Pub Date : 2026-01-05DOI: 10.1016/j.atech.2026.101788
M. Naim , A. Boukhizzou
Agriculture is undergoing a rapid digital transformation that challenges its ecological, social, and ethical foundations. This study explores how the transition from two revolutions, from Agriculture 4.0 (A4.0) to Agriculture 5.0 (A5.0), redefines the relationship between technology and agroecology. The dominant approach of A4.0, driven by automation, big data, and artificial intelligence, has enhanced efficiency but missed many agroecological principles, mainly those contributing to secure social equity and responsibility. Emerging as a corrective paradigm, A5.0 seeks to integrate technological progress with agroecological principles that value the social and human dimension. Adopting a scoping review following PRISMA-ScR guidelines, scientific publications indexed in Scopus and CABI up to October 2025 were screened and coded to assess how current A5.0 research embeds the thirteen agroecological principles defined by the High-Level Panel of Experts in 2019. A total of 136 documents were analysed through bibliometric and thematic synthesis. Results show that A5.0 represents a philosophical and structural evolution beyond the efficiency-oriented logic of A4.0, integrating distributed computing, explainable artificial intelligence, digital twins, and collaborative robotics within ecologically restorative and socially inclusive frameworks. However, while A5.0 strengthens resource efficiency, resilience, and certain social segments through open-source technologies and participatory design, gaps remain in policy coherence, emotional engagement, and human–machine co-learning. To address these, the study proposes two complementary agroecological principles, cognitive symbiosis and emotional ecology, emphasising shared intelligence and affective stewardship between humans, machines, and ecosystems. Overall, Agriculture 5.0 reframes digitalisation as a human-ecological partnership that can operationalise agroecology’s ethical goals if governed by inclusion, transparency, and regeneration rather than control and optimisation.
{"title":"Animating the transition: How agriculture 5.0 revitalises agroecological principles","authors":"M. Naim , A. Boukhizzou","doi":"10.1016/j.atech.2026.101788","DOIUrl":"10.1016/j.atech.2026.101788","url":null,"abstract":"<div><div>Agriculture is undergoing a rapid digital transformation that challenges its ecological, social, and ethical foundations. This study explores how the transition from two revolutions, from Agriculture 4.0 (A4.0) to Agriculture 5.0 (A5.0), redefines the relationship between technology and agroecology. The dominant approach of A4.0, driven by automation, big data, and artificial intelligence, has enhanced efficiency but missed many agroecological principles, mainly those contributing to secure social equity and responsibility. Emerging as a corrective paradigm, A5.0 seeks to integrate technological progress with agroecological principles that value the social and human dimension. Adopting a scoping review following PRISMA-ScR guidelines, scientific publications indexed in Scopus and CABI up to October 2025 were screened and coded to assess how current A5.0 research embeds the thirteen agroecological principles defined by the High-Level Panel of Experts in 2019. A total of 136 documents were analysed through bibliometric and thematic synthesis. Results show that A5.0 represents a philosophical and structural evolution beyond the efficiency-oriented logic of A4.0, integrating distributed computing, explainable artificial intelligence, digital twins, and collaborative robotics within ecologically restorative and socially inclusive frameworks. However, while A5.0 strengthens resource efficiency, resilience, and certain social segments through open-source technologies and participatory design, gaps remain in policy coherence, emotional engagement, and human–machine co-learning. To address these, the study proposes two complementary agroecological principles, cognitive symbiosis and emotional ecology, emphasising shared intelligence and affective stewardship between humans, machines, and ecosystems. Overall, Agriculture 5.0 reframes digitalisation as a human-ecological partnership that can operationalise agroecology’s ethical goals if governed by inclusion, transparency, and regeneration rather than control and optimisation.</div></div>","PeriodicalId":74813,"journal":{"name":"Smart agricultural technology","volume":"13 ","pages":"Article 101788"},"PeriodicalIF":5.7,"publicationDate":"2026-01-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145925629","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
With increasingly stringent requirements for greenhouse gas reduction, assessing current tractor drivetrains and developing cleaner alternatives are of growing importance. Designing such systems requires extensive field-load data, which remain difficult to obtain due to the absence of scalable, long-term measurement methods. This study presents a non-invasive methodology for evaluating tractor field loads using only controller area network (CANBUS) messages, universally available on modern tractors. A static mathematical model of a hydromechanical continuously variable transmission (IHMCVT) is developed, validated through experimental tests, and applied to map drawbar forces under real-world conditions. The model employs convergence-based calculations with multiple feedback loops and shallow neural networks to determine hydrostatic unit efficiencies. Validation against power-based measurements from road and field tests demonstrates accuracy within ±10%, suitable for practical applications. Applied to one year of operational data, the method reveals that tractors mainly operate at low speeds (<10 km h⁻¹) under heavy soil tillage and transport conditions. Specific fuel consumption mapping further highlights inefficiencies due to engine–transmission interactions. Overall, the proposed CANBUS-based model provides a reliable, scalable, and low-complexity approach for real-world mission profiling and future drivetrain optimization.
