Nailing Down The Value Of Data
There’s been quite a dust-up in the past few years about who owns a grower’s data and what agriculture technology providers (ATPs) can do with it. It reached the point where the American Farm Bureau stepped in, leading efforts to draw up guidelines for use of producers’ information in 2014. But just what’s so great about data that everyone wants it?
PrecisionAg® magazine reached out to a number of consultants and growers to see how they measure the value of data. They agreed there are monetary benefits in higher yields and saved input costs. Then too, more precise practices and use of products makes sense for the environment.
Our contacts also agreed that yield data was the Holy Grail, the most important information collected. Some called it a score card, a report card of everything that went before in the season. And as such, how it’s interpreted impacts future agronomic decisions — from variable-rate seeding recommendations to nutrient tracking to simply picking the best variety, says Steve Cubbage, President, Prime Meridian, Nevada, MO.
Planting data came right behind yield information in value. As Tim Norris, CEO of Ag Info Tech LLC, Mount Vernon, OH, puts it: “Planting is the most important job we can do as a grain farmer. If we don’t find issues and correct them during planting, we have limited our yield potential right from the start.”
MORE BY LISA HEACOX
In fact, Norris says producers can do much better collecting data at planting. Thanks to displays with Internet connections, I envision a day when growers can automatically collect and record current weather conditions. In addition, sensors on the planter could automatically collect soil temperature and moisture.
Yield data gets more valuable when it’s supplemented with planter ride, downforce, seeding accuracy, and crop input application information, says Cubbage. “We’re not getting the whole picture,” he worries. “That’s a shame because we’re missing out on what could be some easy answers to manage crops.”
But isn’t there “low value” data in an operation? Definitely not, agreed our contacts, provided it’s quality data.
In fact, a few growers are shooting to record everything that happens in a field. Most recently, Grower Steve Pitstick, Maple Park, IL, began documenting the footprints of machine operations in his fields — movement of tillage units, grain carts, etc. The idea is to look at all the impacts of compaction on yield, to correlate it to imagery or to whatever measure he can.
Norris has clients that have been collecting information on both tillage and grain cart trips and sees traffic patterns show up in row-by-row downforce maps. Having these layers of information can help identify what caused the patterns and possible field problems.
Pitstick admits he might not find value in all data collected today, but in the future event that he can, he says that with such stored information, he’ll be ahead of his peers.
Indeed, Nathan Wentworth, Wentworth Farms, Warrensburg, IL, says that looking back, he wishes he had collected certain data sooner. He won’t label anything as low value if he can gather it easily, economically and accurately. He envisions a time when he’ll be able to analyze it and potentially make actionable decisions with it.
Wentworth is also focused on improving how he collects data. “We have made a significant investment into the Precision Planting and FieldView platforms to improve the simplicity and accuracy of our planting and yield data,” he says. “We’ve been impressed with FieldView and are very excited to gather all of our field data through this platform in the not too distant future.”
Quality The Biggest Limitation
While data may be valuable, growers can face real struggles collecting it right. A number of our experts believe the biggest limitation of data today is its quality.
Jeremy Wilson, Technology Specialist for Crop IMS, Effingham, IL, seriously questions the quality of the data collected by the average grower. “We’re creating mountains, and we’ve got more big data solutions that are going to take these mountains and bring us solutions and answers. But if it’s junk, what do we have to gain?” he says. “I think too many big data solutions are not giving any credit to the quality of data they may be getting, and I don’t know how to address it.”
Poor education by the industry ranked high as a reason for bad data. “Growers are promised things that their monitors and systems can produce and all the things they can do with data. But no one teaches them what it takes to produce good data that can actually be used in making decisions,” says Ben Flansburg, BCA Ag Technologies, Medina, NY. Too many times, the customers’ mounds of data he sees aren’t useable, the problems caused because something wasn’t calibrated or parameters were not set correctly.
There haven’t been incentives to do a good job calibrating because growers just don’t know what they’ll do with the data produced anyway, says Pitstick. “Kind of like the chicken and the egg,” he notes. But as systems become easier to calibrate, quality will get better.
Prime Meridian’s Cubbage has found information submitted to firms for analysis is generally unorganized, uncalibrated, and typically missing large chunks of valuable data. “This is the reality on the ground, and growers and agri-professionals are, in my opinion, not addressing this serious issue,” he says. “Properly collecting a foundational data set such as where different varieties are planted, and what population they are planted at, then collecting calibrated yield data at the end is not the norm.” He adds that the problems are seriously holding back the industry, including the potential promised by Big Data.
Other production information pieces — such as rainfall, planting locations and planting dates — needed for decision-making are missing as well. If yield data is the report card, then all that’s gone before is the lesson plan, says Pitstick. But most growers don’t carefully record the lesson plan. “Guys keep track of stuff on a clipboard on a tractor or a calendar in the shop,” he says. “But that doesn’t do much good because it’s not spatial, it’s not anything we can look at on a digitized map. It’s very hard to do analysis on those parts of the lesson plan that aren’t digitized.”