In this study, a capsule-oriented prototype machine for mechanical harvesting of poppy (Papaver somniferum L.) was developed, and its design features and performance were evaluated. In Turkey, poppy cultivation is mostly conducted in small plots, and harvesting operations are generally performed manually. This situation both increases the need for labor and leads to crop losses. The prototype harvesting system was evaluated for efficiency and capsule loss under varying plant morphologies. During the design process, the physico-mechanical properties of the poppy plant were determined, and cutting blade placements and motor selections were made in accordance with these properties. The harvesting success of the machine was analyzed under different plant heights and capsule numbers through experiments conducted under laboratory and field conditions. Across 360 plant samples tested under laboratory and field conditions, the system achieved a harvesting efficiency of 92% with less than 8% capsule loss. It was determined that the circular cutting system operated with less force and more efficiently compared to vertical cutting. In addition, the AC and DC motor systems used in the machine optimized the plant transport and capsule cutting processes by providing different speed and direction controls. The results show that the prototype machine is successful in terms of both efficiency and quality and can be applied in wider areas by integrating automatic feed systems in the future. These findings indicate that the developed cutting and brush-based transport system can reduce capsule loss more effectively than conventional single-point cutting mechanisms reported in previous studies.

The increasing frequency and intensity of heavy snowfall events due to climate change pose significant risks to agricultural facilities, particularly lightweight structures such as greenhouses. This study develops a probabilistic risk management framework using a Markov Chain approach to analyse snow load dynamics under varying climatic conditions, incorporating the effects of snow density changes due to temperature fluctuations. Time-series meteorological data from 99 weather stations across South Korea over the past 20 years were utilised to calculate hourly snow loads. Cluster analysis was employed to classify snow load states, and transition probabilities between states were derived to construct Markov transition matrices. The framework evaluates failure probabilities of various greenhouse specifications, highlighting the influence of structural thresholds and regional snowfall patterns. Areas prone to heavy snowfall exhibited significantly higher failure probabilities compared to regions with milder conditions. National-scale failure probability maps were developed to provide actionable insights for disaster mitigation, emphasising the importance of region-specific risk management strategies. Results demonstrate the critical role of snow density in snow load evolution and its implications for structural safety. The proposed methodology facilitates early warning systems, resource allocation, and policy recommendations, supporting a proactive approach to disaster risk reduction. This study underscores the necessity of integrating probabilistic models with structural safety assessments to enhance the resilience of agricultural facilities against extreme snowfall events, offering a robust tool for sustainable agricultural practices.

In this paper, an improved lightweight YOLOv8 method is proposed to detect the ripeness of tomato fruits, given the problems of subtle differences between neighboring stages of ripening and mutual occlusion of branches, leaves, and fruits. The method replaces the backbone network of the original YOLOv8 with a more lightweight MobileNetV3 structure to reduce the number of parameters of the model; at the same time, it integrates the convolutional attention mechanism module (CBAM) in the feature extraction network, which enhances the network's capability of extracting features of tomato fruits. At the same time, it introduces the SCYLLA-IoU (SIoU) as a bounded YOLOv8 frame regression loss function, effectively solving the mismatch problem between the predicted frame and the actual frame and improving recognition accuracy. Compared with the current mainstream models Resnet50, VGG16, YOLOv3, YOLOv5, YOLOv7, etc., the model is in an advantageous position regarding precision rate, recall rate, and detection accuracy. The research and experimental results show that the mean values of precision, recall rate, and average precision of the improved MCS-YOLOv8 model under the test set are 91.2%, 90.2%, and 90.3%, respectively. The detection speed of a single image is 5.4ms, and the model occupies less memory by 8.7 M. The model has a clear advantage in both detection speed and precision rate and also shows that the improved MCS-YOLOv8 model can provide strong technical support for tomato-picking robots in complex environments in the field.

