AgriEngineering, Volume 8, Issue 1 (January 2026) – 39 articles

Projects integrating Science, Technology, Engineering, and Mathematics (STEM) are essential to interdisciplinary research. This study presents a STEM (Science, Technology, Engineering, and Mathematics) methodology with the primary objective of designing, constructing, and validating a functional cannabinoid extraction device. To inform the device’s drying parameters, the dehydration kinetics of female hemp buds or flowering buds (FHB) were first analyzed using infrared drying at 100 °C for different durations. The plants were cultivated and harvested in accordance with good agricultural practices using Dinamed CBD Autoflowering seeds. The FHB were harvested and prepared by manually separating them from the stems and leaves. Six 5 g samples were prepared, each with a slab geometry of varying surface area and thickness. Two of these samples were ground: one into a fine powder and the other into a coarse powder. Mathematical fits were obtained for each resulting curve using either an exponential decay model or the logarithmic equation $y\left(t\right)=A\left(e^{-kt}\right)+y_0$ calculate the equilibrium moisture (m_E). The Moisture Rate (MR) was calculated, and by modelling with the logarithmic equation, the constant k and the effective diffusivity (Deff) were determined with the analytical solution of Fick’s second law. The Deff values (ranging from 10⁻⁷ to 10⁻⁵) were higher than previously reported. The coarsely ground powder sample yielded the highest k and Deff values and was selected for oil extraction. The device was then designed using Quality Function Deployment (QFD), specifically the House of Quality (HoQ) matrix, to systematically translate user requirements into technical specifications. A 200 g sample of coarsely ground, dehydrated FHB was prepared for ethanol extraction. Chemical results obtained by Liquid Chromatography coupled with Photodiode Array Detection (LC-PDA) revealed the presence of THC, CBN, CBC, and CBG. The extraction device design was validated using previous results showing the presence of CBD and CBDA. The constructed device successfully extracted cannabinoids, including Δ9-THC, CBG, CBC, and CBN, from coarsely ground FHB, validating the integrated STEM approach. This work demonstrates a practical framework for developing accessible agro-technical devices through interdisciplinary collaboration. Full article

Large language models (LLMs) offer new opportunities for agricultural education and decision support, yet their adoption is limited by domain-specific terminology, ambiguous retrieval, and factual inconsistencies. This work presents AgroLLM, a domain-governed agricultural knowledge system that integrates structured textbook-derived knowledge with Retrieval-Augmented Generation (RAG) and a Domain Knowledge Processing Layer (DKPL). The DKPL contributes symbolic domain concepts, causal rules, and agronomic thresholds that guide retrieval and validate model outputs. A curated corpus of nineteen agricultural textbooks was converted into semantically annotated chunks and embedded using Gemini, OpenAI, and Mistral models. Performance was evaluated using a 504-question benchmark aligned with four FAO/USDA domain categories. Three LLMs (Mistral-7B, Gemini 1.5 Flash, and ChatGPT-4o Mini) were assessed for retrieval quality, reasoning accuracy, and DKPL consistency. Results show that ChatGPT-4o Mini with DKPL-constrained RAG achieved the highest accuracy (95.2%), with substantial reductions in hallucinations and numerical violations. The study demonstrates that embedding structured domain knowledge into the RAG pipeline significantly improves factual consistency and produces reliable, context-aware agricultural recommendations. AgroLLM offers a reproducible foundation for developing trustworthy AI-assisted learning and advisory tools in agriculture. Full article

Facing water scarcity and environmental contamination, a sustainable approach combining bioeconomy and circular economy principles has emerged: the use of onshore oil and gas produced water (PW) to irrigate Nopalea cochenillifera. This study evaluated the ability of Nopalea cochenillifera to phytoextract contaminants, focusing on translocation and bioaccumulation factors for the recovery of degraded soils. The experiment was conducted in a randomized block design with five treatments (T1: 100% supply water; T2: 75% supply water + 25% PW; T3: 50% supply water + 50% PW; T4: 25% supply water + 75% treated PW; T5: 100% PW) and five replicates in 20 L pots. After 240 days, plant and soil samples were analyzed for micronutrients (Cu²⁺, Mn²⁺, Fe²⁺, Zn²⁺ and Na⁺) and heavy metals (Cr, Ni, Cd and Pb). The highest median TF was observed for Mn in treatment T3 (10.55), while the highest median BF occurred for Cu in treatment T2 (10.852). Nopalea cochenillifera effectively translocated Mn, Zn, Ni, Cd, and Pb from roots to shoots and bioaccumulated all analyzed nutrients, particularly Cu, Mn, Fe, and Zn. PW irrigation altered elemental transport and intensified metals accumulation. Thus, Nopalea cochenillifera demonstrates strong phytoextraction potential for environmental remediation in semi-arid regions. Full article

