active lifestyle. If it is profitable enough, they will spend time crunching numbers, but most would prefer not to.”

Three waves

Lowenberg-DeBoer says precision farming is already in the process of becoming embodied-knowledge technology. And it is happening in three waves.

The first precision agricultural technology to be designed and commercialized as embodied-knowledge technology was vehicle-based nitrogen (N) sensing developed in the 1990s. Examples include GreenSeeker, Norsk Hydro (Yara) N sensor, and Crop Circle. These products measure the greenness of leaves, which relates to the amount of nitrogen in a corn plant, and then vary the rate of N on the go when used with variable-rate application equipment.

The second wave was GPS-enabled vehicle guidance technology that assists with steering. He says this technology is well on its way to becoming standard. “The real technology success story of the last decade is guidance,” he says. According to Purdue surveys with custom applicators, use of GPS lightbars grew from 5% in 1999 to 57% in 2006 nationally, and up to 77% in the Midwest. During the same time frame, use of GPS auto-guidance systems, where the equipment takes over steering, grew from 6 to 20%.

“We've also seen great interest among farmers, although we don't have the numbers,” he says. “Low-cost assisted steering systems that sell for around $5,000 are popular even among those you'd never expect to be interested in higher tech. Included in that group are older farmers with back or shoulder problems who say that assisted steering systems are allowing them to stay in farming a few more years.”

Robots are next

Lowenberg-DeBoer believes the third wave of embodied-knowledge technology will be robotics. “Twenty years ago few people would have imagined the current market for GPS guidance,” he says. “If you would have told U.S. farmers that there would be equipment to take over steering of the tractor or combine, most would have said, ‘What do I need that for? I can steer. And I can do a good job at it.’ But the market has seen tremendous growth, and farmers are finding uses for it. I think we may see the same thing with robotics.”

Article continued on next page >

Robotics, he says, are highly developed forms of embodied knowledge in which a device or group of devices performs autonomous tasks under the control of a programmed computer. These tasks could include crop scouting, weeding, irrigation, microspraying or harvesting.

He says that, by reducing labor, robotics may facilitate certain crop management practices that farmers in the U.S. have shied away from simply because of high labor costs. “Organic production is a classic example,” he says. “One of the great barriers toward organic production in the United States is that it requires more labor. With robotics, we may be able to deal with that. Another example may be intensive soil sampling. It's a very repetitive task. Sounds like something a robot could do.”

With robots doing much of the labor, VRT could become more profitable. “Present adoption of VRT depends on making precision agriculture embodied-knowledge technology that the user can simply buy as an input,” Lowenberg-DeBoer says. “I think GPS vehicle guidance will stimulate crop management creativity and improvise the next steps.”

Robotic weed control

THE ILLINOIS Council on Food and Agricultural Research has awarded the University of Illinois $300,000 to develop a special robot for weed control. Agricultural engineers at the U of I already have developed several ag robots that can guide themselves down corn rows using lasers or sensors to gauge the distance to corn plants (see “Robot farmers — cheap,” July/August 2004). U of I ag engineer Tony Grift says the new robot in development will be equipped to detect weeds and apply herbicides while navigating itself up and down the rows.

“The plan is to develop a robot that is smart enough to distinguish weeds that are glyphosate [Roundup] resistant from the nonresistant ones and then use mechanical weed control for the resistant weeds and use chemicals in minute amounts [microspray method] for the rest,” Grift says. “The main weed is waterhemp, which has been shown to be resistant in Illinois through three mechanisms, and more will follow.”

The proposal to develop the “high-efficiency, flexible, intelligent farming tools,” calls for two ag engineers, a geneticist, a systems integrator and a weed scientist. Estimated date of completion is the summer of 2009.

“This robot would be in the $5,000 price range,” Grift says. “We will contact potential manufacturers to produce this type of robot. And hopefully there will be a market, especially once more weeds become resistant in light of the cost of developing a new Roundup, which is astronomical.”

Economics of VRT

VARIABLE-RATE TECHNOLOGY (VRT) using GPS, first developed in the 1990s, allows you to vary the rates of seed, pesticide and fertilizer according to the needs of each area of a field. Those needs are determined by analyzing geo-referenced data that have been collected on your fields. By tailoring the rates, as opposed to using one uniform rate, across a field, you can apply inputs more efficiently and theoretically cut input costs and increase yields.

However, according to a review of more than 200 articles on the topic, conducted by Purdue University, VRT does not always result in an economic benefit. “Most studies to date have shown only modest returns to site-specific variable-rate applications of crop nutrients,” says Jess Lowenberg-DeBoer, agricultural economist at Purdue. He says the reason is because the gains are usually offset by the labor, sampling and equipment costs associated with site-specific management.

But what about the recent rise in fuel and fertilizer prices? To find out whether higher input prices balance out the associated costs of VRT, Lowenberg-DeBoer and colleague Bruce Erickson used earlier studies on VRT profitability and plugged in 2006 prices.

One study in particular, conducted in 1999 in Indiana, measured the net profit of applying phosphorus (P) and potassium (K) at one uniform rate and at varying rates based on 2.5-acre grids and on soil type. “In the 1999 study, varying P and K rates by soil type resulted in a net benefit of $4 to $5/acre over a single uniform rate application,” Lowenberg-DeBoer says. “Varying the rate by 2.5-acre grids showed the least benefit.

“When we plugged in 2006 prices, all three methods were less profitable, which was not surprising because costs today are higher,” he says. “But the ranking was still the same. In other words, varying rates by soil type was slightly more profitable than uniform rate technology, and the three-acre grid was still the least profitable.

“The higher energy and fertilizer prices do make variable application slightly more attractive than in the past,” he continues. “But prices are not high enough…to tip the scale conclusively to the variable-rate technology side. Whether [higher prices] will dramatically change the situation for VRT is still unknown. The associated costs of doing it will have to come down.”