The purpose of the conducted research was to systematize and synthesize the direct and indirect losses incurred by Ukrainian agriculture in 2022-2025 and to assess changes in the sector’s condition – reflected in production dynamics, the structure of producers, and export performance – using available official statistics and institutional reports. The research period focused on the time of the full-scale war in Ukraine (2022-2025). Some comparisons were made also with the pre-war state of agriculture. As a result, the losses of agriculture due to the war were systematized and classified, and the main changes in the branch as a reaction to the challenges, associated with the war, were identified, including specified changes in the structure of Ukrainian agriculture, the dynamics of agricultural production volumes and its sectoral structure and, accordingly, in exports. It was established that, despite the direct and indirect damage, caused by the war to Ukrainian agriculture, which caused significant losses in the branch, it demonstrated an extraordinary ability, in comparison with other sectors, to survive in war conditions and realize its potential to meet domestic and external needs for food products and to ensure food security.
Food authentication and quality control have become major challenges in the global food industry, particularly
due to issues of food adulteration that impact consumer health and product integrity in the market. Portable
Near-Infrared (NIR) spectroscopy has emerged as a rapid, non-destructive, and environmentally friendly
technology with great potential for detecting adulteration and ensuring product quality, directly at various points
along the supply chain. This review article discusses the application of portable NIR spectroscopy in the
authentication of four major categories of food products: liquids, powders, oils, and meats. The review highlights
the types of samples and adulterants used, as well as the classification and prediction performance reported
across studies. The results demonstrate that portable NIR can achieve high classification accuracy (>90%) and
predictive performance with R2 values exceeding 0.90 and Ratio Prediction Deviation (RPD) values above 3,
depending on the product type and analytical approach. Although the performance of portable devices is
generally slightly lower than benchtop instruments, their advantages in mobility and ease of use make them
highly promising for field applications. Future studies are recomended to develop more universal model, broaden
the range of commodities investigated, and integrate this technology with other digital approaches to strengthen
food quality control systems that are faster, more precise, and sustainable.
Keywords: food authentication, food adulteration, portable NIR spectroscopy, food quality control
Usman Idris Ismail, Hui Na Chua, Rosdiadee Nordin
et al.
As a staple food for the majority of the global population, rice plays a vital role in food security. However, rice crop yield is heavily influenced by factors such as soil quality, weather, irrigation, and biological threats like pathogens (fungi, bacteria, viruses). Traditional methods for detecting rice diseases are often labor-intensive, time-consuming, and require expert knowledge, making them inefficient for large-scale or timely response. This process has prompted the adoption of automated techniques that integrate deep learning (DL) vision techniques to improve detection accuracy and efficiency. Deep learning models, particularly artificial neural networks, have shown promising results in detecting diseases from rice leaf images. This review aims to address three core research questions: What are the available open-source datasets for rice disease detection, and how do their characteristics affect model performance? What are the most commonly used deep learning architectures, and what are their advantages and limitations? What challenges exist in dataset generalization and model deployment for real-world applications? In answering these questions, this paper reviews current open-source datasets, highlighting their metadata and geographic coverage. It also compares popular deep learning architectures, discussing their respective strengths and shortcomings in rice disease detection. Furthermore, it explores the limitations of existing models in terms of real-world deployment, including issues related to data diversity, domain adaptation, and hardware constraints. Finally, the paper outlines future directions to improve the robustness and applicability of deep learning models in practical agricultural settings.
In precision agriculture and plant science, there is an increasing demand for wireless sensors that are easy to deploy, maintain, and monitor. This paper investigates a novel approach that leverages recent advances in extremely low-power wireless communication and sensing, as well as the rapidly increasing availability of unmanned aerial vehicle (UAV) platforms. By mounting a specialized wireless payload on a UAV, battery-less sensor tags can harvest wireless beacon signals emitted from the drone, dramatically reducing the cost per sensor. These tags can measure environmental information such as temperature and humidity, then encrypt and transmit the data in the range of several meters. An experimental implementation was constructed at AERPAW, an NSF-funded wireless aerial drone research platform. While ground-based tests confirmed reliable sensor operation and data collection, airborne trials encountered wireless interference that impeded successfully detecting tag data. Despite these challenges, our results suggest further refinements could improve reliability and advance precision agriculture and agrarian research.