This study compared three deep learning architectures—one-dimensional convolutional neural network (1D-CNN), self-supervised learning (SSL), and Vision Transformer (ViT)—to evaluate their ability to predict carotenoid content from visible–near-infrared (VIS–NIR) spectral reflectance data (400–850 nm) acquired non-destructively from tea leaves. Model performance was evaluated using 10-fold cross-validation and analyzed through the mean SHapley Additive exPlanations values to identify key spectral features. The ViT model achieved the highest predictive accuracy (coefficient of determination [R²] = 0.81, root mean square error [RMSE] = 1.04, ratio of performance to deviation [RPD] = 2.32), followed by 1D-CNN (R² = 0.75, RMSE = 1.21, RPD = 1.99), whereas SSL showed substantially lower predictive performance (R² = 0.30, RMSE = 2.01, RPD = 1.20). Feature importance analysis revealed that ViT focused strongly on the red-edge region around 720 nm, which corresponds to spectral features associated with carotenoids and chlorophyll. The 1D-CNN relied mainly on blue (450–480 nm) and red (670–700 nm) regions, while SSL exhibited a broadly distributed importance pattern across wavelengths. These results indicate that ViT’s self-attention mechanism captures long-range spectral dependencies more effectively than conventional convolutional or self-supervised models. Overall, the study demonstrates that transformer-based architectures provide a powerful and interpretable framework for non-destructive estimation of carotenoid content from VIS–NIR reflectance spectroscopy. Full article

The temperate Himalayan states of Jammu and Kashmir, Himachal Pradesh, Uttarakhand, Ladakh, Sikkim, and Arunachal Pradesh in India face unique agro-ecological challenges across agriculture and allied sectors, including pest and disease pressures, inefficient resource use, post-harvest losses, and fragmented supply chains. This review systematically examines 21 critical problem areas, with three key challenges identified per sector across agriculture, agricultural engineering, fisheries, forestry, horticulture, sericulture, and animal husbandry. Artificial Intelligence (AI) and Machine Learning (ML) interventions, including computer vision, predictive modeling, Internet of Things (IoT)-based monitoring, robotics, and blockchain-enabled traceability, are evaluated for their regional applicability, pilot-level outcomes, and operational limitations under temperate Himalayan conditions. The analysis highlights that AI-enabled solutions demonstrate strong potential for early pest and disease detection, improved resource-use efficiency, ecosystem monitoring, and market integration. However, large-scale adoption remains constrained by limited digital infrastructure, data scarcity, high capital costs, low digital literacy, and fragmented institutional frameworks. The novelty of this review lies in its cross-sectoral synthesis of AI/ML applications tailored to the Himalayan context, combined with a sector-wise revenue-loss assessment to quantify economic impacts and guide prioritization. Based on the identified gaps, the review proposes feasible, context-aware strategies, including lightweight edge-AI models, localized data platforms, capacity-building initiatives, and policy-aligned implementation pathways. Collectively, these recommendations aim to enhance sustainability, resilience, and livelihood security across agriculture and allied sectors in the temperate Himalayan region. Full article

Rice–crab co-culture systems (RC) represent promising sustainable intensification approaches, yet their nitrogen (N) cycling and optimal fertilization strategies remain poorly characterized. In this study, we compared RC with rice monoculture system (RM) across four N gradients (0, 150, 210, and 270 kg N·hm⁻²), assessing N dynamics in field water and N distribution in soil. The results showed that field water ammonium nitrogen (NH₄⁺-N) concentrations increased nonlinearly, showing sharp increases beyond 210 kg N·hm⁻². Notably, crab activity in the RC altered the N transformation and transport processes, leading to a prolonged presence of nitrate nitrogen (NO₃⁻-N) in field water for two additional days after tillering fertilization compared to RM. This indicates a critical window for potential nitrogen loss risk, rather than enhanced retention, 15 days after basal fertilizer application. Compared to RM, RC exhibited enhanced nitrogen retention capacity, with NO₃⁻-N concentrations remaining elevated for an additional two days following tillering fertilization, suggesting a potential critical period for nitrogen loss risk. Post-harvest soil analysis revealed contrasting nitrogen distribution patterns: RC showed enhanced NH₄⁺-N accumulation in surface layers (0–2 cm) with minimal vertical NO₃⁻-N redistribution, while RM exhibited progressive NO₃⁻-N increases in subsurface layers (2–10 cm) with increasing fertilizer rates. The 210 kg N·hm⁻² rate proved optimal for the RC, producing a rice yield 12.08% higher than that of RM and sustaining high crab yields, while avoiding the excessive aqueous N levels seen at higher rates. It is important to note that these findings are based on a single-site, single-growing season field experiment conducted in Panjin, Liaoning Province, and thus the general applicability of the optimal nitrogen rate may require further validation across diverse environments. We conclude that a fertilization rate of 210 kg N·hm⁻² is the optimal strategy for RC, effectively balancing productivity and environmental sustainability. This finding provides a clear, quantitative guideline for precise N management in integrated aquaculture systems. Full article

