Final Report for FNE12-757
Our goal was to integrate state of the art GPS precision farming equipment into our organic farm and use it to improve weed control. We believed that while rtk GPS operated tractors for cultivating weeds had been a failure on most farms that tried it, that we had identified the source of most of the inaccuracy that made it unsuccessful. We proved to be only partly correct in our assessment of why it had not been accurate enough to guide tractors to cultivate weeds. We also proved that we could use our GPS rtk system to record the location of our rows and then use that information to guide the tractor with a cultivator down the rows with a high degree of precision.
Unfortunately, the added complexity and demand for accuracy of using our GPS rtk system to guide the cultivator resulted in greatly increased problems and glitches in the system causing us to have too many problems and too much down time to make using the system for guiding our cultivator practical.
We discussed our observations and problems with the GPS system with several large farmers both organic and conventional who had had similar experiences. This was a very valuable exercise for us. I was very fortunate that our son Peter was able to understand the technical documentation and company trouble shooting guides so that he was able to determine the causes of many of our problems. I believe that the real problem with currently available precision agricultural products that use GPS rtk technology is not that he equipment is not accurate enough. It is. It is in the lack of technical support and insufficient training of sales people the manufacturers and their dealer networks provide.
We found that once our system was properly adjusted and debugged, it was able to guide a corn planter and grain drill precisely enough to eliminate the need for markers. Our grain drill has 5” row spacing which demands very precise driving skills. The rtk controlled tractor was guided accurately enough to eliminate the average 1 foot overlaps that were needed to avoid gaps due to driving errors with the grain drill.
We are now using GPS rtk for managing our contour crop strips to eliminate overlaps and gaps in the strip boundaries. When we farmed with chemicals, the inaccuracies in our strips were annoying and often caused us to waste trips and materials in the fields when we farmed them, the losses were small and at that time, unavoidable. With organic management, the inaccuracies in our strip boundaries have become not only annoying but very expensive as well. Even very well laid out strips had an average of more than 5% crop loss due to cultivator damage in row overlaps and wide spots which quickly became infested with weeds.
In corn and soybeans during 2012, with high grain prices and good yields, the lost production on our strip edges averaged over 5%. With total crop value exceeding $2000 per acre, 5% losses due to the relatively small inaccuracies in strip boundaries cost us up to $100 per acre in lost production. In tillage operations, the amount of overlap is generally ignored as a cost factor however, it is not uncommon to make one extra pass in each 150’ strip to compensate for the accumulated driving errors made in the tillage. With our 24’ wide harrow, that equals about one extra acre of coverage for each 8 acres we farm. With a cost per acre of $22, that adds $2.75 of additional cost to each pass of tillage we make.
Smaller and midsized organic farms have been somewhat slower to adopt some of the tools of precision farming than larger commercial operations have been. The first barrier to adoption of this technology (GPS rtk) has been the high cost of the equipment, which must be spread over large acreages in order to be profitable. Finding applications that produce higher per acre returns is one way that smaller farms could benefit enough from adopting GPS rtk to make it become a profitable investment.
One potential high value application of this technology is to use it to guide a cultivator. If successful, this would be of great value to organic farmers. Rtk GPS systems claim to be accurate to within an inch. That is accurate enough to control a cultivator at least as well as a skilled operator can and would greatly reduce level of skill required to achieve good mechanical weed control and it would substantially reduce operator fatigue.
Earlier attempts at cultivator guidance had failed largely because trailed planters commonly drift as much as 8” from the center line of the tractor introducing a large deviation between the theoretical location of the rows and the actual one. This error is in addition to the anticipated best case 1 inch error at planting plus the one inch error in cultivator guidance.
We eliminated the error from equipment drift by mounting a second GPS unit on the planter and recording the actual location of the planted rows hoping that this would improve performance enough to allow it to guide a cultivator with enough accuracy to control weeds effectively.
In addition to cultivator guidance, we identified improving the accuracy of contour strips as an important potential use for rtk GPS technology and reducing wasted trips with tillage equipment as another way that this equipment could pay for itself.
