Sensors helping farmers better understand crops
Just over 25 years ago, a mass flow sensor paired with GPS would forever change how farmers viewed their fields. The marriage gave them the ability to relate harvested grain to an area of a field and begin generating
“Because farmers recognized how much yield variation was really occurring in their fields, suddenly precision agriculture took off,” says Scott Shearer, professor, Ohio State University.
About this same time, product development based on interest in sensing soil fertility on-the-go was also gaining traction.
“Our first product was a soil pH meter that measured spots in the field where corn was stunted and stressed,” says Mike Thurow, who founded Spectrum Technologies in 1987. “Through the years, a lot has changed in the industry.”
Today, sensors – both on and off equipment – measure myriad attributes to help farmers maximize yields with minimum resources. Above and below the ground, sensors can determine when more downforce is needed, define when a crop is thirsty, detect a disease before lesions even appear on the leaves, or guide how chemicals are applied.
“Sensors offer more eyes in critical areas,” says Jesse Haecker, global planter, spraying, and nutrient applicator business manager for John Deere. “For example, the 300 sensors on John Deere’s self-propelled sprayers measure temperature, wind speed, ground speed, spray pressure, flow, and changes in terrain to direct chemical applications in varying conditions.”
A Sense for Sensors
Farming based on sensors is growing increasingly important. Having the ability to precisely monitor in-field variability and make decisions based on data is transforming how farmers manage their operations.
“The confidence of having data to make decisions is a huge value point,” says John Gates, VP of product, CropX. “We can basically augment a farmer’s experience so when he encounters an unusual situation that is less familiar, he has a trustworthy, guess-free source of information on how to handle it.”
With Internet of Things (IoT) strategies taking shape, machine and field intel coming from myriad sensors is rapidly evolving. The first question Matt Darr asks around IoT is, “Are we trying to learn about the science of crop production or influence crop production in a single season?
“Being able to predict and react in season is a totally different value proposition than being able to show why an outcome occurred at the end of the year,” says the Iowa State University professor. “Producers have voted with their checkbook over and over. They will pay for outcome-based solutions but are hesitant to pay for solutions that only deliver knowledge without action.”
Haecker says there are definitely more technologies in the market that collect data than there are those making sense of the data. “We have to move beyond ‘The sensor says this’ to ‘What action should I take?’ ”
A farmer’s decision to adopt, Gates says, often comes down to how he can use technology as an enabler to be more profitable. “There are always going to be interesting tools in the marketplace, but what will hit a farmer’s pocketbook right now?” he asks.
It’s something Shearer believes the soil sensing industry has grappled with. “Areas like near infrared reflectance sensors have matured to where we can measure organic matter content on-the-go,” he says. “Unfortunately, to date we have not experienced the same takeoff in measuring soil nutrient levels.”
Darr struggles with the value proposition of field-installed sensors for management decisions. “They have a competing issue of scalability to get to sampling resolutions that matter while also fighting real issues with logistics around placement and physics associated with power requirements in standing crops.”
Innovations like Teralytic’s wireless probe, Shearer says, may turn that around. Its 26 sensors provide detailed soil quality data, including soil moisture, salinity, and NPK at three different depths, as well as aeration, respiration, air temperature, light, and humidity. “Over time, I think it’s going to give farmers a good estimation of the nutrient levels in their soil,” he says.
The biggest leap going forward is going to take place below the ground, says Tomer Tzach, CropX CEO. “There is still much more to learn about factors that contribute to soil health,” he says, adding he too feels that scalability, as well as cost, is an issue.
The end game, Darr believes, is machine-mounted solutions that capture and react to data in real time.
“This makes actionable data walk-up ready for producers and much easier to implement,” he says, adding that while machine-based sensing
is harder than the alternatives, in the end it provides more value opportunities to directly influence the financials of a farm in a given year.
Near term, Trimble’s Wade Stewart surmises the biggest adoption we’ll see is in real-time adjustment. Some examples are the ability to adjust your planting depth on-the-go based on available moisture, having a way to dynamically change tillage based on compaction, and being able to change the angle of your disc based on the amount of residue.
“These are practices that really make the operation better, but they’re done behind the scenes,” says the ag division market manager.
Sensors at Work
The information Brandon Hunt is collecting across his western Kentucky farm not only helps identify efficiencies that lead to higher productivity and profitability but also lowers input costs and optimizes fertilizer use. “I know technology is going to make us better, but it’s a whole lot easier conversation when it is integrated inside our existing system.”