Non-destructive estimation of high-temperature stress effects on tobacco plants is crucial for both scientific research and practical applications. Normalized difference vegetation index (NDVI), chlorophyll index, and spectra in the range of 900–1700 nm of Burley, Oriental, and Virginia tobacco plants under control and high-temperature stress conditions were measured using portable instruments. NDVI and chlorophyll index measurements indicate that young leaves of all tobacco types are tolerant to high temperatures. In contrast, the older leaves (the fifth leaf) showed increased sensitivity to heat stress. The chlorophyll content of these leaves decreased by 40 to 60% after five days of stress, and by the seventh day, the reduction reached 80% or more in all plants. The vegetative index of the fifth leaf also decreased on the seventh day of stress in all tobacco types. Differences in near-infrared spectra were observed between control, stressed, and recovered plants, as well as among different stress days, and among tobacco lines. The most significant differences were in the 1300–1500 nm range. The first characterization of heat-induced changes in the molecular structure of water in tobacco leaves using an aquaphotomics approach was conducted. Models for determining days of high-temperature treatment based on near-infrared spectra achieved a standard error of cross-validation (SECV) from 0.49 to 0.62 days. The total accuracy of the Soft Independent Modeling of Class Analogy (SIMCA) classification models of control, stressed, and recovered plants ranged from 91.0 to 93.6% using leaves’ spectra of the first five days of high-temperature stress, and from 90.7 to 97.7% using spectra of only the fifth leaf. Similar accuracy was obtained using Partial Least Squares–Discriminant Analysis (PLS-DA). Near-infrared spectroscopy and aquaphotomics can be used as a fast and non-destructive approach for early detection of stress and additional tools for investigating high-temperature tolerance in tobacco plants. Full article

Greenhouse pesticide application often suffers from low droplet deposition uniformity and health risks to operators. A self-propelled overhead rail air-assisted sprayer has been designed. The mathematical model based on droplet movement and the DPM are used to analyze the equipment’s working principle. Deposition surfaces at 0.4, 0.5, 0.6, and 0.7 m were used to examine the effects of travel speed, external airflow, and spray angle on droplet deposition uniformity. Through one-way analysis of variance, all variables reached a significant level (p < 0.001). Simulation results identified the optimal operating parameters: travel speed of 0.3 m/s, external air-flow velocity of 0.3 m/s, and spray angle of 5°, resulting in droplet deposition densities of 719, 586, 700, and 839 droplets/cm², with a coefficient of variation of 14.57%. The sediment variation coefficients of both the on-site test results and the simulation results were within 10%, which proved the reliability of the numerical simulation. In conclusion, the device proposed in this study effectively enables targeted fog spraying for multi-layer crops in greenhouses, significantly improving pesticide utilization, reducing application costs, and minimizing environmental pollution. It offers reliable technical support for greenhouse pest control operations. Full article

Sugarcane diseases cause estimated global annual losses of over $5 billion. While deep learning shows promise for disease detection, current approaches lack transparency and confidence estimates, limiting their adoption by agricultural stakeholders. We developed an uncertainty-aware detection system integrating Monte Carlo (MC) dropout with MobileNetV3, trained on 2521 images across five categories: Healthy, Mosaic, Red Rot, Rust, and Yellow. The proposed framework achieved 97.23% accuracy with a lightweight architecture comprising 5.4 M parameters. It enabled a 2.3 s inference while generating well-calibrated uncertainty estimates that were 4.0 times higher for misclassifications. High-confidence predictions (>70%) achieved 98.2% accuracy. Gradient-weighted Class Activation Mapping provided interpretable disease localization, and the system was deployed on Hugging Face Spaces for global accessibility. The model demonstrated high recall for the Healthy and Red Rot classes. The model achieved comparatively higher recall for the Healthy and Red Rot classes. The inclusion of uncertainty quantification provides additional information that may support more informed decision-making in precision agriculture applications involving farmers and agronomists. Full article

Weed management has become a critical agricultural practice, as weeds compete with crops for nutrients, host pests and diseases, and cause major economic losses. The invasive weed Ambrosia artemisiifolia (common ragweed) is particularly problematic in Hungary, endangering crop productivity and public health through its fast proliferation and allergenic pollen. This review examines the current challenges and impacts of A. artemisiifolia while exploring sustainable approaches to its management through precision agriculture. Recent advancements in unmanned aerial vehicles (UAVs) equipped with advanced imaging systems, remote sensing, and artificial intelligence, particularly deep learning models such as convolutional neural networks (CNNs) and Support Vector Machines (SVMs), enable accurate detection, mapping, and classification of weed infestations. These technologies facilitate site-specific weed management (SSWM) by optimizing herbicide application, reducing chemical inputs, and minimizing environmental impacts. The results of recent studies demonstrate the high potential of UAV-based monitoring for real-time, data-driven weed management. The review concludes that integrating UAV and AI technologies into weed management offers a sustainable, cost-effective, mitigate the socioeconomic impacts and environmentally responsible solution, emphasizing the need for collaboration between agricultural researchers and technology developers to enhance precision agriculture practices in Hungary. Full article

Early and accurate date palm disease detection is the key to successful smart farming ecosystem sustainability. In this paper, we introduce DoST-DPD, a new Dual-Stream Transformer architecture for multimodal disease diagnosis utilizing RGB, thermal and NIR imaging. In contrast with standard deep learning approaches, our model receives ontology-based semantic supervision (via per-dataset OWL ontologies), enabling knowledge injection via SPARQL-driven reasoning during training. This structured knowledge layer not only improves multimodal feature correspondence but also restricts label consistency for improving generalization performance, particularly in early disease diagnosis. We tested our proposed method on a comprehensive set of five benchmarks (PlantVillage, PlantDoc, Figshare, Mendeley, and Kaggle Date Palm) together with domain-specific ontologies. An ablation study validates the effectiveness of incorporating ontology supervision, consistently improving the performance across Accuracy, Precision, Recall, F1-Score and AUC. We achieve state-of-the-art performance over five widely recognized baselines (PlantXViT, Multi-ViT, ERCP-Net, andResNet), with our model DoST-DPD achieving the highest Accuracy of 99.3% and AUC of 98.2% on the PlantVillage dataset. In addition, ontology-driven attention maps and semantic consistency contributed to high interpretability and robustness in multiple crop and imaging modalities. Results: This work presents a scalable roadmap for ontology-integrated AI systems in agriculture and illustrates how structured semantic reasoning can directly benefit multimodal plant disease detection systems. The proposed model demonstrates competitive performance across multiple datasets and highlights the unique advantage of integrating ontology-guided supervision in multimodal crop disease detection. Full article