With the capability for mapping that our GPS system gave us we were able to produce detailed maps showing the exact locations of field boundaries and the number of passes in each field. This information allows us to produce planting plans and to reduce the number of wasted passes we make with tillage and planting equipment. This has led to substantial savings.
Our overall objective was to show that we could use an rtk GPS system to improve the management and profitability of our farm and other farms like ours. We identified 5 main ways that this equipment may be able to improve our farm’s profitability.
1) Field mapping and planning field operations
2) Reduction of wasted inputs and passes with equipment
3) Guidance of planting equipment to eliminate the need for markers and to improve spacing accuracy to reduce wasted inputs
4) Maintain the locations and accuracy of our contour strips to eliminate gaps and narrow spots between strips with increased weeds and lost production
5) Accurately guiding a cultivator for improved weed control
The last objective required the most resources and a complex plan of work. The first milestone we had to reach was to record the exact path of the planter in the field to produce a map to guide the cultivator by.
The second milestone was to use the recorded planting data to control the cultivator and make it shift so that the machine stayed matched to the row.
The second milestone proved impossible to achieve because of limitations in the software and the hydraulic interface. This required us to develop an alternative strategy and set a third milestone which was to use the planting data to auto steer the cultivator tractor in the field accurately enough to allow the tractor operator to manually adjust the cultivator’s side shifting hitch.
The fourth milestone was to operate this equipment to cultivate our corn and beans accurately enough to improve our weed control.
To meet our objectives, we installed a Lyca GPS rtk system on our corn planter and also had it set up so that it could be removed and mounted on our other tractors after the planting was finished. The data we gathered during planting was also used for mapping the fields. The highly precise GPS maps we generated in the first year were used to construct planting maps and to correct our crop strips so that the boundaries matched with each other.
This allowed us to plant the crop strips at the exact locations where they had originally been planned and so that the alternating strips that were to be planted to a different crop would fit so that when they were planted at a latter date each matched closely enough so that there were no wide gaps or rows that the cultivator would hit. With our 30 inch row spacing, we had to be a minimum of 18” from the next crop row and a maximum of 40”. We were able to plant our strips within those tolerances.
We mounted the GPS receiver in our larger tractor for use during tillage operations. This was necessary for keeping our strips in the exact locations where they were intended to be and allowed us to control the amount of overlap that occurred during secondary tillage to prevent wasted trips across the field due to un-necessary overlap.
We used the GPS guided tractor for spreading seed, compost and other soil amendments to reduce wasted inputs and follow our nutrient management plan to see if we could save on inputs by reducing waste.
The most demanding work of this project was to use the recorded planting maps to later guide the cultivator in the field to control weeds. Initially this activity failed. The interface between the GPS and the hydraulic cultivator control that we needed to accomplish this did not perform satisfactorily and the company did not offer the needed technical support to fix it. This forced us to find another way to accomplish our goal and we asked for and received a one year extension on the grant to do so.
We planted 3 acres of dry beans in late July of 2012 to allow us to do some additional work on this experiment rather than having to wait until 2013 to resume. This planting was strictly for the experiment not for producing a crop. We successfully mapped the planting and the GPS autosteer guided the tractor between the rows with the cultivator. With the tractor operator free from having to steer the tractor, he easily operated the sideshift on the cultivator to keep it centered on the rows.
In the second season, (2013) we bought a second GPS rtk unit with the intention of running the planter tractor with auto steer from the planting plan and fully eliminating our planter markers while the first unit recorded the exact path of each planter pass.
We also planted our corn in alternating strips leaving the spaces in between for soybeans. When we returned to the unplanted strips 3 weeks later, the rows were accurate enough to eliminate both the gaps and the overlaps and edge rows that were damaged by the cultivator.
