Crop-Monitoring Soil Probes for In-Season Decision Making

Under the Canopy

Joel Reichenberger
By  Joel Reichenberger , Progressive Farmer Senior Editor
Raj Khosla, head of agronomy at Kansas State University, inspects a prototype soil sensor in a lab at the university. (Joel Reichenberger)

Raj Khosla has for decades had an idea of what the ultimate potential of precision agriculture can be.

He imagines a drastic reduction in resource use coupled with increased production thanks to the ability to farm a field pixel by pixel, not with guesstimates and folklore, and not even with fuzzy color maps and vague recommendations, but with precise, updated and consistent information and algorithm.

Now, Khosla says he's starting to see technology come to the table that could finally turn potential into payoff, particularly in one project he's working on: biodegradable soil sensors with the potential of providing farmers with reliable, cheap and critical data.

"If we are successful," he says, "the sky is the limit."

Khosla's been a part of one research team that's spent three years working on such devices. Partners include Gregory Whiting, an associate professor of mechanical engineering focused on materials at the University of Colorado, and Ana Claudia Arias, an electrical and computer sciences professor at the University of California Berkeley. The idea is for the probe to be cheap enough for farmers to install dozens -- even hundreds -- per field each season, maybe for just $1 or $2 per probe.

Then, that network of sensors could collect and transmit soil information -- in this case, critical moisture and nitrogen levels.

Finally, perhaps most importantly, it's supposed to vanish after the growing season, literally melting away into the soil.

The team's work is part of a wave of technological innovation that could sweep over American agriculture. It includes a bevy of probes and sensors, even robots, built to monitor crops and soil in season to provide farmers with a meaningful spotlight where they've never really had one: under the canopy.

"There have been advances in technology that could really change what you can do," says Cherie Kagan, a University of Pennsylvania professor and the leader of another academic team developing similar tools. "If you look at information technology, whether that's on the hardware side, or if you think about the data science side, it's completely changed."


The goal of a biodegradable soil probe is easy to sum up.

"Inexpensive," Khosla says, "and rapid in near real time."

There's potential for vast gains over current practices. Aerial drone, airplane and satellite imagery all have their place on the farm currently, along with myriad other ways producers map their ground, from yield monitors to a handful of different soil-sampling options. But, there are many limits, either from timing or, in the case of aerial cameras, by the canopy. Some are expensive, and some are labor intensive.

Khosla dreams of getting much better data faster, in time to solve problems and save crops in a way that makes applications of all sorts more efficient -- getting under the canopy and into the soil.

At Whiting's lab in Boulder, Colorado, a passion to match Khosla's enthusiasm has taken hold.

Whiting wasn't born a farmer; he barely knew any when he started with the project. A farm tool wasn't necessarily in his plans, but now the laboratory he shares with his research fellows and students is blooming, quite literally. Tomato vines climb from every square foot of empty floor space, and maturing corn stalks fall out of storage cabinets reimagined and outfitted as grow houses.

"We're kind of getting into it," he says, grinning.

Exactly what a commercially available probe would look like and be made of is still very much up in the air, and the visions have changed as the technology has been tested.

Some earlier versions, somewhat smaller than the size of a deck of cards, were encased in a thick wax designed to degrade at a predictable rate. The idea was the soil's microbes would lick their way to the center to lead the probe to a catastrophic failure and begin the decomposition of the components just in time for harvest.

More central to the idea is some kind of balsa wood circuitry board printed with natural metal and conductive ink. The data would come from a variety of potential sensors, perhaps including a pair of hair-thin wires that extend into the soil to create and measure an electromagnetic field.

It could then relay data using RFID transponders, perhaps to a scanner passing by on an irrigation pivot or an aerial drone, or through some fieldwide Wi-Fi network.


Only, it's not all so easy.

Their project, financed by a $1.69 million grant from the Department of Energy's Advanced Research Projects Agency-Energy, recently ended after three years, and there is not a product ready for a farmer's field.