This study aimed to develop a methodology to evaluate, through RGB image processing, the wheat cultivar TRIO Calibre under three irrigation levels (100, 50, and 25%), with or without the application of Bacillus aryabhattai, in Brazilian Cerrado soil. The experimental scheme was a $3\times 2$ factorial design with five replicates. Images were collected, numbered, and organized into files, which were transformed to grayscale. During processing, the grayscale level co-occurrence matrix (GLCM) technique was applied and implemented in four main directions (0°, 45°, 90°, and 135°), and 13 statistical descriptors were extracted. At physiological maturity, the plants were harvested, and the following yield components were evaluated: plant height (PH), number of spikes per plant (NS), number of grains per spikes (NGS), average grain weight (AGW), and total prodution of grains (TPG). Irrigation influenced all the variables, with higher TPG and NS at 100% and 50% water and higher AGW at 25% water. The results indicated that the “contrast” descriptor in the 90° and 135° GLCM directions was the most efficient in differentiating treatments, which presented better performance in the 90° direction and was significantly correlated with the NS ($r=-0.48$, $p<0.05$) and TPG ($r=-0.46$, $p<0.05$). The analyses demonstrated that the methodology has the potential to be adapted for the analysis of under controlled conditions, contributing to more sustainable agricultural practices. Full article

LoRa-based IoT systems are increasingly used in smart farming, greenhouse monitoring, and large-scale agricultural sensing, where long-range, energy-efficient communication is essential. However, estimating link quality metrics such as PRR, RSSI, and SNR typically requires continuous packet transmission and sequence logging, an impractical approach for power-constrained field nodes. This study proposes a deep learning-driven framework for real-time prediction of link- and network-level performance in multihop LoRa networks, targeting the LoRaDisC protocol commonly deployed in agricultural environments. By integrating Bayesian surrogate modeling with Random Forest-guided hyperparameter optimization, the system accurately predicts PRR, RSSI, and SNR using multivariate time series features. Experiments on a large-scale outdoor LoRa testbed (ChirpBox) show that aggregated link layer metrics strongly correlate with PRR, with performance influenced by environmental variables such as humidity, temperature, and field topology. The optimized model achieves a mean absolute error (MAE) of 8.83 and adapts effectively to dynamic environmental conditions. This work enables energy-efficient, autonomous communication in agricultural IoT deployments, supporting reliable field sensing, crop monitoring, livestock tracking, and other smart farming applications that depend on resilient low-power wireless connectivity. Full article

To address challenges such as small target size, high density, severe occlusion, complex background interference, and edge device computational constraints, a lightweight model, GAFNet, is proposed based on YOLO11n, optimized for rice pest detection in field environments. To improve feature perception, we propose the Global Attention Fusion and Spatial Pyramid Pooling (GAM-SPP) module, which captures global context and aggregates multi-scale features. Building on this, we introduce the C3-Efficient Feature Selection Attention (C3-EFSA) module, which refines feature representation by combining depthwise separable convolutions (DWConv) with lightweight channel attention to enhance background discrimination. The model’s detection head, Enhanced Ghost Detect (EGDetect), integrates Enhanced Ghost Convolution (EGConv), Squeeze-and-Excitation (SE), and Sigmoid-Weighted Linear Unit (SiLU) activation, which reduces redundancy. Additionally, we propose the Focal-Enhanced Complete-IoU (FECIoU) loss function, incorporating stability and hard-sample weighting for improved localization. Compared to YOLO11n, GAFNet improves Precision, Recall, and mean Average Precision (mAP) by 3.5%, 4.2%, and 1.6%, respectively, while reducing parameters and computation by 5% and 21%. GAFNet can deploy on edge devices, providing farmers with instant pest alerts. Further, GAFNet is evaluated on the AgroPest-12 dataset, demonstrating enhanced generalization and robustness across diverse pest detection scenarios. Overall, GAFNet provides an efficient, reliable, and sustainable solution for early pest detection, precision pesticide application, and eco-friendly pest control, advancing the future of smart agriculture. Full article

Waste valorisation has become a key strategy for applying circular economy principles in the agro-industrial field. This study investigated the territorial implementation of the waste composting on a territorial scale. The wastes considered were the post-processing orange waste, spent olive oil pomace, and spent wine grape pomace. Their potential use as soil amendments across the provinces of Sicily was assessed through a GIS-based analysis, taking into account nitrogen (N) application constraints. Moreover, a cascade valorisation scheme was also evaluated: post-processing orange waste was first used as animal feed, and the remaining fraction was directed to composting; olive pomace was first sent to pomace oil extraction mills, and the residual material was subsequently used for composting. Results indicate that N inputs derived from composted residues remain below legal thresholds in all provinces, with relative contributions ranging from 38% to 92% of the regulatory limits. Spatial variability in nitrogen availability reflects the territorial distribution of agro-industrial activities, highlighting the importance of localised management strategies. These findings demonstrate that composting, combined with cascade valorisation, is an effective pathway to close nutrient cycles, reduce waste generation, and support sustainable biomass management in regional agri-food systems. Full article