Agriculture constitutes a critical sector within the economic landscape of Pakistan, engaging 37.4% of the labour force and contributing 22.9% to the Gross Domestic Product (GDP), thus functioning as an essential element of the country's economic architecture and the sustenance of its citizens. The primary agricultural commodities, which encompass wheat, rice, cotton, maize, and sugarcane, occupy extensive tracts of agricultural land, thereby underscoring their paramount importance regarding food security, impacts on territory management and related landscape, as well as income generation through export activities. Despite its crucial role, the advancement of agricultural mechanization in Pakistan is significantly lacking, with the available agricultural power that is a little less than 1.6 kW ha-1, which is below the minimum required farm power of 1.82 kW ha-1. In this study, we seek to investigate the historical evolution, current scenario, and significant obstacles confronting agricultural mechanization in Pakistan, especially right now when innovative trends worldwide are pushing towards a progressive digitisation of the sector. Continuing with the conventional agricultural practices during the period of independence in 1947, the Green Revolution of the 1960s represented a crucial transformation towards mechanization, propelled by the indigenous manufacturing of tractors. Nevertheless, the pace of mechanization has slowed in recent years, primarily due to the prevalence of small farm sizes, economic constraints, and inadequate access to financial resources. While tasks such as land preparation and pesticide application have achieved notable levels of mechanization, fundamental operations including sowing, transplanting, weeding, and harvesting continue to be insufficiently mechanized. Significant initiatives, such as Laser Land Levelling in Punjab, show the considerable impact of focused interventions. To address the mechanization deficit, proposed strategies encompass the enhancement of local machinery manufacturing, the establishment of quality standards, the promotion of advanced imported equipment, an increase in farmer education, and the implementation of comprehensive government support through subsidies, tax benefits, and dedicated research and development efforts. A collaborative approach between governmental bodies and the private sector is imperative for fostering advancements in mechanization, thereby ensuring a more efficient, productive, and sustainable agricultural sector in Pakistan.

Given the significant damage rate observed during the transportation of current garlic combine harvesters in China, this study aims to design a new garlic combine harvester capable of achieving minimal harvest losses. The designed machine can simultaneously complete operations for garlic digging, clamping transport, seedling-bulb separation, soil cleaning, and fruit collection across two rows. Through the use of theoretical analysis and calculation of garlic harvesting operations, the key parameters of soil-breaking device, clamping transport device, length-limiting cutting device, and soil cleaning conyeyor were determined. The Box-Behnken test technique was utilized within Design-Expert software, and orthogonal experiments were conducted with the unit's forward speed, digging depth, and soil-breaking angle as test factors, and the stem cutting rate and bulb damage rate as test indices. The test results showed that when the unit's forward speed, digging depth, and soil-breaking angle were 0.49 m/s, 100 mm, and 20°, respectively, the working parameter combination was the best, and the rate of stem cutting and damage were 95.71% and 3.10%, respectively. The findings from the field experiment and optimization aligned closely. This study can provide reference for the development of mechanized garlic harvesting.

A method combining experimental and simulation optimization was used to calibrate parameters to enhance the accuracy of discrete element model parameters during kiwifruit stem separation. First, physical experiments were conducted to determine the intrinsic and contact parameters of kiwifruit stalk. Second, the mechanical parameters of the kiwifruit stalks were determined using three-point bending and shear tests. On this basis, simulation tests were conducted on kiwifruit stalks by combining the Hertz-Mindlin model with a bonding model, and the optimal combination of bonding parameters was confirmed using the bending strength and maximum shear force. Finally, a discrete element model of the kiwifruit was built with the determined bonding parameters and simulated, and the reliability of the model was verified through mechanical tests. The results showed that the density was 867.5 kg/m3, Poisson's ratio was 0.26, the modulus of elasticity was 3.25 × 108 Pa, the recovery coefficient between the fruit stalks and steel parts was 0.365, and the average values of the static and dynamic friction coefficients between the kiwifruit stalks and steel parts were 0.268 and 0.152, respectively. The kiwifruit stem bonding parameters were normal stiffness per unit area kn=7.201×1011 N/m3, shear stiffness per unit area kt=2.379×1011 N/m3, critical normal stress σmax=5.937×108 Pa, critical shear stress tmax = 2.354×109 Pa, and bonded disc radius Rj=0.164 mm. Compared with the results of the mechanical tests, the relative errors of the bending strength and maximum shear of the discrete element model were 2.13% and 2.84%, respectively. The results showed that the discrete element model improves the simulation of the bending and shearing processes of kiwifruit stalks and is capable of characterizing the physical properties of kiwifruit stalks. The results of this study provide a theoretical foundation for the optimal design of end effectors.

The increase in uncertain weather affects agriculture, impacting crop yield and quality, mainly due to the interaction of climatic variables such as temperature, wind speed, and humidity. In addition, soil erosion and nutrient loss are regional problems aggravated by inadequate agricultural practices in developing sugarcane agriculture. The present research proposes an Intelligent System based on a detection algorithm and a fuzzy model to estimate the quality of the sugarcane crop and the probability of the presence of pests and diseases through the prediction of uncertain variables. Wind speed, cloudiness, humidity, and thermal amplitude were considered variables of interest because parameters out of control of these variables generate a state of thermal stress, triggering pests and diseases that reduce crop quality and sugar production. This research uses geospatial information to simplify the exchange of information through a detection algorithm using real-time satellite images and a fuzzy model to estimate crop quality and prevent climate change-related problems. The variables humidity and cloudiness determine sugarcane quality as they are related to crop phenology and the probability that the crop will develop pests and diseases. In contrast, the intelligent system showed a correlation of over 93% for predicting the variables of interest.

