By Yiannis Ampatzidis and Sandra M. Guzmán
Pests and diseases, nutrient deficiencies, tree stress from environmental conditions and other factors can be overwhelming. The arrival of sensors and small unmanned aerial vehicles (UAVs) make it possible to gather large amounts of data to assess risk factors and estimate possible crop performance scenarios based on the availability of resources. At the same time, there are computer and cellphone-based tools for the practical management of this data. Detecting, counting and assessing individual trees in groves allows appropriate horticultural practices such as the timely application of chemicals and precision irrigation scheduling to be implemented.
This large amount of data can be utilized to ease the process of assessing and identifying crop management strategies. Technologies including artificial intelligence, big data analytics tools and cloud computing can be used to generate recommendations based on the real-time status of the field. In the long run, these technologies help the field be managed more accurately, more rapidly than with manual labor-based processes, and more cost effectively by optimizing repetitive tasks and potentially increasing production.
The simulation of human intelligence processes by machines, particularly computer systems, is known as artificial intelligence (AI). Artificial intelligence systems work by learning from a massive amount of training data, analyzing the data for correlations and patterns, and then using these patterns to predict future statuses. Many cell phone apps we use for daily life have AI-based systems in the backend. In citrus, for example, by evaluating millions of samples, an image recognition technology may learn to identify and characterize objects in pictures.
MAKING THE MOST OF DATA
The University of Florida Institute of Food and Agricultural Sciences (UF/IFAS) Southwest Florida Research and Education Center (SWFREC) has developed a cloud and AI-based application called Agroview to process, analyze and visualize data collected from UAVs and other platforms (e.g., satellites, smart sprayers and fertilizers). Among other features, this interactive and user-friendly application can:
- Detect, count and geo-locate trees and tree gaps
- Estimate tree height and canopy size (plant inventory)
- Assess individual tree stress
- Determine plant nutrient concentrations and create fertility maps
Agroview converts aerial and ground-based data into practical and useful information. For example, Figure 1 shows a large citrus production area with specific information for a selected area (outlined in red) displayed in the window on the left. This information includes the total number of trees and gaps (spaces with no trees), the UAV flight date, the total size of the area, average tree size and density, and average plant nutrient concentrations.
Figure 2 shows an individual field and provides information for every tree (e.g., tree height, canopy area, leaf density, etc.). The blue dots represent gaps. Gaps are places with no trees; growers had to remove dead or non-productive trees because of HLB.
Agroview can also be used to estimate plant nutrient concentrations and create fertility maps with discrete management zones. An example of a fertility map for potassium can be seen in Figure 3. Five zones (Deficient, Low, Optimum, High and Excess) were determined based on UF/IFAS guidelines. The figure depicts the range of values for each zone.
The tree ratio for each zone is the number of trees divided by the number of tree spaces. If the tree ratio is 100%, it means that the entire zone has zero gaps. As can be seen in Figure 3, the tree ratio for the Deficient and Low zones is lower (60.8% and 82.6%, respectively) than the Optimum and High zones (89.9% and 87.8%, respectively) in this grove. This indicates that this area of the grove might be problematic and require different management. Similar patterns can be observed for nitrogen and phosphorus.
The main advantage of this UAV- and AI-based technology is that large populations of citrus trees can be assessed quickly and at a low cost while reducing inaccuracies that result from sampling a small subset of plants. The AI model developed with this technology had an error of less than 15% for most of the nutrients analyzed. Researchers expect to improve the accuracy as they continue evaluations. This new technology generates prescription maps for the variable rate application of fertilizers based on UAV imagery.
This novel platform can serve as a digital twin of an agricultural field and solve several problems that the citrus industry faces. The first problem is the regulatory requirement to provide a complete tree inventory to obtain crop insurance. The traditional methods are labor-intensive, slow, inaccurate and expensive. Agroview’s patented AI provides accurate, ground-truth data points such as tree counts, tree gap counts, tree height, tree canopy size and leaf density.
The Agroview platform is a fundamental change in the way a grower can interact with their groves. With Agroview’s growing ability to survey fields and assess plant growth and health in high temporal and spatial resolution, growers can see each growing season evolve in high definition. Agroview can compare one growing season to another, meaning growers are able to calibrate their strong intuitions about what’s happening in their fields with plant-specific accuracy.
ASSESSING IRRIGATION NEEDS
One example of the tools available for citrus water management is IrrigMonitor (Figure 4). This decision support system was developed at the UF/IFAS Indian River Research and Education Center (IRREC). IrrigMonitor combines data from different types and brands of in-field sensors, including soil moisture and weather sensors, for the assessment of the water status of the grove. IrrigMonitor provides real-time information on the tree water needs, and the display can be personalized based on the specific soil conditions of each grove.
The color-based displays allow the rapid assessment of irrigation needs at the soil layers where the roots are mostly located. By making decisions with the top layer, it is possible to introduce high-frequency, low-volume irrigation schedules automatically. This type of irrigation is especially recommended to reduce stress in trees with citrus greening disease.
The bottom layer color-based display allows visualization of excess water applied, nutrients being lost, or water table rise. For example, in Figure 4, the bottom layer has a higher moisture percentage than the top layer, and electrical conductivity increased at the same time in the bottom layer. This means that for a fertigation event, some of the nutrients were lost to the deeper layers of the soil. Graphs are available for more detailed assessments of weather and real-time sensor information.
You don’t have to be familiar with AI to take advantage of its capabilities. The AI-based decision support tools developed by UF/IFAS researchers allow growers to be ahead of the game in terms of optimizing citrus management practices. This could reduce tree stress caused by citrus greening disease and provide a set of competitive solutions under the crop’s current reality. Please consult with the authors or UF/IFAS Extension for more information if you are interested in implementing any of the solutions presented here.
Yiannis Ampatzidis is an associate professor at the UF/IFAS SWFREC in Immokalee. Sandra M. Guzmán is an assistant professor at the UF/IFAS Indian River Research and Education Center in Fort Pierce.
Share this Post