He also does a lot of “smoke tests” to evaluate a technology before he jumps on board. For example, he spent five years testing Trimble’s GreenSeeker on wheat. Starting with a few acres, Hunt added more acres only as the technology was validated. (See “Sensing Across the Acres” to learn how Hunt is utilizing other sensors.)
“It’s about proving the technology is making an operator as efficient as possible, and every dollar invested in the operation is being maximized,” Stewart says. “That is never going to change.”
Comprehending the return on investment for sensors attempting to gain intel on crops can be elusive, Darr says. “In crop protection, we have to produce an ROI that generates enough yield advantage to overcome both product and application costs.”
IoT solutions for identifying weed, insect, and disease issues need to be not only accurate enough to detect the outbreak, but also specific enough to predict whether the treatment will create a positive ROI. “In many areas we simply lack the foundational agronomic expertise to precisely identify these breakpoints between pest pressure and guaranteed positive ROI if treated,” Darr says.
We also have to get to a point where we can execute on a recurring basis, Shearer says. “The one wild card has always been the weather. If I knew what the weather was going to be, it would make it a lot easier to manage nitrogen.”
It’s also why an on-farm weather station has become a critical tool.
“It really empowers a farmer,” Thurow says. “We are seeing more intense rain events, and a thunderstorm can be a game changer. An on-farm weather station gives him a different perspective on what Mother Nature is delivering and how he might alter his plans.”
Connecting the Dots
As sensor technology continues to evolve, it’s unlikely one measurement will be able to answer all the questions a farmer may have. “The power comes from combining information in new ways,” Gates says.
To achieve that, connectivity is key.
“When it comes to connecting all of this information through the internet of tomorrow, our industry has a huge challenge in front of it,” Haecker says. “Sensors that are on equipment or offboard, processing capabilities, and cloud infrastructure … there are a lot of things that have to come together.”
Sensing across the acres
Brandon Hunt’s planters (Case IH 2140, Case IH 2150, and Kinze 3660) are equipped with Precision Planting technology. While data from the planters let him know how well each is performing, Hunt takes the most away from DeltaForce. The downforce system controls and adjusts pressure instantly based on changing soil conditions. This provides the proper planting depth every time regardless of field conditions.
Because of the rolling hills and side slopes Brandon Hunt has to contend with in western Kentucky, staying on his guidance line when planting corn in the strip can be
An implement’s following error can be five to 10 times the tractor guidance error, according to Scott Shearer, professor, Ohio State University. “Depending on the operation, that implement could be wandering from side to side quite a bit,” he says.
By employing TMX-2050 TrueTracker, active steerable hitches, and CenterPoint RTX, Hunt can repeatedly position the planter to within 1 inch every time.
“We’re doing everything we can to keep that planter in the center of the strip 100% of the time,” he says.
However, Shearer notes, implement guidance systems are really not yet a mainstay in ag primarily due to cost.
Weather Station: WatchDog 2900ET
A WatchDog 2900ET weather station feeds data directly iinto Hunt Farms' irrigation control platform.
Center Pivots: AquaSpy
Brandon Hunt has used both watermark and capacitance-based sensors to measure soil moisture on his irrigated acres. Unlike the watermark sensor, which measures soil moisture at a single location, the capacitance-based sensor measures soil moisture and soil temperature every 4 inches along the length of the probe.
“While the watermark sensors, which were placed at 12, 24, and 36 inches, worked well, I was looking for something different,” he says.
A couple of years ago Hunt began using AquaSpy probes on 13 center pivots. AquaSpy is a 48-inch probe equipped with sensors every 4 inches that evaluate moisture, nutrients, and temperature. The probe sends real-time data to Hunt, which he can access through a smartphone, tablet, or laptop.
“There are a couple of things I thought we gathered better by going with AquaSpy,” Hunt says. “First, I get to see the root depth as it grows throughout the season because the capacitance-based sensor sees where roots are wicking away moisture. I know if roots are shallow or if they are deep in the soil like they should be.”
Second, the real-time data is very straightforward. “I don’t have to do a lot of digging or interpolation,” Hunt says. “Because the information is easy to understand, I am able to make more timely decisions on when to irrigate based on what I’m being shown.”
Combine: Yield Data
"We’ve been collecting yield data for a long time. It’s still pretty powerful information because it’s our unit of measure at harvest,” Brandon Hunt says. “Good yield data quantifies all of the things we are doing.”