{"title":"CANBUS to drawbar load estimation: Mapping real-world tractor loads for mission profiling","authors":"Luca Colendi , Manuel Tentarelli , Massimiliano Varani , Michele Mattetti","doi":"10.1016/j.atech.2026.101777","DOIUrl":"10.1016/j.atech.2026.101777","url":null,"abstract":"<div><div>With increasingly stringent requirements for greenhouse gas reduction, assessing current tractor drivetrains and developing cleaner alternatives are of growing importance. Designing such systems requires extensive field-load data, which remain difficult to obtain due to the absence of scalable, long-term measurement methods. This study presents a non-invasive methodology for evaluating tractor field loads using only controller area network (CANBUS) messages, universally available on modern tractors. A static mathematical model of a hydromechanical continuously variable transmission (IHMCVT) is developed, validated through experimental tests, and applied to map drawbar forces under real-world conditions. The model employs convergence-based calculations with multiple feedback loops and shallow neural networks to determine hydrostatic unit efficiencies. Validation against power-based measurements from road and field tests demonstrates accuracy within ±10%, suitable for practical applications. Applied to one year of operational data, the method reveals that tractors mainly operate at low speeds (<10 km h⁻¹) under heavy soil tillage and transport conditions. Specific fuel consumption mapping further highlights inefficiencies due to engine–transmission interactions. Overall, the proposed CANBUS-based model provides a reliable, scalable, and low-complexity approach for real-world mission profiling and future drivetrain optimization.</div></div>","PeriodicalId":74813,"journal":{"name":"Smart agricultural technology","volume":"13 ","pages":"Article 101777"},"PeriodicalIF":5.7,"publicationDate":"2026-01-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145925730","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Accurate identification of wheat varieties at the seedling stage is crucial for maintaining seed purity and optimizing field management. However, the subtle phenotypic variations among seedlings present a significant challenge for visual recognition. To address this, we propose SeedlingNet, a novel deep learning model specifically designed for fine-grained wheat seedling variety classification. The core innovations of SeedlingNet include: The Kolmogorov-Arnold-based Convolutional Attention (KCA) mechanism, which dynamically enhances feature representation by replacing static activation functions with learnable, adaptive ones; A multi-scale feature fusion architecture that integrates hierarchical features to capture both global and local characteristics. We established a comprehensive image dataset of 13,600 images representing 17 wheat varieties at the early growth stage. Experimental results demonstrate that SeedlingNet achieves a remarkable classification accuracy of 99.26 %, outperforming traditional machine learning methods and mainstream deep learning models. Ablation studies confirm the significant impact of the KCA module and the multi-scale fusion structure on the model's performance. This research provides an effective, non-destructive tool for early-stage variety identification, with strong potential for precision agriculture applications.The dataset is licensed in Zhang, Wenbo (2025), "Seedings", Mendeley Data, V1, doi: 10.17632/f8ykx4sz6w. 1.
