This review is motivated by the urgent need to improve soil organic carbon (SOC) assessment methods, which are vital for enhancing soil health, addressing climate change, and promoting carbon farming. By employing a structured approach that involves a systematic literature search, data extraction, and analysis, 86 relevant studies were identified. These studies were evaluated to address the following specific research questions: (a) What are the state-of-the-art approaches in sampling, modeling, and data acquisition? and (b) What are the key challenges, open issues, potential advancements, and future directions needed to enhance the effectiveness of carbon farming practices? The findings indicate that while traditional SOC assessment techniques remain foundational, there is a significant shift towards incorporating model-based methods, machine learning models, proximal spectroscopy, and remote sensing technologies. These emerging approaches primarily serve as complementary to laboratory analyses, enhancing the overall accuracy and reliability of SOC assessments. Despite these advancements, challenges such as soil spatial and temporal variability, high financial costs, and limitations in measurement accuracy continue to hinder progress. This review also highlights the necessity for scalable, cost-effective, and precise SOC measurement tools, alongside supportive policies and incentives that encourage farmer adoption. Finally, the development of a “System-of-Systems” approach that integrates sampling, sensing, and modeling offers a promising pathway to balancing cost and accuracy, ultimately supporting carbon farming practices.

Accurate crop yield prediction is vital towards optimizing agricultural productivity. Machine Learning (ML) has shown promise in this field; however, its application to legume crops, especially to lupin, remains limited, while many models lack interpretability, hindering real-world adoption. To bridge this literature gap, an interpretable ML framework was developed for predicting lupin yield using Sentinel-2 remote sensing data integrated with georeferenced yield measurements. Data preprocessing involved computing vegetation indices, removing outliers, addressing multicollinearity, normalizing feature scales, and applying data augmentation techniques to correct target imbalance. Subsequently, six ML models were evaluated representing different algorithmic strategies. Among them, XGBoost showed the best performance ($R^2$ = 0.8756) and low error values across $MAE$, $MSE$, and $RMSE$ metrics. To enhance model transparency, SHapley Additive exPlanations (SHAP) values were applied to interpret the feature contributions of the XGBoost model. The Enhanced Vegetation Index ($EVI$) and Normalized Difference Vegetation Index ($NDVI$) were found to be key predictors of crop yield, both showing a positive correlation with higher values reflecting greater vegetation vigor and corresponding to increased yield. These were followed by $B03$ (green) and $B12$ (short-wave infrared), which captured key reflectance properties associated with chlorophyll activity and water content, respectively. Both of them substantially influence photosynthetic efficiency and plant health, ultimately affecting yield potential.

This study addresses a critical gap in human activity recognition (HAR) research by enhancing both the explainability and efficiency of activity classification in collaborative human–robot systems, particularly in agricultural environments. While traditional HAR models often prioritize improving overall classification accuracy, they typically lack transparency in how sensor data contribute to decision-making. To fill this gap, this study integrates explainable artificial intelligence, specifically SHapley Additive exPlanations (SHAP), thus enhancing the interpretability of the model. Data were collected from 20 participants who wore five inertial measurement units (IMUs) at various body positions while performing material handling tasks involving an unmanned ground vehicle in a field collaborative harvesting scenario. The results highlight the central role of torso-mounted sensors, particularly in the lumbar region, cervix, and chest, in capturing core movements, while wrist sensors provided useful complementary information, especially for load-related activities. The XGBoost-based model, selected mainly for allowing an in-depth analysis of feature contributions by considerably reducing the complexity of calculations, demonstrated strong performance in HAR. The findings indicate that future research should focus on enlarging the dataset, investigating the use of additional sensors and sensor placements, and performing real-world trials to enhance the model’s generalizability and adaptability for practical agricultural applications.

Enhancing agricultural machinery field efficiency offers substantial benefits for farm management by optimizing the available resources, thereby reducing cost, maximizing productivity, and supporting sustainability. Field efficiency is influenced by several unpredictable and stochastic factors that are difficult to determine due to the inherent variability in field configurations and operational conditions. This study aimed to simplify field efficiency estimation by training machine learning regression algorithms on data generated from a farm management information system covering a combination of different field areas and shapes, working patterns, and machine-related parameters. The gradient-boosting regression-based model was the most effective, achieving a high mean $R^2$ value of 0.931 in predicting field efficiency, by taking into account only basic geometric field indices. The developed model showed also strong predictive performance for indicative agricultural fields located in Europe and North America, reducing considerably the computational time by an average of 73.4% compared to the corresponding analytical approach. Overall, the results of this study highlight the potential of machine learning for simplifying field efficiency prediction without requiring detailed knowledge of a plethora of variables associated with agricultural operations. This can be particularly valuable for farmers who need to make informed decisions about resource allocation and operational planning.

Soybean expeller (SBE), a by-product of soybean oil extraction through the extruding-expelling process, is widely used as a protein source in animal feed and soy-based foods. This study evaluated fungal contamination and mycotoxin levels in SBE samples from 11 extruding-expelling facilities in Argentina, assessing fungal load, moisture content (MC), and mycotoxin profiles. Fungal biota was quantified through colony forming unit (CFU) counts and identified via morphological analysis, while mycotoxins were quantified using liquid chromatography and tandem mass spectrometry. CFU counts were low (0 to 4 CFU g⁻¹ DM), with Penicillium spp. (28.0%) and Mucoraceae (family) (25.6%) being the most frequently isolated genera. Deoxynivalenol (DON) and aflatoxins (AFB₁, AFB₂, AFG₁, and AFG₂) were detected in 20% to 40% of the samples. The average concentration was 215.19 µg kg⁻¹ for DON and 41.68, 0.39, and 0.34 µg kg⁻¹ for AFB₁, AFG₁, and AFG₂, respectively. Although most mycotoxin concentrations were below regulatory limits, a few samples exceeded the threshold for DON (8.6%) and AFB₁ (2.9%). Co-occurrence of two mycotoxins was observed in 60% of the samples. These results highlight the importance of monitoring fungal contamination and mycotoxin levels to ensure the safety and quality of SBE for feed and food applications.

Reliable and fast communication between unmanned aerial vehicles (UAVs) and unmanned ground vehicles (UGVs) is essential for effective coordination in agricultural settings, particularly when human involvement is part of the system. This study systematically compares two communication architectures representing centralized and decentralized communication frameworks: (a) MAVLink (decentralized) and (b) Farm Management Information System (FMIS) (centralized). Field experiments were conducted in both empty field and orchard environments, using a rotary UAV for worker detection and a UGV responding to intent signaled through color-coded hats. Across 120 trials, the system performance was assessed in terms of communication reliability, latency, energy consumption, and responsiveness. FMIS consistently demonstrated higher message delivery success rates (97% in both environments) than MAVLink (83% in the empty field and 70% in the orchard). However, it resulted in higher UGV resource usage. Conversely, MAVLink achieved reduced UGV power draw and lower latency, but it was more affected by obstructed settings and also resulted in increased UAV battery consumption. In conclusion, MAVLink is suitable for time-sensitive operations that require rapid feedback, while FMIS is better suited for tasks that demand reliable communication in complex agricultural environments. Consequently, the selection between MAVLink and FMIS should be guided by the specific mission goals and environmental conditions.