Data is nothing new to agriculture and, specifically, farming operations. Historically, hand-written notes have represented the data used by farmers to evaluate their operation and yield. With the arrival of precision agriculture (PA) in the early 1990s, however, farmers began using technology to collect site-specific, electronic data. As PA technology has evolved, internet connectivity and cloud technology have enhanced the access and portability of data collected on farm machinery and through mobile applications.
Most farm machinery today comes from the manufacturer connected to the internet using telematics or wireless technologies, allowing data to be moved on and off machinery easily. Furthermore, original equipment managers (OEMs), agriculture technology providers (ATPs), and retailers provide software platforms to store, visualize and analyze farm data. Additionally, mobile applications (i.e., APPs) are widely used on farms. This technological access allows farmers and their consultants to use data to support decisions within farm operations on a field-by-field basis.
Today, machines and other on-farm technologies can generate large volumes of data, coupled with cloud platforms that allow the data to be stored, accessed, visualized, and shared. These advances make it necessary for farm operations to develop a digital strategy, which is important from legal and data usage perspectives. This fact sheet, the second in a series of three (FABE 555-556-557) to look at digital strategies, details the information a farm needs to get started on developing its unique digital strategy.
The first steps for developing a digital strategy involve documentation:
- Identify the PA technologies (i.e., the hardware on equipment) being used on the farm.
- Identify the types and format of the data collected by these technologies.
Data collected for use in agricultural settings comes from a variety of sources and can be used for multiple applications. Several types of data currently used in PA are outlined in the next sections. A good, initial resource for reviewing common types of data is AgGateway’s “Data Privacy and Use White Paper” at go.osu.edu/AgGatewayWhitePaper.
Most, if not all, site-specific data files have geo-referenced points with attributes or variables that relate to field operations or characteristics of the field. Most of these points have more information than can be visualized in one single map. Elements collected at each point often include date/time, GPS location, application rate, elevation, ground speed, product, and more.
PA data can be placed in one of several types of data categories including agronomic, machine, prescription, remote sensed imagery, production, and others. Figure 1 gives examples of the data that would be included in each category.
Figure 2. An example of agronomic data. This map was generated from as-planted data and indicates the planted population of a cornfield. (Collected during field research by the Ohio State University Extension Digital Ag Team)
Agronomic data represents data compiled at the field level that is related to agronomy-based information. Types of this data include hybrid (or variety), planted population (as shown in Figure 2), yield, soil type, soil texture, soil fertility, pesticides applied, fertilizer applied, scouting information, and more. Data generated from a yield monitor connected to a GNSS receiver can document yields spatially across a field. As-applied and as-planted data represent agronomic data that can be collected during field operations. Soils data can be used to support fertilizer and regional environmental compliance decisions. Scouting data can be used to track pest and disease pressures and help make spraying decisions.
Many farm machines in use today collect crop and soil data in new and notable ways. Sensors, imagery, and other technologies work collaboratively to provide farmers with details about soil nutrients, moisture, weeds, insects, sunlight, shade, and other factors. When analyzed, the data can help farmers adjust their practices during the planting season, thus providing a more rewarding harvest.
Several agriculture companies provide services that collect data from farm machines and pair it with other data sets, such as weather, productivity potential, harvest yields and satellite imagery. Soil moisture sensors, for example, can help farmers monitor the need for irrigation and, in some cases, automate the process. These types of sensors allow companies to advise farmers on inputs, application rates and practices that can maximize yields and profits. Recommendations are then provided to the farmer in the form of a prescription file for input into the machine’s display.
|Figure 3. Illustration of ground speed (mph) as an example of machine data during planting. (Collected by the Ohio State University Extension Digital Ag Team using field machinery and a mobile APP)
Machine data is compiled using multiple sensors and communication networks located on agricultural machinery. Most of this data relates machine data to the information that can be collected from the controlled area network (CAN) on machines and implements. These include fuel usage, engine RPM, engine load, ground speed, slip, gear selection, and much more. Machine data can also indicate whether the autosteer is engaged, GPS location, machine heading, implement status, 3-point hitch status, and hydraulic pressure and flow rate. Machine data can be an effective tool to evaluate operating costs and field capacity (ac/hr) when coupled with telemetry/wireless technology. Figure 3 illustrates a map of ground speed as an example of machine data. If you have a new tractor, combine, cotton picker or sprayer, machine data can most likely be collected, especially if using a telemetry option on the machine.
