Autonomous Agricultural Spraying UAV: Design, Implementation and Performance Analysis

This article discusses the importance of spraying in precision agriculture to optimize land use, particularly in response to increasing population and declining agricultural land. A six-rotor unmanned aerial vehicle (UAV) was designed to maximize spraying efficiency and minimize waste. The required pesticide amount was determined based on the number of trees in the field, and UAV components capable of autonomous spraying were selected accordingly. Autonomous flight tests were conducted using a color-based object detection algorithm for tree identification. Success rates are calculated by the ratio of color-changing areas in images captured by the thermal camera to the total area. The results indicate that in low-wind conditions, the spraying success rate can reach 92%, whereas in high-wind conditions, it drops to 20%. Comparisons with traditional spraying methods reveal that tractor-based spraying achieves the same efficiency (92%) but requires 1.5 times longer spraying time and twice the pesticide amount. In contrast, hand-pump spraying reaches 97% efficiency but requires 7.5 times longer and consumes 3.5 times more pesticide. In addition, when comparing spraying to be done on large agricultural lands such as 10 acres, in addition to the amount of spraying and water, diesel fuel is added for spraying with a tractor, personnel costs are added for spraying by hand, while only the electricity cost to charge the battery is added for spraying with a UAV. The effect of wind speed on the success rate can be ensured by revising the UAV position after the calculations are made after the wind direction and speed are determined, and stability can be ensured in future studies.