Radiometric infrared (IR) imaging is a valuable technique for remote-sensing applications in precision agriculture, such as irrigation monitoring, crop health assessment, and yield estimation. Low-cost uncooled non-radiometric IR cameras offer new implementations in agricultural monitoring. However, these cameras have inherent drawbacks that limit their usability, such as low spatial resolution, spatially variant nonuniformity, and lack of radiometric calibration. In this article, we present an end-to-end pipeline for temperature estimation and super resolution of frames captured by a low-cost uncooled IR camera. The pipeline consists of two main components: a deep-learning-based temperature-estimation module, and a deep-learning-based super-resolution module. The temperature-estimation module learns to map the raw gray level IR images to the corresponding temperature maps while also correcting for nonuniformity. The super-resolution module uses a deep-learning network to enhance the spatial resolution of the IR images by scale factors of x2 and x4. We evaluated the performance of the pipeline on both simulated and real-world agricultural datasets composing of roughly 20,000 frames of various crops. For the simulated data, the results were on par with the real-world data with sub-degree accuracy. For the real data, the proposed pipeline was compared to a high-end radiometric thermal camera, and achieved sub-degree accuracy. The results of the real data are on par with the simulated data. The proposed pipeline can enable various applications in precision agriculture that require high quality thermal information from low-cost IR cameras.
Cropland non-agriculturalization refers to the conversion of arable land into non-agricultural uses such as forests, residential areas, and construction sites. This phenomenon not only directly leads to the loss of cropland resources but also poses systemic threats to food security and agricultural sustainability. Accurate identification of cropland and non-cropland areas is crucial for detecting and addressing this issue. Traditional CNNs employ static convolution layers, while dynamic convolution studies demonstrate that adaptively weighting multiple convolutional kernels through attention mechanisms can enhance accuracy. However, existing dynamic convolution methods relying on Global Average Pooling (GAP) for attention weight allocation suffer from information loss, limiting segmentation precision. This paper proposes Frequency-Aware Dynamic Convolution (FADConv) and a Frequency Attention (FAT) module to address these limitations. Building upon the foundational structure of dynamic convolution, we designed FADConv by integrating 2D Discrete Cosine Transform (2D DCT) to capture frequency domain features and fuse them. FAT module generates high-quality attention weights that replace the traditional GAP method,making the combination between dynamic convolution kernels more reasonable.Experiments on the GID and Hi-CNA datasets demonstrate that FADConv significantly improves segmentation accuracy with minimal computational overhead. For instance, ResNet18 with FADConv achieves 1.9% and 2.7% increases in F1-score and IoU for cropland segmentation on GID, with only 58.87M additional MAdds. Compared to other dynamic convolution approaches, FADConv exhibits superior performance in cropland segmentation tasks.
Da Tan, Michael Beck, Christopher P. Bidinosti
et al.
The success of agricultural artificial intelligence depends heavily on large, diverse, and high-quality plant image datasets, yet collecting such data in real field conditions is costly, labor intensive, and seasonally constrained. This paper investigates diffusion-based generative modeling to address these challenges through plant image synthesis, indoor-to-outdoor translation, and expert preference aligned fine tuning. First, a Stable Diffusion model is fine tuned on captioned indoor and outdoor plant imagery to generate realistic, text conditioned images of canola and soybean. Evaluation using Inception Score, Frechet Inception Distance, and downstream phenotype classification shows that synthetic images effectively augment training data and improve accuracy. Second, we bridge the gap between high resolution indoor datasets and limited outdoor imagery using DreamBooth-based text inversion and image guided diffusion, generating translated images that enhance weed detection and classification with YOLOv8. Finally, a preference guided fine tuning framework trains a reward model on expert scores and applies reward weighted updates to produce more stable and expert aligned outputs. Together, these components demonstrate a practical pathway toward data efficient generative pipelines for agricultural AI.
The variable and unpredictable load demands in hybrid agricultural tractors make it difficult to design optimal rule-based energy management strategies, motivating the use of adaptive, learning-based control. However, existing approaches often rely on basic fuel-based rewards and do not leverage expert demonstrations to accelerate training. In this paper, first, the performance of Q-value-based reinforcement learning algorithms is evaluated for powertrain control in a hybrid agricultural tractor. Three algorithms, Double Q-Learning (DQL), Deep Q-Networks (DQN), and Double DQN (DDQN), are compared in terms of convergence speed and policy optimality. Second, a piecewise domain-specific reward-shaping strategy is introduced to improve learning efficiency and steer agent behavior toward engine fuel-efficient operating regions. Third, the design of the experience replay buffer is examined, with a focus on the effects of seeding the buffer with expert demonstrations and analyzing how different types of expert policies influence convergence dynamics and final performance. Experimental results demonstrate that (1) DDQN achieves 70\% faster convergence than DQN in this application domain, (2) the proposed reward shaping method effectively biases the learned policy toward fuel-efficient outcomes, and (3) initializing the replay buffer with structured expert data leads to a 33\% improvement in convergence speed.
