These days, it's not out of the question for implements to draw on three sources of power from a tractor — the drawbar, PTO and hydraulics. Take modern planters, for example.
"You might be feeding part of the hydraulics off the SCV [selective control valves] valves on the tractor. But you might actually throw another hydraulic pump on the PTO to get some more. Of course, there's some drawbar loading just to drag it through the field," says Roger Hoy, biological systems engineering professor at the University of Nebraska-Lincoln and director of the Nebraska Tractor Test Laboratory. "In that case, you're really taking power out of all three outlets simultaneously."
However, Hoy notes, tractor testing hasn't yet advanced to test all three power sources simultaneously. At the moment, NTTL tests each, but tests them individually.
So, a new project, funded by a four-year, $472,887 grant from the National Institute of Food and Agriculture's Agriculture and Food Research Initiative, aims to advance tractor testing methods to accommodate newer equipment. Testing all three, what's referred to as mixed-mode testing, will first require tests in the field to determine optimum loads to start at. It's not the first time someone's developed a mixed-mode test, but one of the first mixed-mode tests, the PowerMix test developed in Germany, was developed specifically for Europe.
"What is that percentage of hydraulic, versus PTO, versus drawbar from the engine gross power?" asks Santosh Pitla, UNL biological systems engineering assistant professor and project lead. "To answer that question, if we have a load cart that can apply all of these powers, how do we set up those percentages? For different operations, it's going to be different."
That's why Pitla and Hoy are starting out by working with farmers in the field, collecting CANBUS (control area network bus) data, or messages, from the tractors through data-logging device. However, CANBUS data is calculated, not measured. So, they're also using pressure transducers to measure hydraulic pressure drop, flowmeters to measure hydraulic flow rate, an instrumented drawbar equipped with a strain gauge to measure force exerted on the drawbar, and a meter to measure PTO torque. Using this information, they can calibrate the data from the CANBUS.
Another kind of big data
However, the data collected from the CANBUS can be unwieldy, and needs to be sorted. And it isn't as simple as pulling an exact figure for hydraulic flow rate, for example.
"We collected CANBUS data for 10 minutes, and there was around 1 gigabyte of data, if you record everything on the bus," Pitla says. "That's why we need to figure out what are the messages we need before we even start the data collection."
This means filtering the messages to receive only what's needed. This involves using message identifiers for components like engine fuel rate and engine torque. For a few of these measurements, Pitla notes the message identifiers have already been determined with the help of protocol like Society of Automotive Engineers J1939 — a communication protocol for off-road machinery CANBUS systems. For others, however, like hydraulic flow rate and pressure, there is still some testing need.
"This project is different. Because in big data, we talk about agronomy, we talk about nitrogen application rates, we talk about seeding rate. But with this, we're talking about machine data, and that's also going to be a big factor," Pitla says. "If you think about driverless tractors of the future, this is another layer that's going to be important. Right now, there's an operator doing everything. But if there's no driver, I need to keep track of the engine fuel rate, whether the robot is loaded or not. How do we know all that? We need to get all that data from machine data."
After testing in the field, different loads will be tested for different implements at the Tractor Test Lab.
Matching up appropriate equipment
The end goal, Hoy says, is helping farmers improve their bottom line by matching the right tractor with the implement. It could potentially involve software or a smartphone app that provides recommendations to farmers based on tractors they have available. For example, the fuel efficiency is typically at its best when the most power is being exerted. So, using an overpowered tractor for lighter field operations wouldn't be the most fuel efficient.
"Let's say you're looking for a row crop tractor and you're going to pull a planter with it, an anhydrous toolbar with it, some grain carts during harvest, and maybe you want to pull a deep ripper with it sometime," Hoy says. "You put all that in, and you'll get an answer back for a fairly large tractor. Then you say, 'I'm not going to worry about deep ripping.' Maybe that 8370 turns into a 7000 Series tractor. So, you've reduced purchase cost, and you increase the likelihood of operating that tractor the most efficiently."
"From some of the very preliminary data we collected from the CANBUS from field operations, we have seen producers oversizing their tractors," Pitla says. "I've seen some anhydrous applications where it's only a 50% load."
"A tractor by itself is a pretty useless piece of equipment. It's only useful when it's delivering power to an implement. So, drawbar, PTO and hydraulics are primarily the methods today," Hoy says. "We're not really focused on engine power. That's not power the farmer can access. Applying those correct loads to represent real-world applications is the real goal."