8/31/2013 7:00 AM
By Deborah Jeanne Sergeant New York Correspondent
AUBURN, N.Y. — Farmers monitor crop yield for different reasons, but regardless of why they do it, they want an accurate measurement.
That’s what Emma Long, a Cornell graduate student, is working to discover. Her project’s goal is to analyze the data provided by yield monitor equipment developed by John Deere, which she said is the most widely used brand, to see if it matches the lab results she generates at Cornell.
The hope is if farmers can accurately measure yields themselves, they can get closer to real-time results and spare their operations the hassle of sending samples away to a lab.
Long was one of several presenters at a corn silage field day Aug. 22 at O’Hara Machinery, here.
She was joined by Quirine Ketterings and Karl Czymmek of Cornell’s Nutrient Management Spear Program, Dave Russell of Agrinetix LLC, and Clair Culver of O’Hara Machinery.
The moisture sensing component of yield monitoring equipment debuted with John Deere’s first attempt in 2007. Claas followed in 2009 and New Holland in 2012. But no research indicates how accurate the data is. John Deere’s advertising claims its data is accurate within plus or minus 2 percent. Long wants to find out if that holds true.
Ketterings and Czymmek discussed why keeping track of yields is important.
“The more data we have, the more precise we can be,” Czymmek said.
He added that knowing more can help farmers “do a better job of managing.”
Russell led a session on how John Deere’s equipment works, both in the meeting room and surrounding a chopper outfitted with the yield measuring tools.
To measure yield, the equipment must accurately monitor moisture. John Deere’s equipment uses a near-infrared sensor that combined with a mast flow sensor — a sensor that measures the speed rolls as well as how much they open and close to determine the amount of material going through the machine — located just behind the chopper head, it can measure tons of dry matter per acre. A GPS tracks where in a field the equipment harvests, and a display in the cab helps the equipment operator set up and keep an eye on the monitor.
“It works nice a lot of the time,” Russell said. “But it can get frustrating when it’s not working well.”
Clement Gervais, hailing from northern Vermont, complained that the $1,500 lens on his equipment broke during his first cutting.
“Our experience hasn’t been too great,” Gervais said.
Russell said he hadn’t heard of other farmers experiencing that problem.
Another farmer said he lost three weeks worth of data waiting for a replacement lens and felt that his overall data would be skewed because of the omission.
Culver recommended adjusting the lens so crops passing by it would not scrape against the quartz and scratch it. Rocks and mud, however, are difficult to avoid.
All the presenters agreed that measuring hay using this equipment is harder than measuring corn.
Maintaining equipment and checking how often it’s calibrated can make a difference in how long it lasts and how accurate the data will be. Long said that among the 10 farms she has studied, the ones that calibrate the equipment more often tended to have better accuracy when compared with Cornell’s lab test results.
She said she could spend years studying the myriad of variables that affect accuracy; however, a few tips for calibration such as avoiding turning at headlands, not calibrating when starting out, filling the truck to normal capacity and operating at normal speed, can improve accuracy.
“Take time to do good data collection to begin with so you can use the data later,” Long said. “Take time to preload the variety, crops and fields so you don’t have to take that time while you’re busy harvesting.”
There are several challenges to using field monitoring equipment. For instance, if a farmer labels a field as “North40” but another operator calls it “north40,” the equipment will register two different fields.
Also, at least in John Deere’s case, data is not transferable to a smartphone or tablet computer; just to a desktop or laptop via USB. The company hopes to make that upgrade sometime next year.
Although she does not have statistics yet, Long said that “the moisture sensor is accurate on some farms and variable on others.”
She hopes more farms will participate in her study. Emma Long may be reached at 585-813-7228.