Shumila Shahid, Abubakar Dar, Azhar Hussain
et al.
The pollution of heavy metals (HMs) is a major environmental concern for agricultural farming communities due to water scarcity, which forces farmers to use wastewater for irrigation purposes in Pakistan. Vegetables grown around the cities are irrigated with domestic and industrial wastewater from areas near mining, paint, and ceramic industries that pollute edible parts of crops with various HMs. Cadmium (Cd) is an extremely toxic metal in arable soil that enters the food chain and damages the native biota, ultimately causing a reduction in plant growth and development. However, the use of microbes and growth regulators enhances plant growth and development as well as HM immobilization into the cell wall and hinders their entry into the food chain. Thus, the integrated use of bacterial consortium along with exogenously applied jasmonic acid (JA) mitigates the adverse effect of metal stress, ultimately reducing the metal mobility into roots by soil. Therefore, the current study was conducted to check the impact of Cd-tolerant bacteria and JA on the growth, nutrient status, and uptake of Cd in the cauliflower (Brassica oleracea). Our results demonstrated that increasing concentrations of Cd negatively affect growth, physiological, and biochemical attributes, while the use of a bacterial consortium (SS7 + SS8) with JA (40 μmol L−1) significantly improved chlorophyll contents, stem fresh and dry biomass (19.7, 12.7, and 17.3%), root length and root fresh and dry weights (28.8, 15.2, and 23.0%), and curd fresh and dry weights and curd diameter (18.7, 12.6, and 15.1%). However, the maximum reduction in soil Cd, roots, and curd uptake was observed by 8, 11, and 9.3%, respectively, under integrated treatment as compared to the control. Moreover, integrating bacterial consortium and JA improves superoxide dismutase (SOD) (16.79%), peroxidase dismutase (POD) (26.96%), peroxidase (POX) (26.13%), and catalase (CAT) (26.86%). The plant nitrogen, phosphorus, and potassium contents were significantly increased in soil, roots, and curd up to 8, 11, and 9.3%, respectively. Hence, a consortium of Klebsiella strains in combination with JA is a potential phytostabilizer and it reduces the uptake of Cd from soil to roots to alleviate the adverse impact on cauliflower’s growth and productivity.
Alexandre de Oliveira Bezerra, Vanessa Ap. de Moraes Weber, Fabricio de Lima Weber
et al.
Assessing the phenotype of cattle through human visual inspection is a very common and important practice in precision cattle breeding. This paper presents the results of a correlation analysis between scores produced by humans for Nelore cattle and a variety of measurements that can be derived from images or other instruments. It also presents a study using the k-means algorithm to generate new ways of clustering a batch of cattle using the measurements that most correlate with the animal's body weight and visual scores.
Anna-Akrivi Thomatou, Eleni C. Mazarakioti, Anastasios Zotos
et al.
Adulteration of high-value agricultural products is a critical issue worldwide for consumers and industries. Discrimination of the geographical origin can verify food authenticity by reducing risk and detecting adulteration. Between agricultural products, beans are a very important crop cultivated worldwide that provides food rich in iron and vitamins, especially for people in third-world countries. The aim of this study is the construction of a map of the locally characteristic isotopic fingerprint of giant beans, “Fasolia Gigantes-Elefantes PGI”, a Protected Geographical Indication product cultivated in the region of Kastoria and Prespes, Western Macedonia, Greece, with the ultimate goal of the discrimination of beans from the two areas. In total, 160 samples were collected from different fields in the Prespes region and 120 samples from Kastoria during each cultivation period (2020–2021 and 2021–2022). The light element (C, N, and S) isotope ratios were measured using Isotope Ratio Mass Spectrometry (IRMS), and the results obtained were analyzed using chemometric techniques, including a one-way ANOVA and Binomial logistic regression. The mean values from the one-way ANOVA were <i>δ</i><sup>15</sup>N<sub>AIR</sub> = 1.875‰, <i>δ</i><sup>13</sup>C<sub>V-PDB</sub> = −25.483‰, and <i>δ</i><sup>34</sup>S<sub>V-CDT</sub> = 4.779‰ for Kastoria and <i>δ</i><sup>15</sup>N<sub>AIR</sub> = 1.654‰, <i>δ</i><sup>13</sup>C<sub>V-PDB</sub> = −25.928‰, and <i>δ</i><sup>34</sup>S<sub>V-CDT</sub> = −0.174‰ for Prespes, and showed that stable isotope ratios of C and S were statistically different for the areas studied while the Binomial logistic regression analysis that followed correctly classified more than 78% of the samples.
