Advances in yield mapping tech

Such advances, however, didn’t happen in isolation. As our collective computing capability continues to increase at an exponential rate, GPS and yield-mapping technologies today look nothing like their early predecessors.

Little resemblance between old and new yield mapping systems

The first yield mapping technology was produced by the company Ag Leader, becoming commercially available to farmers in 1992. However, the company didn’t add a GPS component until several years later, largely because GPS technology was itself in a comparative infancy at the time – says Karon Tracey-Cowan, a 30-year veteran of using, marketing, and training farmers and agronomists on precision-agriculture technologies in Canada.

“We were just starting to merge location with that information. The equipment development somewhat parallels the improvement over time of measuring distance,” she says.

Differential GPS

Generally speaking these improvements came after the United States military – the original developers of GPS – rescinded the right to occasionally scramble signals in the name of national security. The launch of stationary satellites under the WAAS system also allowed for more secure secondary connections (already known as “differential GPS”), which helped correct spatial anomalies.

“All of a sudden we were in a whole new way of seeing maps,” Tracey-Cowan says. “We would be anywhere from one to three metres out in the mid to late 1990s. We basically dropped to sub-metre once we had WAAS.”

_{Text continues underneath image}

The first yield mapping technology was produced by Ag Leader, becoming commercially available to farmers in 1992. This is the Yield Monitor 2000, with a screen (upper right) for GPS yield monitoring. This is a harvest of maize (Zea mays) in Illinois, USA, early nineties. – Photo: Hollandse Hoogte / Science Photo Library

Struggles associated with early yield-monitoring systems

Improvements in other data collection systems were also major factors. Indeed, she reiterates the struggles associated with early yield-monitoring systems – such as inaccurate spatial readings and field maps, or a general inability to use data in any meaningful way – have largely been erased as computing technology has advanced by leaps and bounds.

_{Karon Tracey-Cowan:}

The technology is always improving, but to compared what we have now to what we had in the 1990s, its not even the same planet

“It wasn’t just about equipment. There’s an agronomy component. All these types of data gathering were happening in parallel,” she says. “It’s not about gears and leavers anymore. It’s the computer driven management of those systems, and they are easy to manage. You just have to learn how to use them. The technology is always improving, but to compared what we have now to what we had in the 1990s, its not even the same planet.”

Advances in accuracy and calibration ease

So what does this different planet look like? For Greg Kitching, another veteran crop advisor and technology consultant with Premier Equipment – a sizeable equipment dealer and ag-service provider in Southwestern Ontario – accuracy and data processing improvements have been the most significant changes over the last two decades.

“I’ve seen it evolve. We’ve gone from plus or minus five per cent accurate to plus or minus one per cent,” says Kitching. “Current systems we can be quite accurate.”

_{Text continues underneath image}

Cabin of the MY 2021 John Deere X Series combine harvester. The new Integrated StarFire 6000 Receiver and Generation 4 Display provide harvest data. – Photo: John Deere

Shift to self-calibrating yield monitors

Aside from more refined GPS technology, that accuracy derives in part from a shift to self-calibrating yield monitors. This is in contrast to more traditional systems requiring repeated calibrations, and which are more susceptible to user-error. Kitching says John Deere was a main driver of this change, though it has been widely adopted by other large equipment manufacturers.

“It’s become more common. A barrier to accuracy was the grower having to calibrate at the beginning of season and for all these different crops. With these systems that’s no longer necessary,” he says, adding the popularity of yield monitoring systems continues to rise as the technology is now standard in new equipment.

Retrofitted systems are popular too, and have been for some time. Ag Leader’s original system, for example, was designed to fit on existing, GPS-free machinery. Other companies also offered similar hardware at the time, and more continue entering this space. One such example is FarmTRX, a Canadian company producing a comparatively inexpensive, simple to calibrate yield-monitoring system based on optical sensors.

Correcting overlap

Modern guidance systems also contain what Tracey-Cowan calls a built-in “autoswath” feature, which reduces swatch width in increments as it enters a previously harvested or otherwise worked area, such as headlands. This makes for more geographically refined data.

“As soon as the monitor passes into an area already harvested, it knows that spatially and adjust accordingly. If the monitor is of an older vintage, this feature may not be available and the operator has the option to manually change swath width,” she says.

Older monitors still have settings to mitigate overlap inaccuracies but are just not automated. “On most newer models, say 2012 and newer, this is not an issue unless the operator complete ignores the basic set up steps like entering row or header width. If a user follows the very basic setup and calibrations this does not have to be an issue.”

_{Text continues underneath image}

The first yield mapping technology was produced by Ag Leader. Today it is installed on many makes and models at the factory and aftermarket yield monitoring kits install on most combines made in the last 35 years. – Photo: Ag Leader

Advances in data processing and handling

Gathering accurate data is one thing, but doing something with it is another. Indeed, a historic criticism and barrier to adoption for yield monitoring systems has been the complexity of data processing. To this end, Kitching says equipment manufacturers have also put a lot of work into new software platforms.

“The software to deal with the data has improved tremendously. Deere has invested significant dollars in that development because, since it’s in their equipment, it’s important for them to have the tools to use it.”

Tracey-Cowan reflects a similar statement. Improvements in GPS, sensor, and general agronomic technologies mean higher-quality data is delivered in greater amounts. As a consequence, artificial intelligence (AI) itself has and continues to improve.

“We have better platforms to allow us to use and combine information for a clearer picture. The biggest changes now are probably less with hardware than with software, the visualization of that information, the transfer, and the analysis,” she says.