Precision agriculture has increasingly adopted controlled agricultural microclimates, particularly smart greenhouses, as a strategy to enhance crop yields while maintaining environmental conditions within suitable ranges for each crop. Among the variables that govern the water balance in these systems, air humidity plays a critical role; therefore, accurate humidity forecasting is essential for implementing timely control actions that support productivity levels. However, greenhouse conditions are frequently perturbed by extreme weather events, which lead to nonlinear and non-stationary humidity dynamics. In this context, the aim of this study was to design an optimized evolving fuzzy inference system for humidity forecasting that can adapt to changing and unforeseen situations in agricultural microclimates. A prototyping-based methodology was followed, including phases of communication, quick planning, modeling and quick design, construction of the prototype, and deployment. A hybrid genetic algorithm was used to optimize the parameters of an evolving Mamdani-type fuzzy inference system, extended to handle missing values in online data streams. Thirty independent optimization runs were performed, and the best configuration achieved a mean squared error of 1.20 × 10⁻² in humidity forecasting using one minute of data for three months. The resulting model showed high interpretability, with an average number of 1.35 rules, tolerance for missing values, imputing 2% of the data, and robustness to sudden changes in the data stream with a p-value of 0.01 for the Augmented Dickey–Fuller test at alpha = 0.05. In general, the optimized evolving fuzzy inference system obtained an effectiveness rate greater than 90% and demonstrated adaptability to extreme weather conditions, suggesting its applicability to other phenomena with similar characteristics. Full article

This study presents a field-oriented AI web system for durian orchard management that recognizes leaf health from on-orchard images under variable illumination. Two complementary pipelines are employed: (1) a rule-based module operating in HSV and CIE Lab color spaces that suppresses sun-induced specular highlights via V/L* thresholds and applies interpretable hue–chromaticity rules with spatial constraints; and (2) a Deep Feature (PCA–SVM) pipeline that extracts features from pretrained ResNet50 and DenseNet201 models, performs dimensionality reduction using Principal Component Analysis, and classifies samples into three agronomic classes: healthy, leaf-spot, and leaf-blight. This hybrid architecture enhances transparency for growers while remaining robust to illumination variations and background clutter typical of on-farm imaging. Preliminary on-farm experiments under real-world field conditions achieved approximately 80% classification accuracy, whereas controlled evaluations using curated test sets showed substantially higher performance for the Deep Features and Ensemble model, with accuracy reaching 0.97–0.99. The web interface supports near-real-time image uploads, annotated visual overlays, and Thai-language outputs. Usability testing with thirty participants indicated very high satisfaction (mean 4.83, SD 0.34). The proposed system serves as both an instructional demonstrator for explainable AI-based image analysis and a practical decision-support tool for digital horticultural monitoring. Full article

Early and non-destructive detection of plant-parasitic nematodes is critical for implementing site-specific management in coffee production systems. This study evaluated the potential of unmanned aerial vehicle (UAV) multispectral and thermal imaging, combined with textural analysis, to detect Meloidogyne exigua infestation in Coffea arabica (Topázio variety). Field surveys were conducted in two contrasting seasons (dry and rainy), and nematode incidence was identified and quantified by counting root galls. Vegetation indices (NDVI, GNDVI, NGRDI, NDRE, OSAVI), individual spectral bands, canopy temperature, and Haralick texture features were extracted from UAV-derived imagery and correlated with gall counts. Under the conditions of this experiment, strong correlations were observed between gall number and the red spectral band in both seasons (R > 0.60), while GNDVI (dry season) and NGRDI (rainy season) showed strong negative correlations with gall density. Thermal imaging revealed moderate positive correlations with infestation levels during the dry season, indicating potential for early stress detection when foliar symptoms were absent. Texture metrics from the red and green bands further improved detection capacity, particularly with a 3 × 3 pixel window at 135°. These results demonstrate that UAV-based multispectral and thermal imaging, enhanced by texture analysis, can provide reliable early indicators of nematode infestation in coffee. Full article

For high-value-added products such as matcha, scientific confirmation of the origin is essential for quality assurance and fraud prevention. In this study, three nondestructive analytical techniques, specifically fluorescence (FF), near-infrared (NIR), and Fourier transform infrared (FT-IR) spectroscopy, were combined with machine learning algorithms to accurately identify the origin of Japanese matcha. FF data were analyzed using convolutional neural networks (CNNs), whereas NIR and FT-IR spectral data were analyzed using k-nearest neighbors (KNNs), random forest (RF), logistic regression (LR), and support vector machine (SVM) models. The FT-IR–RF model demonstrated the highest accuracy (99.0%), followed by the NIR–KNN (98.7%) and FF–CNN (95.7%) models. Functional group absorption in FT-IR, moisture and carbohydrates in NIR, and amino acid and polyphenol fluorescence in FF contributed to the identification. These findings indicate that the selection of an algorithm appropriate for the characteristics of the spectroscopic data is effective for improving accuracy. This method can quickly and nondestructively identify the origin of matcha and is expected to be applicable to other teas and agricultural products. This new approach contributes to the verification of the authenticity of food and improvement in its traceability. Full article

