6.1   Outreach Education for Precision Agriculture

Investigators

            Timothy S. Stombaugh, Biosystems and Agricultural Engineering, tstomb@bae.uky.edu
            Scott A. Shearer, Biosystems and Agricultural Engineering, shearer@bae.uky.edu
            John Fulton, Biosystems and Agricultural Engineering, jfulton@bae.uky.edu
            Thomas G. Mueller, Agronomy, mueller@pop.uky.edu
            Carl Dillon, Agricultural Economics, cdillon@ca.uky.edu

Cooperators

            Charles W. Byers, Agricultural Economics
            Wendy Stivers, Extension Specialist for 4-H/Youth Development Education

Introduction

Precision agriculture (PA) is rapidly gaining popularity among progressive farm managers.  The concept of optimizing profitability through spatial management is favorable to many agriculturists.  However, the large amounts of data that must be managed and the complexity of spatial decision-making are sources of frustration to those who are using the technology, and a major hindrance to those contemplating adoption of the technology.

Spatial data may be obtained from several sources.  Much of the data are collected at the farm.  These data could include GPS field boundaries, yield maps, field history, and intensive soil sampling.  Farm-specific data can also be obtained from other sources such as the Internet.  For example, the Kentucky Office of GIS (KY OGIS) provides many free spatial data sets on their web site including digital elevation models (DEM’s), Digital Ortho Quads (DOQ’s), Hydrography maps, Land cover maps, and Tiger files.  In addition, satellite imagery is available from a variety of sources for a fee.

Numerous PA software packages are available ranging from simple mapping programs to more powerful Geographic Information System (GIS) engines (Table 1).  Unfortunately, there are many challenges that arise when using these software packages.  Foremost are the coordinates of the spatial data sets.  Spatial data can be projected into a number of coordinate systems such as State Plane or Universal Transverse Mercator (UTM).  Data sets projected to different coordinate systems do not align, and therefore data visualization or analysis using multiple layers is impossible.  Reprojecting a data set is a cumbersome task even when using proper GIS tools.  The appropriate datum and projections must be selected to ensure the spatial data are referenced to the same coordinate.  To make matters worse, the accuracy and format of much of the data found on the Internet is unknown and one must contact the organization providing the source data to obtain this information.

Other challenges arise when using PA software packages.  Some programs, especially some of the more powerful programs, are very complex.  Even the most basic procedures can be quite difficult, especially for new users.  In addition, there are compatibility issues.  For example, some packages will conflict with other packages preventing both from being installed on a single computer.  Many import and export file formats are incompatible, or many programs simply do not support variable-rate application file formats.

Table 1.  Software companies providing agricultural mapping and GIS packages.

 Objectives

The overall goal of this project is to increase adoption and use of PA practices by helping producers learn to use spatial data sets and analysis software packages.  This goal will be met via two specific objectives: 1) to establish procedures and guidelines for importing data to common PA software packages with emphasis on case studies involving the creation of data files at the farm level, and downloading free data sets from the Internet, and 2) to develop and deliver extension educational programs for PA data management and decision making.

Procedures

Objective 1: Establish procedures and guidelines for importing data to common PA software packages

The principle investigators have experience with several of the PA software packages and can provide support and training on these packages.  There are several packages that we have not used, which are viable alternatives for PA users.  Therefore, our first task in this project will be to acquire copies of the remaining software packages and become familiar with their operation and capabilities.

Our next task will be to catalog the spatial data provided via the Internet by the Kentucky OGIS suitable for use by Kentucky producers.  As we do, we will determine the proper projection and data formats required for each of the PA software packages.  If simple procedures cannot be developed to import the data directly from the KY OGIS Internet server, we will reproject the data and provide this data via the Precision Agriculture web page at the University of Kentucky (www.bae.uky.edu/~precag).

Capable undergraduate students will be supported to assist with software installation and data reprojection.  Software will be purchased as required and installed on a dedicated desktop computer.

Objective 2: Develop and deliver extension educational programs for PA data management and decision making

Given the machine-specific nature of PA software packages and GPS equipment, and the amount of data that would have to be transferred for each training session, it is almost impossible to use the UK general purpose computing laboratories to teach PA database development and analysis.  Therefore, we propose to establish a mobile teaching laboratory comprised of portable computers and hand-held GPS receivers.  The mobile laboratory can be taken to different sites to reach a larger audience.  We will establish 10 workstations so that we can accommodate 20 students (2 per station).  Though the handheld GPS receivers we plan to acquire may not provide adequate accuracy for all PA activities, these low-cost devices will be quite suitable to facilitate an educational experience.  The desktop computer used in Objective 1 will also be used to house hard drive images of the portable computers so that they can be easily rebuilt and maintained in a consistent manner.

We will develop three basic educational modules involving PA practices.  These modules will contain basic content that can be tailored as appropriate to a variety of audiences.

Module 1: Mapping and Scouting.  Spatial data must be accurate before it has any value to PA management.  Therefore, this module addresses issues relating to field data collection and accuracy.  Students who complete this module should be able to make informed selections of GPS equipment, and use that equipment to collect accurate and useful field data.  During more detailed presentations, we will discuss GPS interfacing, the data sentences contained in the NMEA strings, and GPS configurations.  We will use handheld GPS receivers to provide hands-on experience with the following field activities:

1. GPS Basics – What is GPS and how does it work?  What do I need to know to select and use GPS receivers to their fullest potential?