Since Complete Fluidic Sprinklers (CFS) cannot function well in low-pressure environments, Dynamic Fluidic Sprinkler (DFS) were developed to address this issue. In 2021, research in the field and laboratory were conducted to examine how well DFS and CFS performed hydraulically in both indoor and outdoor conditions. In this investigation, a Thiess Clima laser precipitation monitor was used to evaluate the droplet size and velocity distribution of two different types of sprinklers indoors. From the findings, DFS velocities ranged from 0.1 to 4 m/s whereas CFS ranged from 0.1 to 5.3 m/s. The maximum frequency value was obtained at velocities of 1 m/s for each combination. The DFS had a slightly greater discharge coefficient and spray pattern than the CFS. The DFS's maximum spray range was 12.2 m, while the CFS's maximum spray range was 10.8 m, with standard deviations of 1.07 and 1.66, respectively. Under high wind speed conditions, the maximum combined Coefficient of Uniformity (CU) of DFS and CFS were 81.1% and 78%, respectively. For a given pressure and sprinkler spacing, DFS delivered higher CU values than CFS, especially while running at low pressure, demonstrating that DFS offered a more favoured water distribution pattern at low pressure. At different distances from the sprinkler, the highest application rates for DFS and CFS were 6.7 mm h−1 at 7 m and 6.5 mm h−1 at 7 m, respectively. A comparison of DFS and CFS under hydraulic performance indicated that DFS had a better performance than CFS. The study can serve as a guide for how to conserve water in sprinkler-irrigated fields.

Low degree of transplanting automation will affect production efficiency and planting quality in vegetable cultivation. A new seedling pick-up device was designed and constructed to reduce direct grasping damages to seedlings and improve transplanting efficiency. The pick-up device consists of an air jet loosening device, a flexible pick-up manipulator, a parallel feeding device, and a multi-axis motion control system. Its working principle is to use air jet ejection to assist in loosening of seedling roots from the tray cells, grasp their stems for extracting with the pick-up manipulator, and finally transfer them to the delivery device for feeding into the planting device as needed. The mechanical structure and working parameters of each component were designed, and the control system was constructed according to the working requirements of ejecting, extracting, transferring, and discharging operations. A prototype of the new pick-up device was constructed, and its performance evaluation was conducted using an orthogonal experimental design using cucumber, pepper and caluiflower as test objects. The results showed that the test object, the root lump's moisture content and the loosening way (either as a whole or individual loosening of seedlings) had significant effects on the success ratio in picking up seedlings. Overall, the success in picking up seedlings from the cell was found to be influenced by horticultural and mechanical factors. Under the optimal level group, the maximum success ratio for automatic picking up seedlings was up to 94.49% for pepper while that of cucumber and cauliflower recorded 90.75% and 92.62%, respectively. The seedling pick-up performance was satisfactory for efficient transplanting requirements.

Infructescence detection during the early fruiting stage is a necessary preliminary work to estimate the yield of Zanthoxylum. The purpose of this research is to detect and quantify the infructescences on the images of early fruit-bearing branches of Zanthoxylum which are collected in the natural environment. Thus, a machine vision-based algorithm for detecting Zanthoxylum infructescences is proposed, which contains of two phases. The first is to segment fruit-like and non-fruit region from Zanthoxylum branch image by extracting the Histogram of Oriented Gradient (HOG) feature map and Excess Green minus Excess Red (ExGR) index to obtain the fruits of Zanthoxylum infructescences. The second is to identify fruits adaptively and roughly in accordance with the density of their position distribution, then an optimization model is established to merge rough clusters and the optimal clustering result is obtained. Experiments with normal samples demonstrate that the proposed approach receives a Precision of 96.67%, a Recall of 91.07%, and an F1-score of 0.93. Compared with ADPC-kNN, DBSCAN and OPTICS, the results reveal that the proposed algorithm has an improved robustness and achieves higher Recall and F1-score. Meanwhile, its competitiveness is displayed in the experiments with deep learning-based methods. It can adaptively detect the infructescences of Zanthoxylum branch images, and the experiments prove that it is effective.