Plasma-assisted nitrogen fixation is emerging as a promising alternative to the dominant industrial method of the Haber–Bosch (H–B) process, which is energy-intensive and environmentally detrimental. Nonthermal plasma technology represents a cutting-edge innovation with the potential to revolutionize nitrogen fertilizer (N-fertilizer) production, offering a more sustainable approach by operating under mild conditions, making it suitable for decentralized N-fertilizer production. Toward the goal, in this study, we demonstrate our development and test of a novel nonthermal plasma system for continuous on-site production of N-fertilizer. This technology results in a product of aqueous N-fertilizer on-site, from only air, water, and electricity without carbon emissions, directly applicable to plants, bypassing costly and hazardous multiple steps in the production and transportation of the industrial N-fertilizers. Full article

The low level of mechanization in the production process of cumin seeds is one of the primary factors limiting their yield and economic efficiency. To enhance the mechanization of cumin seed production, this study focused on cumin seeds as the research subject. Physical parameters of cumin seeds were determined through physical experiments; based on these parameters, a discrete element model of cumin seeds was established, and the shear modulus was calibrated using angle of repose tests. The established model was used to simulate the seeding process of a seed drill, the model’s accuracy was verified by analyzing the seed trajectory, movement velocity, seeding quality, and the dynamic angle of repose of seeds inside the drill. Results indicated that the collision recovery coefficient, static friction coefficient, and rolling friction coefficient between cumin seeds and ABS plastic, stainless steel plates, and other cumin seeds were 0.3, 0.35, and 0.21; 0.49, 0.39, 0.24; and 0.24, 0.38, 0.18, respectively. Calibration via simulated cylinder accumulation tests yielded a deviation of 0.28% between the simulated accumulation angle and the physical accumulation angle at a shear modulus of 100 MPa; the simulated seed trajectory during dispensing closely matched physical dispensing tests. The average deviation in particle drop velocity within the bridge channel region was 4.23%, with a maximum deviation of 6.07%; the average deviation in dynamic packing angle from start to finish for the particle group was 2.84%, with a maximum deviation of 4.18%; and the average mass discharged from the 14 simulated seed nozzles was 0.0446 g, compared to 0.043 g in physical tests, with a deviation of 3.72%. These results demonstrate the high accuracy and reliability of the established cumin discrete element model and its parameters, providing technical support for the design and optimization of full-process mechanical cumin production systems. Full article

Semantic segmentation of UAV–acquired RGB orthomosaics is a key component for quantifying vegetation cover and monitoring phenology in precision agriculture. This study evaluates a representative set of CNN–based architectures (U–Net, U–Net Xception–Style, SegNet, DeepLabV3+) and Transformer–based models (Swin–UNet/Swin–Transformer, SegFormer, and Mask2Former) under a unified and reproducible protocol. We propose a transfer–and–consolidation workflow whose performance is assessed not only through region–overlap and pixel–wise discrepancy metrics, but also via boundary–sensitive criteria that are explicitly linked to orthomosaic–scale vegetation–cover estimation by pixel counting under GSD (Ground Sample Distance) control. The experimental design considers a transfer scenario between morphologically related crops: initial training on Opuntia spp. (prickly pear), direct (“zero–shot”) inference on Agave salmiana, fine–tuning using only 6.84% of the agave tessellated set as limited target–domain supervision, and a subsequent consolidation stage to obtain a multi–species model. The evaluation integrates IoU, Dice, RMSE, pixel accuracy, and computational cost (time per image), and additionally reports the BF score and HD95 to characterize contour fidelity, which is critical when area is derived from orthomosaic–scale masks. Results show that Transformer-based approaches tend to provide higher stability and improved boundary delineation on Opuntia spp., whereas transfer to Agave salmiana exhibits selective degradation that is mitigated through low–annotation–cost fine-tuning. On Opuntia spp., Mask2Former achieves the best test performance (IoU 0.897 +/− 0.094; RMSE 0.146 +/− 0.002) and, after consolidation, sustains the highest overlap on both crops (IoU 0.894 +/− 0.004 on Opuntia and IoU 0.760 +/− 0.046 on Agave), while preserving high contour fidelity (BF score 0.962 +/− 0.102/0.877 +/− 0.153; HD95 2.189 +/− 3.447 px/8.458 +/− 16.667 px for Opuntia/Agave), supporting its use for final vegetation–cover quantification. Overall, the study provides practical guidelines for architecture selection under hardware constraints, a reproducible transfer protocol, and an orthomosaic–oriented implementation that facilitates integration into agronomic and remote–sensing workflows. Full article