{"title":"Multi branch model based on cross scale feature fusion for wheat seedling variety recognition","authors":"Zhang Wenbo , Zhang Ziyang , Xi Chengyu , Zhang Qingshan","doi":"10.1016/j.atech.2026.101785","DOIUrl":"10.1016/j.atech.2026.101785","url":null,"abstract":"<div><div>Accurate identification of wheat varieties at the seedling stage is crucial for maintaining seed purity and optimizing field management. However, the subtle phenotypic variations among seedlings present a significant challenge for visual recognition. To address this, we propose SeedlingNet, a novel deep learning model specifically designed for fine-grained wheat seedling variety classification. The core innovations of SeedlingNet include: The Kolmogorov-Arnold-based Convolutional Attention (KCA) mechanism, which dynamically enhances feature representation by replacing static activation functions with learnable, adaptive ones; A multi-scale feature fusion architecture that integrates hierarchical features to capture both global and local characteristics. We established a comprehensive image dataset of 13,600 images representing 17 wheat varieties at the early growth stage. Experimental results demonstrate that SeedlingNet achieves a remarkable classification accuracy of 99.26 %, outperforming traditional machine learning methods and mainstream deep learning models. Ablation studies confirm the significant impact of the KCA module and the multi-scale fusion structure on the model's performance. This research provides an effective, non-destructive tool for early-stage variety identification, with strong potential for precision agriculture applications.The dataset is licensed in Zhang, Wenbo (2025), \"Seedings\", Mendeley Data, V1, doi: 10.17632/f8ykx4sz6w. 1.</div></div>","PeriodicalId":74813,"journal":{"name":"Smart agricultural technology","volume":"13 ","pages":"Article 101785"},"PeriodicalIF":5.7,"publicationDate":"2026-01-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145925729","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-01-02DOI: 10.1016/j.atech.2026.101782
Qing Yang , Abdullah Al Mamun , Zafir Khan Mohamed Makhbul , Mohammad Nurul Hassan Reza , Muhammad Mehedi Masud
This study investigates the impact of big data analytics capabilities (BDAC) on the adoption of Internet of Things-enabled agro-technologies (AIoT) and their subsequent influence on sustainability performance, using dynamic capabilities theory as the conceptual framework. A self-administered survey questionnaire was distributed online, gathering 285 responses from farms in five key regions of China: Henan, Shandong, Anhui, Hebei, and Jiangsu. The data were analyzed using partial least squares structural equation modeling. The findings show that data, technology, technical skills, managerial skills, and a data-driven culture are positively and significantly associated with AIoT, which in turn, significantly influences sustainability performance. Interestingly, basic resources were found to have no significant effect on AIoT adoption. Mediation analysis indicated that AIoT acts as a significant mediator in the relationship between BDAC and sustainability performance, with the exception of basic resources. This study is pioneering in its empirical examination of how distinct dimensions of BDAC influence AIoT adoption to enhance farm sustainability performance, offering one of the first large-scale, farm-level validations of dynamic capabilities theory in agriculture. The findings suggest that developing managerial and technical skills, along with fostering a data-driven culture, is more critical for achieving sustainable digital transformation in agriculture than simply possessing basic resources.
{"title":"The nexus of big data, internet of things-enabled agro-technologies, and farm performance","authors":"Qing Yang , Abdullah Al Mamun , Zafir Khan Mohamed Makhbul , Mohammad Nurul Hassan Reza , Muhammad Mehedi Masud","doi":"10.1016/j.atech.2026.101782","DOIUrl":"10.1016/j.atech.2026.101782","url":null,"abstract":"<div><div>This study investigates the impact of big data analytics capabilities (BDAC) on the adoption of Internet of Things-enabled agro-technologies (AIoT) and their subsequent influence on sustainability performance, using dynamic capabilities theory as the conceptual framework. A self-administered survey questionnaire was distributed online, gathering 285 responses from farms in five key regions of China: Henan, Shandong, Anhui, Hebei, and Jiangsu. The data were analyzed using partial least squares structural equation modeling. The findings show that data, technology, technical skills, managerial skills, and a data-driven culture are positively and significantly associated with AIoT, which in turn, significantly influences sustainability performance. Interestingly, basic resources were found to have no significant effect on AIoT adoption. Mediation analysis indicated that AIoT acts as a significant mediator in the relationship between BDAC and sustainability performance, with the exception of basic resources. This study is pioneering in its empirical examination of how distinct dimensions of BDAC influence AIoT adoption to enhance farm sustainability performance, offering one of the first large-scale, farm-level validations of dynamic capabilities theory in agriculture. The findings suggest that developing managerial and technical skills, along with fostering a data-driven culture, is more critical for achieving sustainable digital transformation in agriculture than simply possessing basic resources.</div></div>","PeriodicalId":74813,"journal":{"name":"Smart agricultural technology","volume":"13 ","pages":"Article 101782"},"PeriodicalIF":5.7,"publicationDate":"2026-01-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145925630","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-31DOI: 10.1016/j.atech.2025.101773
Renato Herrig Furlanetto , Ana Claudia Buzanini , Arnold Walter Schumann , Nathan Shawn Boyd
The development of vision-guided smart sprayers depends heavily on diverse and high-quality datasets, either collected in the field or sourced from public repositories. However, for certain crops, acquiring such datasets is both labor-intensive and challenging. Text-to-image generative models, such as those based on Stable Diffusion (SD), offer a promising alternative by generating synthetic images from a limited number of real samples. This study explores the potential of two SD models (SDXL and SD3.5L) to generate realistic images of Geranium carolinianum growing alongside strawberry plants. The goal was to assess the YOLOv11 model performance by evaluating training and validation strategies involving: 1) Real images only, 2) 50 % Real + 50 % SD3.5L, 3) 50 % Real + 50 % SDXL, 4) SD3.5L synthetic only, and 5) SDXL synthetic only. A total of 400 high-resolution field images were collected during the 2024-2025 strawberry season in Florida, with 50 representative images of G. carolinianum at various growth stages selected to fine-tune the SD models. The quality of generated images was evaluated using Fréchet Inception Distance (FID), Inception Score (IS), and Structural Similarity Index Measure (SSIM). YOLOv11 model performance was assessed using F1-scores and mean average precision at IoU thresholds of 50 % to 95 % (mAP50, mAP50-95 %). SDXL achieved superior FID and IS scores and produced more photorealistic images, whereas SD3.5L generated images with greater morphological accuracy, variability, and higher SSIM values. YOLOv11 models trained on a mix of real and synthetic images performed comparably to models trained only on real data, with no statistically significant differences in detection accuracy. Models trained exclusively on synthetic data yielded lower performance. These results highlight the potential of generative models to supplement dataset creation, significantly reducing field collection efforts, particularly valuable during off-season periods, for species with limited geographic distribution, under environmental constraints, or for rare or newly discovered species.
