The advancement of drone technology has seen many emerging use cases including the expanding use of drones in agriculture. The availability of imaging sensors provide farmers with new opportunities to increase crop yields, minimize crop losses, and thereby maximize their profits. Farmers are now using technology that was once reserved for the military to monitor their crops from the air, instead of visually inspecting their crops on foot. The information gained from drone crop imaging provides a larger, and more accurate view of crop health.
Drones can fly up to 400 feet without the need for special clearance from the Federal Aviation Administration. From this low altitude, higher resolution images can be obtained that are superior and cheaper than high altitude aircraft, or satellite imagery. Also, the low altitude eliminates the issues of clouds that satellite imagery has to contend with.
The availability of inexpensive, but high functioning drones can be traced to the advancements in technologies such as micro electro mechanical systems, smaller GPS modules, more powerful computer processors, and miniaturized digital radios. The advancement and accessibility of these components is mainly due to their use smartphones, and the economies of scale of mass produced parts. The key functionality that makes crop monitoring highly useful for farmers is the autopilot, which automates drone flying. Autopilot software can predetermine the flight pattern that will maximize coverage of the crop fields, imaging software can stitch together aerial imagery into a mosaic map. The generated map is key to improving crop performance, and reducing costs. It shows exactly which areas of the crops need further attention. Farmers can spend more time treating their plants, and less time scouting.
Where Drones Can Help
Some farmers are reluctant to utilize drones for their crop management due to initial cost of a drone. When you consider that the cost for a visual walking inspection or an aerial survey is an average of $2 an acre, the return on the investment of a drone purchase can be met quickly. This can usually be accomplished within one growing season, or even less. The information obtained from the drone can then be utilized to reduce operating costs, and improve crop yields.
There are numerous activities that farmer routinely do, that can be supplemented through the use of drones. To maintain crop health, pests, disease, and weeds need to be identified in order to develop a plan to address these problems. The collection of tissue and soil samples are needed to identify fertility, disease, and pest issues. The root structure of plants need to be inspected for signs of compaction, disease and pests. Soil erosion is always a concern, so erosion channel width and depth needs to be measured. Finally crop yields, and population need to be estimated accurately for resource management. These are the points that are important to farmers:
- Confirmation: Frequently checking that plants are growing at the rate expected
- Early Detection: This is key to addressing plant health issues in order to limit the impacts, and provide time to implement a solution
- Fertilizer Planning: Crops seldom grow evenly, and distrusting fertilizer based on plant health, instead of spreading evenly and reduce costs.
Drones are also starting to be used to apply nutrients and pesticides to plants, instead of using workers on foot, or tractors. Aerial spraying is much faster and cheaper than using traditional methods, such as tractors. Crop dusting drones have been in operation in Japan for over two decades, and are finally starting to make their way for use in the United States.
Drone Imaging Sensors
The information that can be gained within the human visual spectrum pales to what imaging sensors can do. The availability of different camera sensors allow drones to provide even more information than what the human eye can observe. The greater the sensors, the more is the information obtained from the acquired images. The following sensors are available:
- RGB (Red, Green Blue): for plant counting, elevation modeling, and visual inspection
- NIR (Near infra-red): for water management, erosion analysis, plant counting, soil moisture analysis, and crop health
- RE (red edge): plant counting, water management, and crop health
- Thermal infrared: irrigation scheduling, plant physiology, and yield forecasting
The Normalized Differential Vegetation Index (NDVI) shows the difference between red light reflected from plants, and near-infrared light. Healthy leaves with an active photosynthesis process absorb red light, and strongly reflect near infrared light. Dead, or unhealthy leaves reflect both wavelengths of light. This property can be used to measure the health of crops. It does not take specialized sensors to obtain this data. A regular camera modified with a simple filter can be made into a near infrared camera.
Drones in Agriculture Challenges
While the prospect of using drones in agriculture sounds promising, keep in mind that this application is still an emerging technology, and as such has a number of challenges that must first be addressed.
Currently drones equipped with the right sensors can support a farmer to navigate to a location in his fields, observe it, and generate statistics regarding the health and status of the crops, but only on a small scale. Under the current FAA regulations, all measurements and observations have to be done using a drone that is within visual line of sight (VLOS) of the drone operator. The problem is that the majority of American farms are much larger than the VLOS distance. It is possible to conduct multiple operations by day, and stitching together a larger map. However, this is costly in terms of manpower and time. Even though the utility of drones in agriculture is apparent, there is always a reluctance to spend income on anything beyond basic needs.
Drones need to have an active pilot, both from a legal and safety point of view. There has to be someone available if something goes wrong. Even if a farmer setup a drone to execute an autopilot flight, he simply can’t walk away to take care of something else. Using a drone in agriculture does not facilitate multitasking.
There is a further problem that estimates of flight durations are usually based on ideal conditions. A software may predict a time of flight estimate over a given area of interest, but in non-ideal conditions it may take four or five times as long. Once the images are acquired, they still need to be processed to extract useful information. An average computer will still take a few hours to analyze thousands of photos.
The Future of Drones in Agriculture
The American crop value lies in the hundreds of billions of dollars per year, so even a small yield improvement will have a significant economic benefit. However, it is still not clear if in fact drones can provide more usable information at a cost benefit as compared to current aircraft and satellite imaging.
The piece of the puzzle that is currently missing is the lack of expertise to analyze the data, compare it with what is happening on the ground, and recommend a proper course of action to address crop issues. There are services that provide imaging data, but an investment needs to be made in terms of time, money, and software to be able to extract meaningful information from the data. This is currently not being done that well from sources who provide imagery from satellites or aircraft. It is not clear how this is going to change when crop imagery is obtained from drones. However, the technological advancements of drones in agriculture is progressing more rapidly than from satellite or aircraft imagery sources.
There are further problems that still need to be addressed. The generation of calibrated images that provide the added value of monitoring crops over time is still under investigation. Image variability is also affected by different sun angles, and the effect of cloud cover. The advancement of drone image calibration will soon address these technical problems, and provide meaningful crop performance metrics.
Drones in Agriculture Summary
Improving crop yield is an ever pressing problem for farmers and always will be. However, for the first time in generations, digital and technological advancements enable farmers to make large improvements in their performance. On a small scale it is important for an individual farmer to profitable. On a global scale, this ties into the problem of feeding an ever-increasing global population. This is further complicated with the issues of environmental impact, the need to reduce water waste, eliminate chemical run off, and carbon dioxide emissions. Digital technology can contribute to addressing these problems to establish sustainable farming processes.
All members of the agriculture system need to embrace new digital technologies in order to make their operations more productive, profitable, and efficient. Large farms can benefit from a range of applicable data to increase the amount of food produced from the same area. Doing so will potentially result in an improved financial agriculture performance, and higher profits. Smaller farms can also benefit from digital technology to improve their crop yields.