Kazakhstan’s growing potato industry underscores the need to develop and apply digital solutions that boost grading efficiency. A comparison between two traditional deep neural network architectures used to classify color images of potatoes from Kazakhstan is discussed in the paper. Ten representative varieties of Kazakhstani potatoes were selected as objects of study: Alians, Alians mini, Astana, Astana mini, Edem, Edem mini, Nerli, Nerli mini, Zhanaisan, and Zhanaisan mini. Two convolutional neural network (CNN) models, SqueezeNet and GoogLeNet, were refined via transfer learning employing three optimization approaches. Then, they were used to classify the potato images. A comparison of the two neural networks’ classification performance was conducted using common evaluation criteria—accuracy, precision, F1 score, and recall—alongside a confusion matrix to highlight misclassified samples. The comparative analysis demonstrated that both CNN architectures—SqueezeNet and GoogLeNet—achieve high classification accuracy for Kazakhstani potato varieties, with the best performance on Astana and Zhanaisan (>97%). The study confirms the applicability of lightweight CNNs for digital varietal identification and automated quality assessment of seed potatoes under controlled imaging conditions. The developed approach is the first comparative CNN-based varietal identification of Kazakhstani potato tubers using transfer learning and contributes to the digitalization of potato breeding, and provides a baseline for future real-time sorting systems using deep learning. Full article

Spatial monitoring of olive systems in arid regions is essential for understanding agricultural expansion, water pressure, and productive sustainability. This study aimed to map coverage and estimate olive plantation density (Olea europaea L.) in the Atacama Desert, Tacna (Peru) through the integration of UAV-satellite multispectral images and machine learning algorithms (CART, Random Forest, and Gradient Tree Boosting). Forty-eight optical, radar, and topographic covariates were analyzed. Fifteen were selected for coverage classification and 16 for plantation density, using Pearson’s correlation (|r| > 0.75). The classification maps reported an area of 23,059.87 ha (38.21%) of olive groves, followed by 5352.10 ha (8.87%) of oregano cultivation and 725.74 ha (1.20%) of orange cultivation, with respect to the total study area, with overall accuracy (OA) of 86.6% and a Kappa coefficient of 0.81. Meanwhile, the RF and GTB regression models showed R² ≈ 0.89 and RPD > 2.8, demonstrating excellent predictive performance for estimating tree density (between 1 and 8 trees per 100 m²). Furthermore, the highest concentration of olive trees was found in the central and southern zones of the study area, associated with favorable soil and microclimatic conditions. This work constitutes the first comprehensive approach for olive mapping in southern Peru using UAV–satellite fusion, demonstrating the capability of ensemble models to improve agricultural mapping accuracy and support water and productive management in arid ecosystems. Full article

Artificial intelligence (AI) shows great promise for transforming the agriculture sector and can enable the development of many modern farming practices over conventional methods. Nowadays, AI agents and agentic AI have attained popularity due to their autonomous structure and working mechanism. This research work proposes an agentic AI framework that integrates multiple agents developed for farming land to promote climate-smart agriculture and support United Nations (UN) sustainable development goals (SDGs). The developed structure has four agents: Agent A for monitoring soil properties, Agent B for weather sensing, Agent C for disease detection vision sensing in rice crops, and Agent D, a multi-agent supervisor agent chatbot connected with the other agents. The overall objective was to connect all agents on a single platform to obtain sensor data and perform a predictive analysis. This will help farmers and landowners obtain information about weather conditions, soil properties, and vision-based disease detection so that appropriate measures can be taken on agricultural land for rice crops. For soil properties (nitrogen, phosphorus, and potassium) from Agent A and climate data (temperature and humidity) from Agent B, we deployed the long short-term memory (LSTM), gated recurrent unit (GRU), and one-dimensional convolutional neural network (1D-CNN) predictive models, which achieved an accuracy of 93.4%, 94%, and 96% for Agent A; a 0.27 mean absolute error (MAE) for temperature; and a 2.9 MAE for humidity on the Agent B data. For Agent C, we used vision transformer (ViT), MobileViT, and RiceNet (with a diffusion model layer as a feature extractor) models to detect disease. The models achieved accuracies of 95%, 98.5%, and 85.4% during training respectively. Overall, the proposed framework demonstrates how agentic AI can be used to transform conventional farming practices into a digital process, thereby supporting smart agriculture. Full article

The compression deformation of seed cotton has been identified as a key factor affecting the working reliability of the baling device and the quality of bale molding. However, due to the complex working conditions of seed cotton in the continuous compression process in a confined space, it has proven to be difficult to study the compression molding mechanism of machine-harvested seed cotton in the baling process. The present study employs a universal testing machine to compress the seed cotton. In addition, pressure sensors are utilised to ascertain the internal axial load transfer law of the seed cotton. Furthermore, the internal density distribution equation of the seed cotton is established. Moreover, the Fiber model is employed to establish a spatial helix structure model of the cotton fibre. Finally, the compression simulation test is conducted to calibrate its material parameters. The results of the study indicate that seed cotton exhibits hysteresis in its internal stress–strain transfer. Through the polynomial fitting of the compression–displacement curve, it has been demonstrated that as the seed cotton approaches the compressed side, the rate of change in compression increases. The internal density distribution of the seed cotton must be calculated when it is compressed to a density of 220 kg·m⁻³. It is found that the density of the upper layer of the seed cotton is slightly greater than that of the lower layer of the seed cotton. The density distribution equation must then be obtained through regression fitting. The parameters of the compression model must be calibrated by means of uniaxial compression tests. Finally, the density distribution equation of the cotton fibre must be obtained through the compression test. The parameters of the simulation model, as determined by the uniaxial compression test calibration, are of significant importance. This is particularly evident in the context of the Poisson’s ratio of cotton fibre and the cotton fibre elastic modulus under pressure. The regression equation was obtained through analysis of variance, and the simulation of contact parameter optimisation. The optimal parameter combination was determined to be 0.466, and the pressure at this time. The relative error was found to be 2.96%, and the compression of specific performance was determined to be 10.14%. These findings serve to validate the simulation model. The findings of this study have the potential to provide a theoretical foundation and simulation assistance for the design and optimisation of cotton picker baling devices. Full article