Yield data is another layer of information that helps determine which hybrids will be planted the following season. It also helps Hunt evaluate the gains made in strip-till and verify how well he did managing moisture probes and his 13 center pivots.
As Brandon Hunt works to improve profitability, strip-till has become a go-to solution. Besides optimizing yield, the practice lowers fuel and labor expenses. There are also agronomic advantages.
Strip-till lets the Kentucky farmer place nutrients exactly where they’re needed. It keeps crop residue in place over much of the soil,
which reduces erosion and increases water infiltration. More organic matter is preserved, and carbon dioxide remains in the soil.
“Water infiltration is really important in growing corn,” says Hunt, who also raises soybeans and wheat. “Wherever we don’t have the water-holding capacity because of shallow topsoil, strip-till has enabled me to make a bigger
The practice also requires a high level of accuracy and repeatability.
“One objective of strip-till is creating a quality seedbed. It is desirable to place the seed along the center of the bed to maximize yield potential,” says John Fulton, professor, Ohio State University. “Research indicates that seed placed at the edge of the strip or off the strip reduces crop yield. So it’s important to use the same AB lines and a high-accuracy GNSS correction service.”
Blending RTK and RTX
Hunt Farms began using RTK and installed base stations in 2003 when it transitioned to auto steer. Today, the sub-inch accuracy ensures his yield-till HDC strip-till machine and planters are precisely aligned, so seed is placed at the center of the strip.
While RTK has proven its worth, Hunt recently began experiencing
signal loss in some parts of his operation. To offset the problem, he subscribed to Trimble’s CenterPoint RTX. Today, he uses both.
“With the satellite-delivered correction service, you’ve got sub-inch accuracy in less than two minutes,” says Michael Bruno, channel program manager, Trimble.
As Hunt works toward banding fertilizer in the strip, reliable accuracy is even more important. “There is a yield penalty depending on where that row is planted in reference to where the center of the band of fertility is,” he explains.
Precise positioning is the cornerstone of precision ag. But how accurate is accurate enough?
“It really comes down to the perception of the operator. Depending on the operation you’re performing, we have recommendations on how accurate you should be,” Bruno says. “Farming is all about margin. What seems like a marginal increase in accuracy really can impact the bottom line.”
For example, if you’re combining wheat and 6 inches of the cutter bar isn’t cutting the crop, it’s not a big deal in one pass. If you make 100 passes over five seasons, it adds up.
“Over time, that waste would have more than compensated for the cost of higher accuracy while making the operation more efficient. You could apply that same example to a sprayer, a fertilizer application, or any piece of tillage equipment. Tillage is probably the biggest one, especially if you practice conventional till,” Bruno says. “As farmers try more accurate services, it’s unlikely they’ll move back down the accuracy chain. Accuracy is addictive.”
Ultimately, if you want to do more intensive management, that’s where higher accuracy will pay off.
“The goal is to get each farmer in the solution that best fits his operation and budget,” says Wade Stewart, market manager, ag division, Trimble.
Revolutionizing Plant Management
Image-based sensors with artificial intelligence behind them is a concept Scott Shearer believes could revolutionize the way farmers manage their crops. See & Spray is the first sign of a tangible application in agriculture, says the Ohio State University professor.
Developed by Blue River Technology, See & Spray brings together cameras, computers, and artificial intelligence to recognize every plant and determine the right treatment for each, continually learning as it goes. With pinpoint accuracy, robotic nozzles target unwanted plants in real time as the sprayer passes through the field, applying a herbicide while avoiding the crop or areas without weeds.
Acquired by John Deere in 2017, the technology currently identifies the differences between cotton plants and weeds of many species and sizes. It has also been tested in soybeans.
Deere wants a broad set of examples in order for the artificial intelligence to learn what a weed looks like and what it doesn’t look like, so See & Spray is currently being tested in the Midwest as well as other parts of the world on additional crops. This rich intelligence will accelerate deployment into several markets, says Julian Sanchez, director, emerging technology, John Deere.
“A system that can be very selective and only hit an area that really needs it can lead to 50% to 60% savings in chemicals alone,” Sanchez says. “Once we get to where we can do it postemergence, we will see a 90% savings.”
“If you look at the value of genetically modified crops, it is in herbicide resistance,” Shearer says. “If John Deere successfully deploys the technology at scale, 50% of the value of genetically modified crops might evaporate. That’s a game changer.”
Once proven, Shearer believes John Deere will move the technology into areas like nutrient deficiency and disease detection.