视觉引导智能喷雾器的开发在很大程度上依赖于各种高质量的数据集，这些数据集要么在现场收集，要么来自公共存储库。然而，对于某些作物，获取这样的数据集既费力又具有挑战性。文本到图像生成模型，例如基于稳定扩散（SD）的生成模型，通过从有限数量的真实样本生成合成图像，提供了一个有前途的替代方案。本研究探索了两种SD模型（SDXL和SD3.5L）的潜力，以生成与草莓植物一起生长的天竺葵卡罗莱纳的真实图像。目标是通过评估训练和验证策略来评估YOLOv11模型的性能，包括：1)仅真实图像，2)50% Real + 50% SD3.5L, 3) 50% Real + 50% SDXL, 4)仅SD3.5L合成，5)仅SDXL合成。本研究收集了美国佛罗里达州2024-2025年草莓季的400张高分辨率田间图像，选取了50张不同生育期具有代表性的G. carolinianum图像，对SD模型进行了微调。生成图像的质量使用fr起始距离（FID）、起始分数（IS）和结构相似指数测量（SSIM）进行评估。使用f1分数和IoU阈值为50%至95% （mAP50, mAP50- 95%）时的平均精度来评估YOLOv11模型的性能。SDXL获得了更高的FID和IS分数，生成的图像更逼真，而SD3.5L生成的图像具有更高的形态学准确性、可变性和更高的SSIM值。使用真实和合成图像混合训练的YOLOv11模型的表现与仅使用真实数据训练的模型相当，在检测精度方面没有统计学上的显著差异。仅在合成数据上训练的模型产生了较低的性能。这些结果突出了生成模型补充数据集创建的潜力，显着减少了野外收集工作，特别是在淡季期间，对于地理分布有限，受环境约束的物种，或稀有或新发现的物种。
{"title":"Enhancing weed detection in strawberry plasticulture: Evaluating stable diffusion models for synthetic image generation of Geranium carolinianum","authors":"Renato Herrig Furlanetto , Ana Claudia Buzanini , Arnold Walter Schumann , Nathan Shawn Boyd","doi":"10.1016/j.atech.2025.101773","DOIUrl":"10.1016/j.atech.2025.101773","url":null,"abstract":"<div><div>The development of vision-guided smart sprayers depends heavily on diverse and high-quality datasets, either collected in the field or sourced from public repositories. However, for certain crops, acquiring such datasets is both labor-intensive and challenging. Text-to-image generative models, such as those based on Stable Diffusion (SD), offer a promising alternative by generating synthetic images from a limited number of real samples. This study explores the potential of two SD models (SDXL and SD3.5L) to generate realistic images of <em>Geranium carolinianum</em> growing alongside strawberry plants. The goal was to assess the YOLOv11 model performance by evaluating training and validation strategies involving: 1) Real images only, 2) 50 % Real + 50 % SD3.5L, 3) 50 % Real + 50 % SDXL, 4) SD3.5L synthetic only, and 5) SDXL synthetic only. A total of 400 high-resolution field images were collected during the 2024-2025 strawberry season in Florida, with 50 representative images of <em>G. carolinianum</em> at various growth stages selected to fine-tune the SD models. The quality of generated images was evaluated using Fréchet Inception Distance (FID), Inception Score (IS), and Structural Similarity Index Measure (SSIM). YOLOv11 model performance was assessed using F1-scores and mean average precision at IoU thresholds of 50 % to 95 % (mAP50, mAP50-95 %). SDXL achieved superior FID and IS scores and produced more photorealistic images, whereas SD3.5L generated images with greater morphological accuracy, variability, and higher SSIM values. YOLOv11 models trained on a mix of real and synthetic images performed comparably to models trained only on real data, with no statistically significant differences in detection accuracy. Models trained exclusively on synthetic data yielded lower performance. These results highlight the potential of generative models to supplement dataset creation, significantly reducing field collection efforts, particularly valuable during off-season periods, for species with limited geographic distribution, under environmental constraints, or for rare or newly discovered species.</div></div>","PeriodicalId":74813,"journal":{"name":"Smart agricultural technology","volume":"13 ","pages":"Article 101773"},"PeriodicalIF":5.7,"publicationDate":"2025-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145925632","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-31DOI: 10.1016/j.atech.2025.101776
Waqar Ahmed Qureshi , Jianmin Gao , Mazhar Hussain Tunio , Osama Elsherbiny , Xiangxin Gao , Luxi Wang , Abdallah Harold Mosha
This study evaluates plasma-integrated aeroponics as a strategy to improve nitrogen management, productivity, and food quality in controlled-environment lettuce (Lactuca sativa L.) production. Soil cultivation wastes water and leaches nutrients, hydroponics can limit root oxygenation, and aeroponics while improving aeration may increase leaf nitrate. A smart aeroponic platform integrating non-thermal plasma (NTP) modules with real-time gas sensing was developed. Four plasma delivery modes (direct vs. indirect; dry vs mist) at 15 and 38 kV were compared with soil, hydroponic, and plasma-OFF aeroponic controls. Optical emission spectroscopy confirmed RONS generation (e.g., NO and OH), and NH₃/O₃ were continuously monitored. The 38 kV indirect mist configuration (T5) gave the best overall performance, producing 259.4 g plant⁻¹ fresh biomass and 15.1 g plant⁻¹ dry biomass, with stronger root development (70 cm