Sofia Winnie Kosasih, Terip Karo-Karo, Nauas Domu Marihot Romauli
The aim of this study was to determine the quality of guava jelly after the addition of papaya peel extract and citric acid. The study used a two-factor, randomized factorial design. The first factor was the addition of papaya peel extract (P): (2%;4%;6%;8%), while the second was the addition of citric acid (A): (1%;1,5%;2%;2,5%). The parameters analyzed were moisture content, ash content, total soluble solids, vitamin C content, crude fiber content, degree of acidity (pH), and total acid. According to the results, the addition of papaya peel extract had a highly significant effect on the water content, crude fiber content, degree of acidity, total acid, and had a differ significant effect on ash content and total dissolved solid. The addition of citric acid also had a highly significant effect on the total dissolved solid, the content of Vitamin C, degree of acidity, the organoleptic test of taste, and had a differ significant effect on water content and total acid. The interaction between the addition of papaya peel extract and citric acid had a highly significant effect on the degree of acidity. Guava jelly with 8% papaya peel extract and 2% citric acid had the optimum quality characteristics.
Boje Deforce, Bart Baesens, Estefanía Serral Asensio
The recent surge in foundation models for natural language processing and computer vision has fueled innovation across various domains. Inspired by this progress, we explore the potential of foundation models for time-series forecasting in smart agriculture, a field often plagued by limited data availability. Specifically, this work presents a novel application of $\texttt{TimeGPT}$, a state-of-the-art (SOTA) time-series foundation model, to predict soil water potential ($ψ_\mathrm{soil}$), a key indicator of field water status that is typically used for irrigation advice. Traditionally, this task relies on a wide array of input variables. We explore $ψ_\mathrm{soil}$'s ability to forecast $ψ_\mathrm{soil}$ in: ($i$) a zero-shot setting, ($ii$) a fine-tuned setting relying solely on historic $ψ_\mathrm{soil}$ measurements, and ($iii$) a fine-tuned setting where we also add exogenous variables to the model. We compare $\texttt{TimeGPT}$'s performance to established SOTA baseline models for forecasting $ψ_\mathrm{soil}$. Our results demonstrate that $\texttt{TimeGPT}$ achieves competitive forecasting accuracy using only historical $ψ_\mathrm{soil}$ data, highlighting its remarkable potential for agricultural applications. This research paves the way for foundation time-series models for sustainable development in agriculture by enabling forecasting tasks that were traditionally reliant on extensive data collection and domain expertise.
Md. Toukir Ahmed, Md Wadud Ahmed, Ocean Monjur
et al.
As the demand for food surges and the agricultural sector undergoes a transformative shift towards sustainability and efficiency, the need for precise and proactive measures to ensure the health and welfare of livestock becomes paramount. In the context of the broader agricultural landscape outlined, the application of Hyperspectral Imaging (HSI) takes on profound significance. HSI has emerged as a cutting-edge, non-destructive technique for fast and accurate egg quality analysis, including the detection of chick embryo mortality. However, the high cost and operational complexity compared to conventional RGB imaging are significant bottlenecks in the widespread adoption of HSI technology. To overcome these hurdles and unlock the full potential of HSI, a promising solution is hyperspectral image reconstruction from standard RGB images. This study aims to reconstruct hyperspectral images from RGB images for non-destructive early prediction of chick embryo mortality. Firstly, the performance of different image reconstruction algorithms, such as HRNET, MST++, Restormer, and EDSR were compared to reconstruct the hyperspectral images of the eggs in the early incubation period. Later, the reconstructed spectra were used to differentiate live from dead chick-producing eggs using the XGBoost and Random Forest classification methods. Among the reconstruction methods, HRNET showed impressive reconstruction performance with MRAE of 0.0955, RMSE of 0.0159, and PSNR of 36.79 dB. This study motivated that harnessing imaging technology integrated with smart sensors and data analytics has the potential to improve automation, enhance biosecurity, and optimize resource management towards sustainable agriculture 4.0.
