Using computational intelligence and UAVs to reduce drift on the application of pesticides

Abstract

The use of pesticides in agriculture is essential for production of quality and in high quantity. Spraying these products from aircraft provides greater speed to this process and prevents soil compaction. On the other hand, meteorological factors (e.g. wind speed and direction) may hinder the application of pesticides in the area of cultivation target, causing the drift of the product to neighboring areas. It is believed that most of the pesticides sprayed suffer drift to outside of target area and that only a small portion is actually effective for the control of pests. Therefore, if the agricultural spraying is performed with greater precision is possible to decrease the amount of pesticides used and increase the quality of the cultivation process, besides decrease environmental impact. With the proposed a set of concepts that can be used in distinct architectures, this Thesis Doctoral proposes a Framework for intelligent spraying of pesticides based on Unmanned Aerial Vehicle (UAV) in the areas of agricultural cultivation in order to optimize the application coverage. The concepts proposed in this thesis will be evaluated on a case study using an architecture composed by a rotary wing UAV and a Wireless Sensor Network (WSN) deployed around the field crop. Preliminary results demonstrate that the use of techniques to adjust policy, that corrects the route of the UAV to weather conditions, provides a spray more precise than using static parameters defined empirically. (AU)