The current iteration of sensors tend to degrade in days and weeks instead of months, and their data, particularly for nitrogen, can prove inconsistent even before that.

"We've made some progress forward, answering some questions," Khosla says. "Is it possible? Yes. Is the engineering possible? Yes. But, there are lots of challenges that still need to be addressed."

A search continues for different materials that will be more stable in a variety of soils while proving sturdy enough to last through a growing season.

Still, sometimes with an initially far-fetched idea like a biodegradable soil probe, it's about the lessons you learn along the way.

For instance, Whiting's team learned to extract data not just from sensors it had buried awaiting decomposition but also from that decomposition itself. As the soil microbes gnawed at the conductive ink, the electrical resistance changed in a measurable way.

The sensor can be a sensor.

"That's been one of the best threads we've pulled," Whiting says. "Microbial activity is a proxy for soil health. We can connect and wirelessly send signals related to the amount of microbial activity in the soil. I don't think anybody's had a way to do that before."

Whiting is partnering with teams at Vanderbilt and in the United Kingdom to push that element of the research.

Other experiments have involved small plastic probes that slide into plant stalks to offer constant data, but "plants don't like having a lump of plastic shoved in them," Whiting says. A different, more promising take on that same idea involves a sensor in a substance with the consistency of Jell-O.

One effort, with the group at UC Berkeley, is focused on tracking nitrous oxide emissions, oxygen content and pH levels in soils, and another drops the biodegradability but adds the ability to test for nitrogen (N), potassium (K) and phosphorus (P).

"We feel good about the N and the K, and we're working on the P," Whiting says. "This whole probe project was quite successful and interesting, and, man, it really spawned a lot of new directions for us."

This all hints at a deeper understanding of plants and the soil they grow in.


Khosla can't shake the vision.

"Inexpensive and rapid," he says repeatedly. "Inexpensive. Rapid. Diagnostic. Biodegradable. Nondestructive."

He says it's coming.

"I'm still looking forward to the day we can have diagnostic, quantitative measures of soil moisture and soil nitrate, and instantly," Khosla says.

Because of experiments and visions like his, "under the canopy" is shaping up to be one of the next big areas of focus for farmers.

It's a belief shared by Cherie Kagan at the University of Pennsylvania and her team, Iot4Ag, or Internet of Things for Precision Agriculture, a National Science Foundation Research Center (NSF).

That group encompasses researchers at Purdue University, the University of Florida, the University of California-Merced and the University of Pennsylvania all working on a $26-million grant from the NSF that finances the study for five years. It's being supported and sponsored by several of the ag industry's big names, as well, from Syngenta to John Deere to Corteva.

The research effort aims to fill the same information gap targeted by Khosla using sensors both in the soil and above. The group wants to use tiny versions, small enough to be dubbed "smart chaff" (perhaps tinier than a fingernail), and it wants them spread across the field "in close contact with plants and beneath the soil surface," according to its website.

Early prototypes are being sent to farms this summer in hopes of gathering data such as potassium, pH, soil compaction and even signs of disease.

It's also studying whether the answers go beyond probes, working with swarms of autonomous robots, both aerial and ground-based.

"It's all about how getting at understanding what stresses influence the yield and health and resiliency of crops," Kagan says. "How do we get that kind of data where it's some of the things you want to know but you can't see directly? You see the plant when it's distressed, or you can see signs of disease, but how do we get at the target variables that will help us prevent crops from being stressed, to catch it early."

Success could signal a new age of precision ag, one that could make the current status quo look like an 8-bit video game in a high-definition world.

"All of these technologies, they're very different than what we called precision ag 30 years ago," Kagan says.

"We call it the internet of things, this idea you can get all these signals and get a very deep and different understanding of the crop."


-- Follow Joel on Twitter @JReichPF


Joel Reichenberger