2. Mapping field boundaries – How do I get an accurate map of a field boundary and calculate the area within that boundary?

3. Locating Features – Can I locate field features and obstructions such as waterways, outcroppings, and sinkholes?

4. Establishing grids and soil sampling – What are the different techniques for laying out grids on a field?  What is the best way to collect soil samples in the grids?

5. Scouting – How could I use GPS to delineate pest infestations, wet spots, and other yield-limiting factors?

Module 2: Basic PA software.  The purpose of this module will be to facilitate hands-on experience with the basics of PA software and analyses.  We will choose a software package for use in the training program that has ample capability yet is simple and readily available to producers.  We will make every effort NOT to endorse specific products; rather, our focus will be to build a knowledge base that encompasses all products.  We will use a case study approach to build a spatial database for a mock farm in Kentucky.  Essentially, we will pick a farm location in Kentucky and build a database for that location that includes field boundaries, DEM’s, DOQ’s, Soils maps, land use maps, and yield maps.  The following are specific topics that will be addressed in this module:

1. Importing data – How do I get the data such as my GPS field boundaries and yield maps into the software?  What about manual input of data such as planting varieties and field history?

2. Data Projections – How do I make sure that the scale of all my data sets match?

3. Mapping Techniques – How do I correct errors in maps?  How many colors should I have on a map?  Will the data ranges I select affect the appearance of the map and interpretation of the data?

4. Internet Resources – What kind of information is available via the Internet and how can I get it?

5. Remote Sensing – How can I incorporate satellite and aerial images into my database?

6. Basic Analysis – How do I perform basic analyses such as overlaying soil maps with yield maps?

Module 3: Advanced PA Analysis.  PA techniques become more valuable as producers collect spatial data sets from their fields.  Unfortunately, the increased number of data sets in a database make analysis and data management much more difficult.  The purpose of this module will be to show how to perform more complex PA analyses using commercially available GIS software packages.  We expect successful students to be able to utilize multiple data sets to make decisions about land use, cropping systems, and variable rate prescriptions.  Though the primary focus will be on data management and software functionality, some agronomic discussions will help producers make more effective use of the data they have collected.  We will extend the case study initiated in Module 2 by utilizing that database to make spatial decisions.  Some of the topics that will be covered include:

1.       Geostatistics – Are my sampling densities adequate?  Is my data valid?  What is causing the yield variation?

2.       Land Use and Profitability – Which parts of my field are profitable?  Which should be enrolled in CRP or other programs?

3.       Variable Rate Prescriptions – How do I create variable-rate prescriptions for my inputs?

Each of these educational modules can be tailored to a variety of audiences.  Module 1 will involve a combination of indoor learning and outdoor data collection activities.  Though Module 1 will be a helpful precursor to Module 2, it will not be a strict prerequisite.  Module 2 or similar field experience will be prerequisite to Module 3.  Each module will be initially delivered as a half-day workshop; however, the level of detail can be adjusted to obtain various lengths of workshops.  The content of the workshops will be tailored to a variety of audiences including producers, service providers, agents, and high school groups.

The educational modules will be delivered in several different ways.  Foremost, we will conduct in-service training for Kentucky agents so that they can provide much of the support that farmers request.  These in-services will be conducted in several locations throughout the agricultural regions of Kentucky and will be repeated as warranted by demand.  The mobile laboratory will then be available for county level activities and training for producers.  Upon request from agents or other farmer groups such as the newly formed Kentucky Precision Agriculture Network, training sessions will be provided for producers.

In addition, we will conduct a minimum of two annual training sessions in conjunction with major agricultural events in the state such as field days and conferences.  We will attempt to gain national exposure for the program by providing training sessions at the National Farm Machinery show in Louisville, and reporting results of the program at professional conferences and Southern Region Extension Meetings.

Another important outreach channel will be to high schools.  PA is becoming a larger part of the skill set required to be successful in FFA and 4-H competitions.  We will partner with 4-H and FFA groups in key agricultural regions to provide training for high school teachers.  We will also target several key Vocational Agriculture programs whose students might benefit from PA training and pilot programs for high school students.  Dr. Byers and Dr. Stivers will be the points of contact with the FFA and 4-H programs, respectively.  Both collaborators have given a verbal commitment to participate in this phase of the project.

This project will be conducted over a three-year period with full implementation of all educational modules expected by the end of Year 2.  Dr. Stombaugh will be the primary facilitator of the training sessions.  Drs. Shearer, Mueller, and Dillon will provide technical information from their respective expertise areas, and will assist with some of the teaching activities.  Mr. Fulton will oversee the software evaluation and reprojection activities.

Expected Benefits

The educational programs developed through this project will be a direct benefit to Kentucky producers, agents, service providers, and high schools students.  The portability of the teaching laboratory will facilitate outreach to a wide and diverse audience.  The data sets provided by the Kentucky OGIS will be more fully utilized by KY producers.  University personnel will increase their knowledge base for PA products making them a better resource for agents and producers.  The high school outreach programs will also enhance recruitment activities for the University of Kentucky.

Deliverables

One key deliverable is the mobile PA computing laboratory, which can be used in a variety of venues through out Kentucky.  The results of this study will be summarized in two extension publications.  One publication will detail the necessary procedures to ensure proper formatting of spatial data provided by the OGIS.  Another publication will overview procedures for importing the KY OGIS data sets into agricultural software packages.  Also, some of the OGIS data sets may be reprojected and provided as appropriate from University of Kentucky file service resources.