plant⁻¹ root length) and higher photosynthetic capacity (peak 21 µmol CO₂ m⁻² s⁻¹) alongside improved chlorophyll/PSII efficiency. High-voltage plasma induced greater nutrient-solution drift during a 7-day renewal cycle (T5 day-7 pH 5.44; EC 2.11 dS m⁻¹), indicating active modulation of root-zone chemistry. Despite yield gains, leaf nitrite remained very low (2.29 mg kg⁻¹ FW) and nitrate stayed within reported dietary safety limits (2900 mg kg⁻¹ FW). Shoot nutrient and functional quality increased, with higher soluble sugars (29.41 g kg⁻¹; +116% vs. soil), vitamin C, phenolic-related antioxidant potential, and antioxidant capacity (+102% vs. soil). Overall, biomass production, photosynthetic efficiency, and nutrient accumulation are significantly improved by plasma-assisted aeroponics, particularly the 38 kV indirect mist mode, indicating its promise as an effective method for producing high-quality lettuce.
{"title":"Smart plasma-enhanced aeroponic cultivation of lettuce (lactuca sativa l.): Comparative evaluation with soil and hydroponic systems on growth responses, nitrogen absorption, and nutritional quality","authors":"Waqar Ahmed Qureshi , Jianmin Gao , Mazhar Hussain Tunio , Osama Elsherbiny , Xiangxin Gao , Luxi Wang , Abdallah Harold Mosha","doi":"10.1016/j.atech.2025.101776","DOIUrl":"10.1016/j.atech.2025.101776","url":null,"abstract":"<div><div>This study evaluates plasma-integrated aeroponics as a strategy to improve nitrogen management, productivity, and food quality in controlled-environment lettuce (Lactuca sativa L.) production. Soil cultivation wastes water and leaches nutrients, hydroponics can limit root oxygenation, and aeroponics while improving aeration may increase leaf nitrate. A smart aeroponic platform integrating non-thermal plasma (NTP) modules with real-time gas sensing was developed. Four plasma delivery modes (direct vs. indirect; dry vs mist) at 15 and 38 kV were compared with soil, hydroponic, and plasma-OFF aeroponic controls. Optical emission spectroscopy confirmed RONS generation (e.g., NO and OH), and NH₃/O₃ were continuously monitored. The 38 kV indirect mist configuration (T5) gave the best overall performance, producing 259.4 g plant⁻¹ fresh biomass and 15.1 g plant⁻¹ dry biomass, with stronger root development (70 cm plant⁻¹ root length) and higher photosynthetic capacity (peak 21 µmol CO₂ m⁻² s⁻¹) alongside improved chlorophyll/PSII efficiency. High-voltage plasma induced greater nutrient-solution drift during a 7-day renewal cycle (T5 day-7 pH 5.44; EC 2.11 dS m⁻¹), indicating active modulation of root-zone chemistry. Despite yield gains, leaf nitrite remained very low (2.29 mg kg⁻¹ FW) and nitrate stayed within reported dietary safety limits (2900 mg kg⁻¹ FW). Shoot nutrient and functional quality increased, with higher soluble sugars (29.41 g kg⁻¹; +116% vs. soil), vitamin C, phenolic-related antioxidant potential, and antioxidant capacity (+102% vs. soil). Overall, biomass production, photosynthetic efficiency, and nutrient accumulation are significantly improved by plasma-assisted aeroponics, particularly the 38 kV indirect mist mode, indicating its promise as an effective method for producing high-quality lettuce.</div></div>","PeriodicalId":74813,"journal":{"name":"Smart agricultural technology","volume":"13 ","pages":"Article 101776"},"PeriodicalIF":5.7,"publicationDate":"2025-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145925634","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-30DOI: 10.1016/j.atech.2025.101755
Khokan Kumar Saha , Rabin Dulal , Leigh Schmidtke , Lihong Zheng , Xiaodi Huang , Andrew Clark
Identifying grape cultivars and vineyard origins is essential for ensuring wine quality, traceability, and authenticity. This study presents a rapid liquid chromatography-quadrupole time-of-flight mass spectrometry (LC-qTOF MS) approach that substantially reduces analysis time compared with conventional methods. Mass spectral data of grape cultivars collected from 11 vineyards across New South Wales, Australia, within an m/z range of 45-1500 were analysed to discriminate cultivars based on their geographical origin. To address the high dimensionality of the dataset, feature extraction was performed using principal component analysis (PCA), supervised stacked autoencoder (SAE), and uniform manifold approximation and projection (UMAP). Among these dimensionality reduction techniques, PCA exhibited the most robust and consistent performance. Fifteen machine learning models were then evaluated to assess classification accuracy. The Random Forest (RF) model, when combined with PCA, achieved the highest accuracy (95.6%), effectively distinguishing grape cultivars from the 11 vineyard sites. Overall, these findings demonstrate that integrating rapid LC-qTOF MS with machine learning provides a powerful and efficient framework for authenticating grape cultivars and classifying vineyard origins, highlighting the potential of data-driven approaches for food provenance and quality assurance.