As with any new technology, the learning curve in integrating GPS equipment into a farming system is very steep at the beginning. In addition to the work we performed as part of the original grant proposal, I interviewed a large number of other organic and conventional farmers who had purchased GPS equipment with the intention of using it to guide a cultivator to achieve better weed control. I found that our experience with this equipment was similar to that of other farmers.
The biggest single complaint was that service and technical support for this equipment was inadequate and that the equipment proved unreliable. The second most common complaint is that the GPS equipment they purchased was far less accurate than they had expected it to be.
When considering the purchase of GPS equipment farmers are often presented with 3 available steps in accuracy starting with basic low cost uncorrected entry level GPS with an accuracy of about plus or minus 3 feet. This, farmers are told is sufficient for spreading seed, fertilizers, manure, etc or for spraying herbicides in fields where there are no rows or other visible navigation aids for the operator to guide by. The next level up is a corrected signal where a satellite is used to broadcast a basic correction signal along with the GPS signals. These are described to be accurate within 6 inches. The third level up is what we call “rtk” or “real time kinematics” rtk has a an on site generated correction signal or a ground based network such as the COORS system generated correction signal. We discovered that there are many other factors that determine the actual accuracy of a GPS system on the farm.
Several people trying to sell us their products had told us that these were relatively conservative advertised accuracy ratings and that at times we could expect much better performance much of the time. What we learned from our work with this project was that these rated accuracy figures actually represented the best possible performance when the many variables that could reduce performance were all optimized.
We found that there were many important software adjustments and settings that had to be made by the dealer and then carefully maintained or upgraded as upgrades become available. If these upgrades are not installed as they become available and if the adjustments are not kept up, performance degrades and problems such as lost signal, slow startups, and reduced accuracy begin to build up. Several farmers I talked with complained about dealers delivering equipment and simply not checking it for accuracy or not even completing the startup adjustments after it was installed. Some examples of adjustments or upgrades that can affect performance are being linked to the wrong correction tower. This would greatly increase the distance from the tower and increase the correction error. Not having the needed access codes loaded. Most GPS receivers can receive signals from both the American satellites and from the Russian GLONAST system. That doubles the possible number of positioning points and really improves accuracy when the American satellites are very low on the horizon. Without the proper access codes installed, the GPS receiver shows that it is picking up the Russian satellites but can’t use them.
The other factor responsible for less than rated accuracy was the hardware itself. If a GPS rtk system was sold with for example with a lower grade antennae to be more competitively priced, it reduced maximum possible accuracy to below what the GPS itself was capable of. If it was installed with any other lower grade components, each would further reduce the system’s performance.
For more information on factors that affect the accuracy of GPS systems See the two documents attached below:
While the advertising statements and sales literature may technically be true, they are very misleading to customers who are not well versed in technical language. To be fair to the equipment sales people, they may not understand the limitations of their products or what the accuracy ratings actually mean in the real world much better than the customers do.
An accurate understanding of this subject was very important for the achievement of our goals. Each of our goals required certain minimum standards of accuracy to be successful. Without adjusting the expected accuracy to reflect reality and including all factors that modify performance we could not expect to achieve them.
In addition to problems with accuracy, we encountered reliability issues. This was another area where product support was inadequate for controlling a cultivator. When interrupted service is not corrected quickly, the operation can’t wait until service is restored to resume.
In conclusion, we found that our system was adequate for achieving our first 4 objectives: 1) Field mapping and planning field operations 2) Reduction of wasted inputs and passes with equipment 3) Guidance of planting equipment 4) Maintain the locations and accuracy of our contour strips
Unfortunately, it was not adequate to reliably guide our cultivation equipment.
We feel that we accomplished a great deal more than we expected in some areas but were disappointed to conclude that our most anticipated goal was premature for the current state of GPS rtk products.
It is important for farmers considering the purchase of this type of equipment understand its limitations and go forward with an accurate understanding of what they are buying, what they can reasonably expect to be able to do with it and why.
We showed that GPS rtk can be a profitable investment for even smaller to midsized farms when used to reduce wasted inputs, extra tillage passes, and lost production at strip edges.
