Agriculture business and drones

Unmanned aerial vehicles have quickly entered our daily reality. And if a few years ago a flying drone in the sky gathered the attention of passers-by, nowadays the squeezing of drones is just another component of the city noise. But the drones used by wedding photographers and video bloggers are just the tip of the iceberg. For many logistics, construction, security and insurance companies, UAVs have become the main tool for obtaining data. And the list of companies that find tasks for the drones in their business processes is constantly expanding. For example, some internet providers have recently begun using drones to lay cables and monitor their condition. In fact, in any type of activity that requires fast, detailed and cheap spatial data, there is room for the use of unmanned aerial vehicles.

But while for many spheres a drone is only a promising tool that can improve the efficiency of the company, for the agricultural sector, instead of the question “Do you use drones?”, the following questions come up more and more often: “Which drones to use?”, “How often to survey the fields?”, “What other crop data can be collected?” According to a poll conducted by the Farm Journal Pulse magazine in early 2017, almost one-third of US farmers are already using &UAVs in their fields, and another third is about to begin using them in the coming year. Such survey results speak for themselves, as the owners of small farms are the most conservative representatives of the agricultural industry, who are cautious about the introduction of new technologies. But even if a farmer from the Midwest in the United States is considering buying a drone, it becomes clear that the technology is becoming a truly mass solution not only for large agribusinesses.

The assessment of the world market for agricultural drones is significantly different depending on the company that carried out the research. According to MarketsandMarkets, the market is estimated at about $800 million, and by 2022 it is expected to grow to $4.2 billion. The results of the Zion Research analysis are somewhat restrained and state that by 2021 the market for drones will increase to $2.978 billion. But, regardless of the financial results, all analysts agree that the agricultural drone market is expecting significant growth. Thus, according to the AgFunder, in 2015, the capitalization of companies focused on attracting drones to agricultural production increased by 189%.

Estimation of the market for agricultural drones according to Zion Research

 

Еhere are news about more and more new startups that offer implementation of UAV-technologies in agricultural production. To understand what caused such an interest in the agricultural branch of drones, let’s try to consider the problems solved by the drones in the fields.

 

What can drones do in the field?

Mapping

Agriculture is one of the most ancient types of human activity and in recent years it has become one of the most innovative and technological industries. GMOs, robots, big data, AI are just a short list of technologies used by farmers. The emergence of high-performance hybrids, effective fertilizers and pest control has greatly enhanced the efficiency of agricultural activities. Growth in crop yields (as well as the cost of sowing material and fertilizer) has led farmers to cease to consider the field as a homogeneous unit but began to focus on obtaining the maximum yield from each plant. Mechanized systems of sowing, irrigation and nutrition have long been able to differentiate individual areas of crops or even individual plants. But the use of all these features is relevant only if data is available. And here UAVs enter the game. And compared to the satellite data, they have higher detail and efficiency, compared to of the manned aerial photography, their cost is a definite advantage. These made drones the missing part of the puzzle, the absence of which stopped many farmers from conducting high-yielding, precise agriculture.

Initially, the drones were used exclusively to survey fields in the visible range and build detailed cropping maps on the basis of photographs. Thanks to high resolution, such maps have greatly benefited both in traditional topographic maps and in satellite imagery. And the possibility of constructing digital surface models for the fields has allowed improving the performance of hydro-reclamation and a number of other agro-technical measures.

Static maps in the visible range allow you to view the field from the height of the bird’s eye, but still, don’t give so much information about crops. In addition, such maps are often made only once a season, which does not allow to respond promptly to all the challenges faced by agrarians. Therefore, rather quickly the requirements of the market for agricultural drone shifted from simple mapping to monitoring of crops by means of UAVs.

 

Monitoring

It’s physically impossible to carry out a daily review of the condition of each plant, even on one hectare of the field. And if the area of the farm is hundreds or thousands of hectares then the company has to control the state of crops solely from established norms calculated on the basis of averaged statistical data. The result of this approach is that plants in one part of the field can systematically lose fertilizers and moisture, and in other parts, on the contrary, the amount of fertilizers and irrigation volumes will be redundant. The reasons for this, first of all, will be the difference in the soil cover, which is heterogeneous in different parts of the field, but also the significant impact is also made by the relief, quality of the seed, the characteristics of cultivation of the soil. The solution to this problem may be periodic monitoring by means of UAV, preparation of operational maps of crop condition and adjustment of volumes of fertilization and irrigation on their basis.