{"title":"Discrimination of vineyard origin and grape cultivars using rapid mass spectrometry and machine learning","authors":"Khokan Kumar Saha , Rabin Dulal , Leigh Schmidtke , Lihong Zheng , Xiaodi Huang , Andrew Clark","doi":"10.1016/j.atech.2025.101755","DOIUrl":"10.1016/j.atech.2025.101755","url":null,"abstract":"<div><div>Identifying grape cultivars and vineyard origins is essential for ensuring wine quality, traceability, and authenticity. This study presents a rapid liquid chromatography-quadrupole time-of-flight mass spectrometry (LC-qTOF MS) approach that substantially reduces analysis time compared with conventional methods. Mass spectral data of grape cultivars collected from 11 vineyards across New South Wales, Australia, within an <em>m</em>/<em>z</em> range of 45-1500 were analysed to discriminate cultivars based on their geographical origin. To address the high dimensionality of the dataset, feature extraction was performed using principal component analysis (PCA), supervised stacked autoencoder (SAE), and uniform manifold approximation and projection (UMAP). Among these dimensionality reduction techniques, PCA exhibited the most robust and consistent performance. Fifteen machine learning models were then evaluated to assess classification accuracy. The Random Forest (RF) model, when combined with PCA, achieved the highest accuracy (95.6%), effectively distinguishing grape cultivars from the 11 vineyard sites. Overall, these findings demonstrate that integrating rapid LC-qTOF MS with machine learning provides a powerful and efficient framework for authenticating grape cultivars and classifying vineyard origins, highlighting the potential of data-driven approaches for food provenance and quality assurance.</div></div>","PeriodicalId":74813,"journal":{"name":"Smart agricultural technology","volume":"13 ","pages":"Article 101755"},"PeriodicalIF":5.7,"publicationDate":"2025-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145925631","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-30DOI: 10.1016/j.atech.2025.101774
Weixin Zhai , Le Chen , Jinming Liu , Zhou Guo , Qian Qian , Jiawen Pan , Caicong Wu , Xiangzeng Kong
Agricultural machinery trajectory operation mode identification aims to determine the spatiotemporal distribution characteristics of agricultural machinery trajectories, thereby understanding the machinery's activity at each time step. It is a multivariate time series classification task. However, existing methods are limited to capturing feature interactions only within local trajectory points, without fully utilizing the global contextual information hidden within the trajectory. To effectively model the contextual information of trajectories to accurately identify operation modes pattern of agricultural machinery, we propose a dual-domain context-aware network (DCANet). First, we propose a multi-angle feature enhancement method that uses physical kinematics formulas and mathematical statistics methods to expand the feature set of trajectories, thereby fully exploiting the inherent information. We then design a frequency-domain context-aware module consisting of two components: a local awareness bottleneck module and a learnable frequency-domain attention context module. The former is used to model the local dependencies. Based on this, the latter uses a discrete wavelet transform and a multi-head attention mechanism to calculate the correlation and dependency of trajectory features in multiple frequency subspaces, adaptively analysing trajectory feature changes in the time-frequency domain. Finally, we design an information aggregation context module. Based on the construction of the spatiotemporal relationship graph of agricultural machinery trajectory points, we combine it with a multi-filter graph convolution operator to capture the multi-scale features of trajectory points in different channels. To validate DCANet, we evaluated it on 3 datasets against 12 related methods. The results demonstrate that DCANet achieves superior performance, with accuracies of 89.03%, 90.24%, and 90.47%, and F1-scores of 71.81%, 84.62%, and 90.43%, respectively.