It can produce large returns especially for organic farmers with rolling land who have contour strips, conservation structures and other practices that reduce soil erosion. The ability to maintain the location and width of strips with a high degree of precision is of great potential value to all farmers and to soil conservation engineers who work with them. The value to organic farmers is greater because of our reliance on mechanical weed control and the accompanying weed problems and crop loss wherever rows don’t maintain uniform spacing or where they overlap.
Education & Outreach Activities and Participation Summary
Outreach was an ongoing part of our GPS project. We kept a number of other farmers closely informed of our progress on an ongoing basis as we did our grant work. I presented observations we had made in doing our project at a workshop at the MOSES conference in LaCrosse Wisconsin in Feb of 2013.
I have continued to include progress updates from our grant work at workshops and farmer meetings throughout the past 2 years. While none of the workshops I presented at in 2013 were entirely devoted to our findings from this grant work, It was and remains an important means of disseminating the information we had gathered AND is provided us with important feedback and observations from other farmers that proved helpful in our ongoing work.
After completing the field work of our SARE grant last fall, I put together a presentation that closely mirrored the contents of our 2013 year end report plus the more detailed material regarding limitations in accuracy that my son Peter was able to glean from the company tech sheets that he had acquired.
I presented highlights from our findings at the January 14, 2014 NYCO meeting to about 75 organic farmers and researchers and then presented the complete report at the February 11, 2014 meeting of NYCO.
I delivered a series of telephone conferences for our old order Mennonite friends this winter. These calls were participated in by over 70 farmers plus recordings of them were made for further distribution. This has turned out to be a very effective vehicle for outreach. I continue to receive follow up calls from participants regularly to clarify and expand on what I talked about.
I will continue to present this information at workshops that I present at and will add further data to the presentations I give in the future as I add to my experiences.
I feel that our findings about factors that affect the accuracy of GPS and the reliability/technical problems with this technology are important contributions to share with other farmers. Understanding the factors that affect the accuracy of this type of equipment makes it possible to avoid costly gaps in service and can help farmers to proactively maintain a high level of performance from their equipment. This is information that many farmers I have spoken with really appreciated having.
Having an accurate understanding of the true capabilities and limitations of a product is important for farmers considering a purchase. They are also very helpful in establishing reasonable expectations before deciding on a purchase. Promotional literature and over enthusiastic salespeople often miss-lead farmers into buying products that are not well matched with their intended use. This is especially important when such equipment is used in planting applications to eliminate the need for markers on corn planters or to prevent costly overlap when drilling grains.
On the positive side, we have shown that GPS can generate valuable savings by reducing tillage overlap, planting overlap, and wasted trips across fields. The savings from this use of the technology alone are often large enough to pay back the cost of the equipment in a short time.
On big corn planters, (12 rows or more) GPS is cheaper than a new marker system and with 24 row and bigger planters companies are not even selling most of them with markers anymore. Understanding the factors that influence accuracy and reliability is very helpful for preventing service loss which would stop planting and can greatly improve the accuracy of the spacing between passes.
While using GPS to improve the maintenance and accuracy of contour strips pays its greatest returns to organic farmers who rely on mechanical weed control, the value of using rtk GPS to manage farming operations in fields with diversion ditches, terraces, gully plugs, and all kinds of strip cropping or contour farming practices is substantial.
I believe that the knowledge we gained from this project will be useful for any farmers who are considering the purchase of and those who already own GPS equipment to get better performance from it and to use it for additional applications that help them get the most value from it.
While we were not able to make the GPS controlled cultivator guidance work well enough to be a viable use of the technology, we were pleased and surprised to discover that our other uses of the equipment actually proved more successful than we had originally anticipated.
We have changed the way we farm based on what we learned from doing this grant work and feel that the large initial investment will indeed pay of fairly quickly although not in the way that we had planned.
I feel it is important to share what we have learned with other farmers, especially what we have learned about the limitations of this equipment and the very serious problems we have encountered with lack of technical support and disappointing dealer service.