Unmanned aerial vehicles (BSPs) have long been used in farming, especially in Japan, China and the US. In our country, the technology is slowly maturing and generating more and more interest. Farmers are definitely eager to adopt innovative solutions as they look for opportunities to improve the profitability of their production. Digitization and robotization of agriculture is already the order of the day, and the driving force behind the change is the ceiling-piercing prices of fertilizers and crop protection products.
Agricultural entrepreneurs have never been spoiled by fate, but now they find themselves in real trouble. What is currently happening in the world is affecting them tremendously - incredibly high fuel and fertilizer prices are paralyzing their business. It's difficult to adapt to the new situation, with only significant cost-cutting coming into play. How do you cut spending on inputs? Can BSPs somehow help with this?
One way is to implement precision farming methods. By using the latest technologies, farmers can perform precise agrotechnical treatments. This allows them to optimize fertilization and crop protection. Just such data is provided by an unmanned aircraft equipped with a multispectral sensor (camera). Using the reflections of specific light beams from plants, we are able to assess their condition and other important parameters, including the crop's nitrogen requirements. Based on the corresponding vegetation index, we develop a variavle rate application (VRA) map. It is divided into several classes, which are assigned fertilizer rates that the area requires. According to the results of research and our practice, this results in an increase in yields and a reduction in nitrogen consumption of up to 30% (which, at today's prices, is three times better economic results compared to the previous season).
Application maps also refer to the rational use of crop protection products. With conventional methods, once a sufficient number of pests or weeds appear, pesticides and herbicides are applied to the entire field. Threats, however, do not have to be present in every part of it - by determining exactly what area is affected, we can, with a certain buffer, apply agents only where necessary. In this way, we reduce their use by up to 80%.
In the above orthophoto, we present a high-resolution orthophoto of the experimental fields in visible RGB (red, green, blue) colors. The precision of such a photogrammetric product is about 2cm/pixel, so we can already observe anomalies and weeds present with the naked eye. Their colors differ from the crops in the plots. If we add the near-infrared reflection maps, red edge channel and thermal imaging, we are able to classify the pixels responsible for the weeds through appropriate transformations. Once this requirement is met, we have ready material for further analysis and development of zonal or spot spraying maps.
Precision and high resolution data is possible with BSPs - at least for the time being, they will not be replaced by satellites or ground-based sensors. Why? Because drone measurements are relatively inexpensive, fast and accurate. The palette of their applications is, of course, much larger. Using them, we can:
- forecast yields,
- determine the number of seedlings,
- estimate hunting damage and those related to natural disasters,
- Perform spraying or discharge of the biological agent,
- carry out an analysis of the level of decline in plant health resulting from a shortage or excess of water,
- Assess the degree of plant transpiration,
- Develop a map of sowing uniformity,
- optimize the desiccation process
- model the growth of fruit trees, etc.
The possibilities are indeed many, but we must be able to take advantage of them. From our experience, we know that know-how in this type of analysis is the key. Otherwise, we may acquire or process data in such a way that a perfectly inaccurate procedure comes out of our precise one.