The dosage sprayed upon per unit area is an important index to measure the effects of pesticide application. Owing to the fact that parameters such as flight height, flight speed, and spray swath can change at any given time, it is impossible to ensure a consistent amount of pesticide application per unit area during the course of aerial variable spray. In order to ensure a consistent amount of pesticide application per unit area, a set of control models of aerial variable spray using an unmanned aerial vehicle (UAV) was proposed, and the corresponding control system was developed based on the technology of aerial variable spray. According to the change of flight parameters, this system was able to adjust the opening degree of solenoid valve through the control model of aerial variable spray. After that, the amount per unit time would change to ensure a consistent amount of pesticide application per unit area, which effectively avoided the phenomenon of uneven pesticide application and improved the accuracy. According to the actual demand for the area in need of pesticide application, the operator can manually control the amount of pesticide applied and change the dosage sprayed upon per unit area to achieve a better effect. Through field tests, it was verified that the system has high accuracy of variable control. The deviation range was between 0.11% and 9.79%, which met the demands of agricultural aviation pesticide application. Furthermore, the system had strong stability for working continuously for more than 6 h at 30°C to meet the environmental requirements of pesticide application via UAV. All the data related to the pesticide application were stored in this system, which provided a reference for the further study of the precision technology in pesticide application. The model proposed in this paper also provided a theoretical basis for the technology development of aerial variable spray.

Purely soft actuators often suffer from poor grasping stability, low load capacity, and non-adjustable stiffness. To address these limitations, this study presents a pneumatic variable-stiffness soft actuator enabled by a laterally arranged fiber-jamming layer. Under an externally applied vacuum (negative gauge pressure), the fiber contact network is compacted to realize reversible stiffness modulation through frictional fiber jamming, without relying on geometric/structural interlocking. This mechanism also alleviates typical drawbacks of existing jamming media, including crease defects in layer jamming and particle migration/loosening in granular jamming. Based on the principle of virtual work and a Hertzian contact assumption, analytical models are developed for actuator bending and equivalent stiffness. Bending, stiffness, comparative variable-stiffness, and object grasping experiments are conducted to validate the proposed design. The results show that, under random perturbations of the driving pressure (randomly varying within 50–60 kPa), the peak-to-peak fluctuation of the bending angle is reduced by 78% compared with a single actuating layer (pneumatic chambers only). When the jamming-layer gauge pressure decreases from 0 kPa to −80 kPa, the tip equivalent stiffness increases by 220.5%. Under matched structural dimensions and pressure ranges in a controlled benchmarking setup, the stiffness modulation range is improved by 25.24%, 59.44%, and 19.36% relative to layer-, granular-, and vertically aligned fiber-jamming actuators, respectively. Meanwhile, the repeatability index (RSD) is reduced by 20.64%, 41.85%, and 10.78%. These results demonstrate that the proposed actuator provides a large stiffness-tuning capability with good repeatability, thereby improving grasping stability.