To address the issue of low depth estimation accuracy in complex goji berry orchards, this paper proposes a method for identifying and locating goji berries that combines the YOLO-VitBiS object detection network with stereo vision technology. Based on the YOLO11n backbone network, the C3K2 module in the backbone is first improved using the AdditiveBlock module to enhance its detail-capturing capability in complex environments. The AdditiveBlock introduces lightweight long-range interactions via residual additive operations, thereby strengthening global context modeling without significantly increasing computation. Subsequently, a weighted bidirectional feature pyramid network is introduced into the Neck to enable more flexible and efficient feature fusion. Finally, a lightweight shared detail-enhanced detection head is proposed to further reduce the network’s computational complexity and parameter count. The enhanced model is integrated with binocular stereo vision technology, employing the CREStereo depth estimation algorithm for disparity calculation during binocular stereo matching to derive the three-dimensional spatial coordinates of the goji berry target. This approach enables efficient and precise positioning. Experimental results demonstrate that the YOLO-VitBiS model achieves a detection accuracy of $96.6\%$, with a model size of $4.3\mathrm{MB}$ and only $1.856\mathrm{M}$ parameters. Compared to the traditional SGBM method and other deep learning approaches such as UniMatch, the CREStereo algorithm generates superior depth maps under complex conditions. Within a distance range of 400 mm to 1000 mm, the average relative error between the estimated and actual depth measurements is $2.42\%$, meeting the detection and ranging accuracy requirements for field operations and providing reliable recognition and localization support for subsequent goji berry harvesting robots. Full article

This article examines the pressing issue of verifying contact models in the discrete element method (DEM) for modeling soil tillage processes. Due to the lack of a generally accepted methodology for selecting contact models for various soil types, a comprehensive study was conducted combining field experiments and numerical modeling. A verification method was developed and tested based on comparing experimental data on the draft resistance of a plow body with the results of calculations in the Rocky DEM 4.4 software package. The study yielded reliable experimental values for the draft resistance components and established the ranges of variation for their parameters. A comparative analysis of 10 promising combinations of contact models identified in previous studies was conducted. It was found that the improved Hertz-Mindlin model with the JKR adhesion model provides the best fit to the experimental results. Particular attention is paid to analyzing the influence of surface energy in the JKR model on changes in the rheological properties of the soil medium, which opens up the possibility of predicting soil behavior at different moisture levels. The results of the work are of practical value for the design and optimization of agricultural implements at the stage of their numerical modeling. The accuracy of predicting the draft resistance of the plow body during modeling for the studied soils at a moisture content of 18–25% ranged from 80 to 95%. Full article

This study presents a hybrid recurrent neural network (RNN) approach for greenhouse microclimate prediction, combining a mechanistic model with an Elman network. The research addresses the gap in systematic comparisons between hybrid RNN and feedforward neural network (FFNN) architectures for greenhouse climate forecasting. Different network structures with 1, 2, 3, 5, and 7 hidden layers were evaluated using mean absolute percentage error (MAPE), mean square error (MSE), and coefficient of determination (R²). Results demonstrate that hybrid RNNs significantly outperform FFNNs in predicting indoor temperature, with the 2-hidden-layer configuration achieving the best performance (R² = 0.897). For relative humidity prediction, both networks showed comparable results. The hybrid RNN with 3 hidden layers exhibited optimal performance during training, while simpler configurations proved more effective during testing. The integration of mechanistic knowledge with neural networks enhances prediction accuracy, providing a reliable tool for greenhouse climate control systems. These findings contribute to smart agriculture by offering an efficient computational approach for microclimate management. Full article

The use of agrochemicals remains indispensable for ensuring fruit production; however, their excessive or inefficient application poses significant environmental and health concerns. Rapid detection of spray deposition is crucial for assessing sprayer performance, improving precision application, and reducing drift and chemical waste. In this context, real-time monitoring technologies represent a promising tool to promote sustainable and efficient crop protection practices. This study refines previous experiences with an array of resistive sensors to quickly measure spray deposition. First, a multi-point calibration curve is introduced to improve the sensors’ accuracy. Furthermore, a multiplexed acquisition system (Sciospec ISX-5) is employed to enable time-resolved measurements of the whole sensor array. The method is validated by spectrophotometry and weight measurements. Wind tunnel trials with fluorescein (FLU) and fluorescein + potassium chloride (FLU + KCl) tracing solutions were conducted. The conductivity of the latter was higher than the former, without biasing the measurement. Both tracers showed good correlation between deposition and conductivity (R² = 0.997 for FLU and 0.995 for FLU + KCl), and the maximum deviation from the spectrophotometric estimates was <10%. Time-resolved measurement showed the build-up of deposition over time, potentially indicating the dimensional composition of the sprayed cloud. The improved workflow provides array-wide, sequential deposition measurements, enabling faster on-site acquisition and efficient analysis. The results demonstrate strong potential for scaling the method to field applications, supporting its further development into real-time deposition mapping tools that could guide precision spraying, optimize agrochemical use, and reduce environmental drift. Full article