Remotely Sensed Imagery
With the recent surge in drone use, imagery of fields has become relatively easy to collect. Imagery gathered in the visible light spectrum can be used to view soil patterns, drainage patterns, subsurface tile locations, and more. However, relying on visible imagery alone limits farmers to what they can see with the naked eye. By also collecting light outside the visible spectrum, such as infrared, a new perspective on the field can be viewed and analyzed.
One of the more commonly used imagery examples in agriculture includes the normalized difference vegetation index (NDVI). NDVI is a numerical indicator that is useful for distinguishing vegetation from soil by recording the reflective light that falls on it. One of the benefits of using imagery is the ability to develop variable rate prescription maps for seeding in the absence of a yield map. Additionally, remote sensed imagery can help identify man-made variability, aid in the collection of data for on-farm research, provide yield estimates, and direct scouting activities (stand counts, leaf tissue collection, crop health assessment, etc.).
Production data includes all other data, including farm data, notes, weather data, application dates, planting dates, etc. Production data is useful to support and supplement other forms of digital agriculture.
Hurdles with Farm Data
The data learning curve can be steep for farmers. Two major issues include data compatibility and file formats, and both should be considered in a farm’s digital strategy. Additional details are outlined below.
Data Compatibility - Enables any data from any compatible data system to be simply integrated with the data from any other compatible data systems. More simply, it allows data to read or be translated by any system.
File Formats - A standard method for data to be stored in an electronic file. File formats can be either proprietary or free, and either unpublished or open.
- Metadata – A set of data that describes and gives information about other data.
- File Elements – Elements (characters, fields, and records) that make up an electronic file. Simply, it is the information stored within the file
A current focus within the digital agriculture community is achieving interoperability – the ability of a system or a product to work with other systems or products without special effort on the part of the user. One example would be connecting the data stream between a “Brand A” tractor and a “Brand B” planter. Many companies have their own proprietary standards and formatting for data. Efforts should thus be focused on standardizing, formatting, and developing ways to transport information so it can be used to its fullest potential. Currently, lack of interoperability is a major limiting factor for digital agriculture.
Efforts to address the interoperability of farm data, and, specifically, the need for standard formatting, terminology, etc., are being made by multiple organizations and groups. AgGateway, has developed a glossary (agglossary.org) comprised of terminology from across the agricultural industry. AgGateway has also created the Ag Data Application Programming Toolkit (ADAPT) in an effort to eliminate points of difficulty between hardware and software applications (adaptframework.org).
Additionally, there are several global groups dedicated to standardizing farm data storage and sharing. Many companies are working together to support the ISO11783 standard to ensure efficient tractor-implement communication plus data file storage by terminals. The Agricultural Industry Electronics Foundation (AEF) is one independent organization working to improve such cross-manufacturer compatibility (aef-online.org).
A variety of data formats are commonly used by companies. Many are proprietary, requiring specialized software to view and use. Proprietary formats (.dat, .gsd, .rbin, and .agdata, for example) are used for storing and exchanging data between field machinery and farm management software but can be an obstacle to on-farm data use. Some open-source file formats such as .txt, .shp, and .xml are available and can make it easier to use data stored in these formats. But many commonly-used data formats are not interchangeable, making it a challenge to view and analyze data within one’s software package. The tables below list many of the common file formats that can be used for storing agricultural data.
|Extensible Markup Language
|Rich Text Format
|.ver, .gsd, .gsy, .JDL
|.agdata, .ilf, .yld
|.vyg, .vy1, .yld
The volume and variety of data that can be generated by farm machines and PA technologies for an individual field today is significant. In fact, annual data can easily exceed one terabyte (TB) per acre for farmers who have invested heavily in PA technologies and who use digital tools such as APPs and remote sensing. It is therefore crucial for a farm’s digital strategy to outline the technologies that generate data. Clearly identifying what data is being collected, while also understanding the file format in which it is being stored, is important. Many technologies store data in a proprietary file format, requiring the use of either the company’s software or a farm management software package to upload, read, and use the data. Listing the types of data and file formats within a farm’s digital strategy helps make data usable and valuable to a farm.
The authors would like to thank the following for their time and efforts in reviewing this publication: Dr. Ajay Shah and Dr. Sami Khanal, Department of Food, Agriculture and Biological Engineering Department, Ohio State University; Ben Craker, Kuhn North America; Deb Casurella, MyAgData; Jeremy Wilson, EFC Systems; Christopher Zoller and Jason Hartschuh, Ohio State University Extension; Joe Luck, University of Nebraska; and Bruce Erickson, Purdue University.
Follow the Ohio State Twitter page @OhioStatePA and the hashtag #DataIntel for information related to farm data and its value.