{"title":"Dual-domain Context-aware network in time series classification for identifying agricultural machinery trajectory operation mode","authors":"Weixin Zhai , Le Chen , Jinming Liu , Zhou Guo , Qian Qian , Jiawen Pan , Caicong Wu , Xiangzeng Kong","doi":"10.1016/j.atech.2025.101774","DOIUrl":"10.1016/j.atech.2025.101774","url":null,"abstract":"<div><div>Agricultural machinery trajectory operation mode identification aims to determine the spatiotemporal distribution characteristics of agricultural machinery trajectories, thereby understanding the machinery's activity at each time step. It is a multivariate time series classification task. However, existing methods are limited to capturing feature interactions only within local trajectory points, without fully utilizing the global contextual information hidden within the trajectory. To effectively model the contextual information of trajectories to accurately identify operation modes pattern of agricultural machinery, we propose a dual-domain context-aware network (DCANet). First, we propose a multi-angle feature enhancement method that uses physical kinematics formulas and mathematical statistics methods to expand the feature set of trajectories, thereby fully exploiting the inherent information. We then design a frequency-domain context-aware module consisting of two components: a local awareness bottleneck module and a learnable frequency-domain attention context module. The former is used to model the local dependencies. Based on this, the latter uses a discrete wavelet transform and a multi-head attention mechanism to calculate the correlation and dependency of trajectory features in multiple frequency subspaces, adaptively analysing trajectory feature changes in the time-frequency domain. Finally, we design an information aggregation context module. Based on the construction of the spatiotemporal relationship graph of agricultural machinery trajectory points, we combine it with a multi-filter graph convolution operator to capture the multi-scale features of trajectory points in different channels. To validate DCANet, we evaluated it on 3 datasets against 12 related methods. The results demonstrate that DCANet achieves superior performance, with accuracies of 89.03%, 90.24%, and 90.47%, and F1-scores of 71.81%, 84.62%, and 90.43%, respectively.</div></div>","PeriodicalId":74813,"journal":{"name":"Smart agricultural technology","volume":"13 ","pages":"Article 101774"},"PeriodicalIF":5.7,"publicationDate":"2025-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145925563","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-29DOI: 10.1016/j.atech.2025.101765
Chen Gu , Huaiyang Liu , Yunhao You , Qianghao Zeng , Zhenxiang Zhou , Ming Song , Yun Shi , Tong Tian
Leaf nitrogen content (LNC) is an essential physiological indicator for assessing the growth status of wheat. However, the accuracy and generalization of remote sensing-based monitoring models are often constrained by spatial and temporal variability. To overcome these limitations, this study proposes a cascaded optimization framework that integrates signal enhancement, feature selection, and intelligent optimization algorithms. First, the raw spectral data were preprocessed using Savitzky-Golay (SG) smoothing and a first-order derivative transformation, followed by a multi-scale continuous wavelet transform (CWT). Then, relevant spectral bands were identified through Pearson correlation analysis, and further dimensionality reduction was performed using the successive projections algorithm (SPA). Finally, two regression models were developed: an Extreme Gradient Boosting (XGBoost) model optimized with the Sparrow Search Algorithm (SSA) and an Extreme Learning Machine (ELM) optimized with the Artificial Hummingbird Algorithm (AHA). The XGBoost-SSA model demonstrated superior predictive performance on the test set, achieving a coefficient of determination (R²) of approximately 0.79, a root mean square error (RMSE) of approximately 0.14 mg/g, and a mean absolute percentage error (MAPE) of approximately 9.31%. On an independent external validation set, the XGBoost-SSA model also showed strong generalization capability, maintaining an R2 of approximately 0.73, an RMSE of approximately 0.14 mg/g, and a MAPE of approximately 9.63%. These findings underscore the value of medium-scale CWT-SPA in spectral feature extraction and highlight the advantages of swarm intelligence algorithms in enhancing regression model performance. Overall, the proposed approach provides a reliable solution for high-precision nitrogen monitoring using hyperspectral remote sensing and supports data-driven applications in smart agriculture.