Arun N. Sivakumar, Mateus V. Gasparino, Michael McGuire
et al.
We present a vision-based navigation system for under-canopy agricultural robots using semantic keypoints. Autonomous under-canopy navigation is challenging due to the tight spacing between the crop rows ($\sim 0.75$ m), degradation in RTK-GPS accuracy due to multipath error, and noise in LiDAR measurements from the excessive clutter. Our system, CropFollow++, introduces modular and interpretable perception architecture with a learned semantic keypoint representation. We deployed CropFollow++ in multiple under-canopy cover crop planting robots on a large scale (25 km in total) in various field conditions and we discuss the key lessons learned from this.
AbstractRecent Canadian preferential trade agreements (PTAs) include increased market access for imports of supply‐managed products (dairy and poultry). Such agreements are typically expected to create trade flows and increase supply of relatively low‐priced products in Canada. Industry groups representing Canadian producers and processors of supply‐managed products negotiated to receive approximately C$5 billion in payments from the federal government as compensation for the prospects of facing more international competition and reduced domestic sales. We discuss partial‐equilibrium simulation models that are commonly used by academics and governments to project market effects of new trade agreements, and conceptually illustrate how different assumptions about import supply conditions generate different projected market outcomes. We focus on the quota fill rates of new access commitments—most studies, including those used to inform government policies on compensation payments, assume imports increase in an amount equal to new commitments. This is often not the case, including with recent Canadian trade agreements. We apply a conceptual framework to Canada's supply‐management industry by re‐simulating a quantitative model of the Canadian dairy industry with updated information on implementation and quota fill rates. Projected market effects of trade agreements under the assumption of full import quotas are markedly different from projections that account for unfilled quotas. We discuss the political economy and welfare implications of compensation payments in light of our analysis.
Irma Wardati, Fama Rudi Atmaja, Dyah Nuning Erawati
et al.
The impact of cocoa fruit-sucking ladybugs (Helopeltis antonii Signoret) is getting higher, resulting in control by considering costs and abundant raw materials. The study aimed to ascertain the efficiency of botanical insecticides made from neem and wedusan as controllers and their effect on the behavior of cacao fruit-sucking ladybugs. The research was carried out from May to August 2021 in Petungombo Hamlet, Sepawon Village, Plosoklaten District, Kediri Regency. This study used a Randomize Block Design Non-Factorial with the factors tested including I0 (no insecticide), I1 (5% neem leaf botanical insecticide), I2 (10% wedusan leaf botanical insecticide), and I3 (neem leaf vegetable botanical combination 5% and 10% wedusan leaf botanical insecticide). Further testing uses the Least Significant Difference test with a level of 5%. The results showed that the insecticidal substances of neem leaves and botanical insecticides of wedusan leaves were effective against ladybugs sucking cocoa fruit pods with a value of Lethal Time Fifty. The results were the fastest with a combination of the two over 111 hours and significantly affected behavioral changes classified as very low to low changes.
Machine learning is expected to enable the next Industrial Revolution. However, lacking standardized and automated assembly networks, ML faces significant challenges to meet ever-growing enterprise demands and empower broad industries. In the Perspective, we argue that ML needs to first complete its own Industrial Revolution, elaborate on how to best achieve its goals, and discuss new opportunities to enable rapid translation from ML's innovation frontier to mass production and utilization.
Hannah Kerner, Catherine Nakalembe, Adam Yang
et al.
Satellite Earth observations (EO) can provide affordable and timely information for assessing crop conditions and food production. Such monitoring systems are essential in Africa, where there is high food insecurity and sparse agricultural statistics. EO-based monitoring systems require accurate cropland maps to provide information about croplands, but there is a lack of data to determine which of the many available land cover maps most accurately identify cropland in African countries. This study provides a quantitative evaluation and intercomparison of 11 publicly available land cover maps to assess their suitability for cropland classification and EO-based agriculture monitoring in Africa using statistically rigorous reference datasets from 8 countries. We hope the results of this study will help users determine the most suitable map for their needs and encourage future work to focus on resolving inconsistencies between maps and improving accuracy in low-accuracy regions.