_{Text continues underneath image}

Gathering accurate data is one thing, but doing something with it is another. Indeed, a historic criticism and barrier to adoption for yield monitoring systems has been the complexity of data processing. – Photo: Jan Willem Schouten

Combing metrics

Manufacturers are now combining a variety of sensor technologies within precision systems, including but not limited to yield-mapping technologies. John Deere Harvest Lab, for example, adds sugar, starch, crude protein, dry matter and other metrics to actual yield tonnage using an infrared sensor. More information is also being accumulated during planting (e.g. organic matter and soil temperature).

As these additional data sets are collected, Tracey-Cowan says growers, agronomists, and nutritionists are working to evaluate and work with these new ways to assess variability in the field.

_{Karon Tracey-Cowan:}

Nothing is ever ‘final’ when it comes to technology. There will always be improvements over time

“The important thing to remember is, because we are collecting dynamically, GPS data allows us to see where these values are coming from spatially, supporting how we treat those areas of difference for optimized production. It’s a marked improvement to have multiple data at the same place and time to capture those interrelated factors,” she says, later adding barriers to adoption might be listed as cost, demand for such additions, or the mistaken idea that a system has yet to be perfected.

“Nothing is ever ‘final’ when it comes to technology. There will always be improvements over time. Most developers are looking at all ways to improve the ROI of their equipment so additional data sensors is often an incrementally easy add-on.”

_{Text continues underneath image}

X Series Combines are available with integrated precision ag technologies that collect important machine and yield data while helping to improve harvesting results. – Photo: John Deere

More powerful AI and predictive response

So what’s next for yield-mapping tech? Tracey-Cowan believes the algorithms currently being used to measure and combine data are still in their infancy, and the propulsion of AI through increasing volumes of information mean “the sky is the limit.”

Better algorithms, that is, can correct current inaccuracies that come from combining multiple layers of information – some of which, such as soil-type data, are less reliable than others. Optical sensors themselves require advancement and calibration too.

They need to be ‘taught’ how to interpret the readings they produce and often are seen as relative readings, not absolute or not identical to a lab tested procedure. But these imperatives are part of the development of the systems and are showing great promise for efficiency,” says Tracey Cowan.

“There’s still some disparities between information collection, analysis, and delivery. As we improve one thing it allows us to improve the next thing. Together they will give us those next tools, but it all comes back to computing. What we see in five or ten years will be as different as the yield maps we saw in 1996 compared to today.”

Advances tied to developments in satellite and drone imagery

Kitching agrees, and adds universally applicable algorithms, as opposed to siloed or proprietary systems, are another possible area of growth. Advances overall are also tied to further developments in satellite and drone imagery.

Predictive technologies that can anticipate changes in crop quality or other field characteristics are another area of current investment. “I expect the next generation will be software that instead of being reactive it will be a proactive AI system anticipating a change and reacting to it,” he says.

Ag Leader YM2000 – first commercially successful yield monitor launched in 1992 With the advent of the first commercially successful yield monitor, the Ag Leader YM2000, in 1992 – the mapping and software components were unavailable at that time. So, the YM2000 was helpful for the purpose of instantaneous yield and moisture at that time. The YM2000 would also give a field average yield and moisture reading. GPS commercially technology available in 1996 GPS technology became commercially available in 1996 and with a GPS receiver installed on the combine, the YM2000 could now record a location with a yield and moisture value. The YM2000 successor, the PF3000 was released in 1998, and featured an on-screen map showing field boundary and coverage mapping. Text continues underneath image The Ag Leader YM2000 came out in 1992. It was helpful for the purpose of instantaneous yield and moisture at that time. The YM2000 would also give a field average yield and moisture reading. – Photo: AG Leader In the summer of 2000, SMS (Spatial Management System) was released, allowing farmers to take the data recorded from their YM2000 or PF3000 display andsee the map. SMS would print yield maps, see load or region totals and provide some basic analysis to help aid farmers in understanding correlations of management decisions and yield. Ag Leader InSight display – the first color touchscreen yield monitor The next yield monitor evolution came in 2004 with the Ag Leader InSight display, the first color touchscreen yield monitor. Now farmers could readily see a color keyed yield and moisture map for easy visualization on the go. In 2005, when Ag Leader introduced the SeedCommand planting products for the InSight, growers could now plant with the InSight, log varieties and locations of hybrids and in the fall, get Variety Tracking in the combine. This allowed the growers to record and map the yield and moisture of varieties separately for a quick and easy comparison. Other tasks like Autosteer, Direct Liquid and Dry application control, swath section control to prevent skips and overlaps were also added throughout this time, and are still used today. AgFiniti Mobile iPad app Since then, the InCommand families in 2015 brought with them an AgFiniti Mobile iPad app wich allows syncing with the iPad for the grower to take yield and moisture maps with them as soon as they leave the combine. This is very handy for on-the-go decision making in-season when early order seed purchasing incentives are high. Text continues underneath images The AgFiniti Mobile iPad app allows syncing with the iPad… … allowing growers to take yield and moisture maps with them as soon as they leave the combine. – Photo: Ag Leader InCommand’s DisplayCast feature released in 2017 allows multiple InCommand displays to share guidance lines, coverage logging, AutoSwath section control, yield totals for a field, moisture content and more. This information aids farmers running two or more combines in a field to keep everything together. Additionally, DisplayCast shares a live map of the combine’s progress and location through InCommand displays, helping the grain cart operator know when and where the combine needs to unload. CartACE For fall 2020 Ag Leader is releasing CartACE, an AgFiniti and InCommand feature that shares the combine’s last pass guidance line with the grain cart tractor automatically when it’s within range. The tractor’s steering will then engage on that line and make unloading on the go much easier. Finally, a new Grain Tank Reset sensor works with the InCommand display in the combine to create a Live Grain Tank Counter, which can be shared with connected displays and AgFiniti Mobile using DisplayCast, helping the cart operator know how full the combine is to know when to move up to the combine and start unloading.