In order to carry out this kind of monitoring, the images are used not only in the visible but also in the infrared spectral range, taken with a time interval from one day to one week. Combining these images with special algorithms allows calculating various vegetation indices, the most popular of which is the NDVI. NDVI (Normalized Difference Vegetation Index) is a vegetation index, calculated by combining band values in the red and near infrared spectral ranges and displays the amount of photosynthetic active biomass. In short, NDVI allows you to determine which plants are healthy and which are not. Sometimes such information can save a significant portion of the crop, for example, as described in this article, where the use of the drone and the NDVI allowed the farmer to save the field of sugar beet from the pests in a timely manner.

NDVI image in the DroneDeploy software

 

Crop spraying

If the use of drones for mapping fields and operational monitoring surprises no one, then the involvement of the UAV for spraying the fields is still exotic. Although the effectiveness of such tech is unquestionable, since the drone can more accurately and quickly perform the sprinkling of the required solution, besides, unlike a person or special equipment, the flying device will not damage the soil and plants. The effectiveness of such a solution is manifested in the fact that, in conjunction with the data of operational monitoring, we can provide the most balanced treatment of crops with the use of the minimum amount of chemicals required.

The insignificant use of drones for crop spraying is due to the fact that just a few years ago, the technology of developing drones did not allow the construction of a device that would have enough battery capacity and lifting force to ensure the spraying of large areas of fields. The use of drones to spray small plots was economically unjustified.

Now, with the advancement of UAV technology development, more and more drone-oriented devices are emerging for crop spraying. For example, the DJI AGRAS MG-1 drone can be mentioned, which allows handling 7-10 acres of crops per hour transferring up to 10 kilograms of a solution.

DJI AGRAS MG-1

 

DJI AGRAS MG-1 is just one of many specialized agricultural drones. It is worth noting that often the farmers use “ordinary” drone or their modifications. Therefore, it is worth to look more closely at what distinguishes the “agricultural” drone from all others, read about it here.

 

How to process survey results?

Ideally, the results of a drone survey should be available immediately after the flight, this will allow you to make decisions on the basis of the data without any delays. Understanding this, software developers prefer “cloud” technology to handle the sensing results. For example, the rather popular Pix4D application allows you to get the NDVI index (as well as DVI, SAVI) immediately in the field.

Calculating vegetation index in Pix4D using Parrot Sequoia imagery

 

But for this, you first need to upload all the images to the server, this requires high-speed internet, which is often not available in the field. Similar opportunities are available from a number of other similar services, like PrecisionMapper or DroneDeploy. They provide the entire workflow of surveying the territory from the planning of height and spatial coverage to exporting the captured images and building 3D models of the territory. Thanks to the “cloud” principles of data processing these services don’t depend on hardware capabilities of the computer since the basic calculations are performed on remote servers.

If for some reason, you do not want to transfer all the received survey data to an unfamiliar server, then there is the possibility of using a “more traditional” software that will handle processing of the results on your own PC. As an example, we can mention Agisoft PhotoScan, TBC Aerial Photogrammetry Module, CORRELATOR3D. Unlike “cloud” solutions, these programs are extremely demanding of PC hardware capabilities. For example, the minimum configuration for the CORRELATOR3D is 8GB of RAM and the graphics card not lower than the GeForce GTX 500 series or AMD Radeon R5 series. But even with the presence of a powerful PC, the processing of data, the construction of image mosaic and 3D models can take many hours. Although, in cloud services, processing is not carried out immediately, with a smaller range of available processing process settings.

As for the price of the software, it is comparable to the cost of the drones themselves. For example, the CORRELATOR3D license costs 5,000 EUR and Pix4D 6,500 EUR (although a significant number of such services allow you to get a monthly subscription for a significantly lower price). Speaking of open, free solutions, we can mention OpenDroneMap, whose functionality is generally similar to its commercial counterparts. However, some complexity of installation and lack of a graphical interface significantly narrow the circle of its potential users.

 

Short summary

It is difficult to cover all aspects of the use of drones for the needs of agriculture in one article. Probably, that is why almost every day, news and articles on the use of drones for monitoring crops, the legal aspects of UAV-survey, and the peculiarities of image processing are being published on various thematic resources devoted to both UAV and agriculture. On the one hand, these articles show the general interest of to community, and on the other hand, they themselves increase this interest. Such a “self-heating” market indicates that it will only grow. And maybe even an unexpected expansion will take place leaving those who are sceptical of the technology behind. And so, if you are wondering when to start using drones for agricultural needs, then the only correct answer will be “Now!”