In recent years, with the rapid development of science and technology, agricultural robots have gradually begun to replace humans, to complete various agricultural operations, changing traditional agricultural production methods. Not only is the labor input reduced, but also the production efficiency can be improved, which invariably contributes to the development of smart agriculture. This paper reviews the core technologies used for agricultural robots in non-structural environments. In addition, we review the technological progress of drive systems, control strategies, end-effectors, robotic arms, environmental perception, and other related systems. This research shows that in a non-structured agricultural environment, using cameras and light detection and ranging (LiDAR), as well as ultrasonic and satellite navigation equipment, and by integrating sensing, transmission, control, and operation, different types of actuators can be innovatively designed and developed to drive the advance of agricultural robots, to meet the delicate and complex requirements of agricultural products as operational objects, such that better productivity and standardization of agriculture can be achieved. In summary, agricultural production is developing toward a data-driven, standardized, and unmanned approach, with smart agriculture supported by actuator-driven-based agricultural robots. This paper concludes with a summary of the main existing technologies and challenges in the development of actuators for applications in agricultural robots, and the outlook regarding the primary development directions of agricultural robots in the near future.

Soil surface roughness (SSR) is an important parameter affecting surface hydrology, erosion, gas exchange and other processes. The surface roughness of the farmland environment is directly related to the tillage process. In order to accurately characterize the random roughness (RR) parameters of the surface after ditching, a three-dimensional (3D) digital model of the surface was obtained by laser scanning under the conditions of an indoor ditching test, and the influence of oriented roughness components formed by removing ridge characteristics on the RR of the surface was analyzed by introducing the wavelet processing method. For this reason, four groups of ditching depths and two types of surface conditions (whether the surface was agglomerated or not) were designed in this paper. By comparing the root mean squared height (RMSH) and correlation length (CL) data calculated before and after wavelet processing under each group of tests, it was concluded that the RMSH values of the four groups before and after wavelet processing all change more than 200%, the change amplitude reached 271.02% under the treatment of 12 cm ditching depth, meanwhile, the average CL value of five cross-sections under each group of ditching depths decreased by 1.43–2.28 times, which proves that the oriented roughness component formed by furrows and ridges has a significant influence on the calculation of RR. By further analyzing the roughness value differences of clods and pits in different directions and local areas before and after wavelet transform, it was shown that the wavelet transform can effectively remove the surface anisotropy characteristics formed in the tillage direction and provide a uniform treatment method for the evaluation of surface RR at different ditching depths.

In order to solve the problems of high missed rate and easily damaged seeds in traditional garlic metering devices, an air suction garlic seed metering device was designed. The critical components include a seed tray and a seed disturbing tooth. Seed contour curve fitting and a discrete element method were used to determine the initial parameters. The optimal parameter combinations were obtained through fluid dynamics simulation. With a higher pass seeding rate and a lower missed seeding rate as the goal, the device has been optimized for parameters. The best combination of parameters was obtained: the number of type holes was 9; the diameter of holes was 7.2 mm; the forward speed was 1 km/h; the working negative pressure was −5.5 kPa; the pass seeding rate was 88.54%; and the missed seeding rate was 6.34%. Field trials had shown that the air suction garlic seed metering device designed in this study satisfied the requirements of garlic cultivation. Three tests were conducted and a pass rate of 87.83% and a missed rate of 6.85% were obtained.

Soil block distribution is one of the important indexes to evaluate the tillage performance of agricultural machinery. The traditional manual screening methods have the problems of low efficiency and damaging the original surface of the soil. This study proposes a statistical method of farmland soil block distribution based on deep learning. This method combines the adaptive learning rate and squeeze-and-excitation networks channel attention mechanism based on the original Mask-RCNN and uses the improved model to identify, segment and distribute statistics of the farmland soil blocks. Firstly, the influence of different learning rates and an improved Mask-RCNN algorithm model on training results were analyzed. Secondly, the effectiveness of the model in soil block identification and size measurement was analyzed. Finally, the identified soil blocks were classified accordingly, and the scale problem of soil block distribution after removing edge soil blocks was analyzed. The results show that with the decrease of learning rate, the loss value of model training decreases and the prediction accuracy of model is improved. The average precision value of the improved model increased by 25.29 %, and the recall value increased by 8.92%. The correlation coefficient of the maximum diameter measured by manual measurement and the maximum diameter measured by model algorithm was 0.99, which verifies the feasibility of the algorithm model. The prediction error of the model is the smallest when the camera height is 40 cm. Large-scale detection of soil block size in an experimental field in Hefei, Anhui, with an average confidence of over 97%. At the same time, the soil block is effectively classified according to the set classification standard. This study can provide an effective method for the accurate classification of soil block size and can provide a quantitative basis for the control of farmland cultivation intensity.

Traditional water–fertilizer control systems often suffer from poor precision and slow response, limiting precision agriculture development. This study developed an electrical conductivity (EC) control system for water–fertilizer integration using a fuzzy Proportional-Integral-Derivative (PID) controller optimized by particle swarm optimization (PSO) and integrated with IoT technology. MATLAB/Simulink simulations showed the proposed controller achieved the smallest overshoot (7.64–8.15%), with average settling time reduced by 62.48 s and 20.38 s compared to conventional PID and fuzzy PID controllers, respectively (p < 0.001). Field experiments on winter wheat demonstrated a mean absolute EC deviation of 0.01125 ms/cm, with root-mean-square error (RMSE) of 0.0217 ms/cm, indicating high precision under field conditions. The system also maintained soil moisture in the optimal range (19–25%) with high irrigation uniformity (Christiansen’s coefficient Cu = 97.6%). The system maintained soil moisture in the optimal range (19–25%) while supporting stable soil nutrient levels and crop growth parameters. This study provides a validated technical solution for precision EC control while establishing a foundation for future fully integrated water–fertilizer management systems.