Precision farming is gaining in popularity and is seen by many as the next great revolution in agriculture.....and it will be! Precision farming as practiced by those who can not see the field for the plants has been with us for generations. Ask any gardener of the joy and pride that comes with a well tended garden. Then ask anyone who has hoed a row of sugar beets the meaning of the term 'a-long-row-to-hoe'. The ability to apply "garden" level management to "field scale" production has been the challenge to agricultural producers since the days of our early pioneers. One only needs a genteel stroll though any farms' "machinery bone yard" to realize the efforts extended to help bridge this management gap. To days machinery and technology have come a long way but the challenge remains. Precision farming, that being the ability to better manage small production units on a large scale, will continue to permit greater agricultural production with increased efficiency. To achieve these increases, precision farmers of the future will require more management. Precision farming 'tools' will, however, make this management easier.
During the ongoing development of precision farming technology, today's producers must choose those management practices and 'tools' which fit into one's operations . Some may argue that managers must change to match the 'tools' but wise producers know that maximum production with maximum efficiency is beyond the capability of any one 'tool" or management practice. Choosing 'tools' best suited for one's particular operation can be a challenging task. Exaggerated claims tend to follow the introduction of new technology and are often based on the premise of what the technology "will do". A prudent manager must choose those systems and practices which enhance his operation based on what it "will do for me".
The key to recent advancements in precision farming technology has been the Global Positioning System (GPS). The GPS system is a series of orbiting satellites which transmit signals by which an appropriate receiver is able to compute precise geographic location. This technology opens the door to attach location based information to various agricultural based information sources.
Four years ago, in an attempt to help introduce precision farming technology, Alberta Agriculture, Food and Rural Development (AAFRD) began involvement with GPS based precision farming technology. The intent of those early projects were to establish the requirements for GPS based yield monitoring and variable rate fertilizer applications. Since then, numerous projects are currently underway looking various aspects of precision farming technology. Some aspects include: yield and grain moisture, soil salinity, weed mapping, traction and implement draft, field slope and topography, variable rate applications, machine guidance and control and water management.
Much of the early interest in precision farming technology centered around a GPS receiver, a combine yield monitor and a variable rate controller. AAFRD precision farming projects use Nova-Tel GPS positioning equipment as the basis for gathering location based information. Differential corrections are generated from a base station set over survey control markers or a convenient position near the field location. Pacific Crest radio modems are used to broadcast differential corrections to the mobile receiver. Position accuracies are typically in the 20 cm range when sufficient care is used in setting up the base station. A laptop computer is used to coordinate GPS information with other sensor signals. For harvesting operations, yield sensor and/or manually tagged weed area information is tagged with GPS position. For fertilizing or variable rate applications, GPS position is used as the basis for prescription map data searching and output signal processing. Much of the capabilities of precision farming hardware tools depend on software capabilities of a system. As the technology develops much of the software is being incorporated into the hardware to help make the technology transparent to the operator.
During the last several years that AAFRD has been involved in precision farming, it is increasingly apparent that the amounts of data that can be collected is enormous. Many hundreds and thousands of pages of data can be generated in a short time. Although this may look impressive, it requires considerable effort to sort and verify the data. Numerous errors are inherent in the data gathering process. Yield sensors can be accurate to within several percent but only if calibrated properly and often. Yield sensors typically have a 15-20 second delay from the point in which the crop enters the combine until the yield sensor detects the amount. This delay can be variable depending on the combine and the operating conditions. Yield information is only as accurate as the cutting width. For non-row crop harvesting this can be highly variable depending on operator skill and field layout. Operator errors associated with the harvest operation can cause many opportunities for erroneous data collection as well. Data collection over non-cropped or previously harvested areas requires careful editing to remove unwanted information. Stops and starts as a result of combine plugging cause yield spikes which must be edited.
To be useful in a farm management decision, data has to be as accurate as possible. Sorting and finding erroneous data can be tedious. For yield maps to be useful in the farm decision making process data files have to as accurate as possible. Three-dimensional graphics and GIS packages can produce impressive "yield pictures" but the usefulness of these pictures lies in the ability to make decisions based on the picture. Most farmers know that different portions of their fields produce differently. The key to make yield maps useful is to make the subsequent decisions and operations flexible enough to account for the variations in the yield map. Fertility and weed infestations provided by a tagged yield map are only useful if one can vary the amount of product required at those particular locations. Today's technology is at this point of being able to do this without direct operator intervention. Other field and crop parameters and product applications will continue to be measured and developed over the next few years.
It becomes apparent to those involved in precision farming operations that the equipment and technology to apply variable rate applications will be available. The difficult decision however, is to ascertain exactly how much product should be applied at the various locations. As one gathers site specific information, one quickly realizes that many variables affect production levels. Trying to coordinate the various layers of information into a management decision is no easy task. For example, a sandy clay loam south-facing field slope of 15 degrees had a previous yield of 40 bushels. Wild oats infestations were light but a salinity reading indicated a moderate level of salinity. Manure was applied to the location two years ago and last years rainfall was 13 inches. A fertility test was done but at a location about 300 yards from the location in question. How much and of what type of fertilizer should be applied? The answer is obviously not obvious.
The key to making precision farming a useful tool is to be able to gather site-specific information and then to scientifically provide answers based on that information. Two of the biggest factors limiting maximum crop production are available moisture and fertility. Last year's yield maps are only marginal indicators of present fertility and soil moisture levels. To make precision farming applications useful, today's manager needs this information for each field location. This implies that an on-the-fly moisture and fertility sensor is needed. Unfortunately, these sensors are not presently available and are not likely to be in the near future. Soil fertility and moisture sampling on a grid basis is good interim solution but these tests are costly and tedious. Given a well sampled field, fertility and moisture information can be interpolated to provide reasonably accurate maps. For most farms the cost for determining such information is prohibitive. A next best solution is to perhaps sample locations which are indicative of the more general field locations. "Bench-mark" testing relies on sampling field areas that can be extrapolated to other similar field areas. This method of providing site specific information is probably more realistic for most farms in that it provides useful information at a reasonable cost.
The most important link for making precision farming feasible is to develop the algorithms necessary to use site-specific information in a profitable way. Fertility and moisture information is fairly well understood in terms of their individual effects on crop production. As more information layers are gathered, the effects and the interactions of the variables becomes much more complicated and less well understood. Most scientific studies focus on only several layers of information and in a controlled way. These studies have generated much useful information for the dependent variables tested but become less reliable when extrapolated to include other information layers or variables. Much more research is needed to include the forty or fifty different factors or variables that presently affect crop production levels.
In spite of the fact that agronomic research has not kept up with the change in GPS based precision farming technology, we are making progress. Each bit of gathered information may not be of much use on its own but combined with other layers of information will over time provide the basis for better management decisions. Tools will continue to develop where by data gathering and analysis will be easy and affordable. Accurate crop production algorithms will develop as the data base of information grows. The key for tomorrow's successful producer will be to have a well developed information base from which individualized crop management decisions can be implemented for each small production unit. Precision farming means more management. Knowledge and technology will, however, make the management easier.