{"title":"Hyperspectral inversion of leaf nitrogen content in wheat by integrating CWT-SPA feature optimization and XGBoost-SSA model","authors":"Chen Gu , Huaiyang Liu , Yunhao You , Qianghao Zeng , Zhenxiang Zhou , Ming Song , Yun Shi , Tong Tian","doi":"10.1016/j.atech.2025.101765","DOIUrl":"10.1016/j.atech.2025.101765","url":null,"abstract":"<div><div>Leaf nitrogen content (LNC) is an essential physiological indicator for assessing the growth status of wheat. However, the accuracy and generalization of remote sensing-based monitoring models are often constrained by spatial and temporal variability. To overcome these limitations, this study proposes a cascaded optimization framework that integrates signal enhancement, feature selection, and intelligent optimization algorithms. First, the raw spectral data were preprocessed using Savitzky-Golay (SG) smoothing and a first-order derivative transformation, followed by a multi-scale continuous wavelet transform (CWT). Then, relevant spectral bands were identified through Pearson correlation analysis, and further dimensionality reduction was performed using the successive projections algorithm (SPA). Finally, two regression models were developed: an Extreme Gradient Boosting (XGBoost) model optimized with the Sparrow Search Algorithm (SSA) and an Extreme Learning Machine (ELM) optimized with the Artificial Hummingbird Algorithm (AHA). The XGBoost-SSA model demonstrated superior predictive performance on the test set, achieving a coefficient of determination (R²) of approximately 0.79, a root mean square error (RMSE) of approximately 0.14 mg/g, and a mean absolute percentage error (MAPE) of approximately 9.31%. On an independent external validation set, the XGBoost-SSA model also showed strong generalization capability, maintaining an R<sup>2</sup> of approximately 0.73, an RMSE of approximately 0.14 mg/g, and a MAPE of approximately 9.63%. These findings underscore the value of medium-scale CWT-SPA in spectral feature extraction and highlight the advantages of swarm intelligence algorithms in enhancing regression model performance. Overall, the proposed approach provides a reliable solution for high-precision nitrogen monitoring using hyperspectral remote sensing and supports data-driven applications in smart agriculture.</div></div>","PeriodicalId":74813,"journal":{"name":"Smart agricultural technology","volume":"13 ","pages":"Article 101765"},"PeriodicalIF":5.7,"publicationDate":"2025-12-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145925567","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-12-29DOI: 10.1016/j.atech.2025.101767
Eleni I. Katsarou , Yiannis Kiouvrekis , Daphne T. Lianou , Themistoklis Giannoulis , Efthymia Petinaki , Charalambia K. Michael , Natalia G.C. Vasileiou , Anargyros Skoulakis , Angeliki I. Katsafadou , Nikolaos Solomakos , Dimitriοs C. Chatzopoulos , Maria V. Bourganou , Dimitra V. Liagka , George C. Fthenakis , Dimitris Kalatzis , Vasia S. Mavrogianni
<div><div>Among the zoonotic pathogens, <em>Coxiella burnetii</em> and <em>Listeria</em> spp. are of particular significance, as they may cause relevant infections in people. These often arise as the consequence of consumption of dairy products contaminated with these pathogens. Hence, there is a scope for developing a tool through the application of machine learning methodologies, by means of which one may produce an accurate prediction regarding the presence of the above pathogens in the milk produced at small ruminant (sheep, goat) farms. The objective of the present study was the development of computational models for predictions for the presence of these pathogens in the bulk-tank milk in dairy sheep and goat farms. The presence of <em>C. burnetii</em> or <em>Listeria</em> spp. therein was defined as the target value and computational models were constructed for their prediction, through the use of independent variables (<em>n</em> = 11 for <em>C. burnetii</em> and <em>n</em> = 8 for <em>Listeria</em> spp.) related to factors prevailing at the farms. Seven machine learning tools were employed: Decision Trees (36 different models evaluated for each pathogen), Random Forest (9 models), LightGBM (24 models), XGBoost (54 models), k-Nearest Neighbours (32 models), Support Vector Machines (3 models) and Neural Networks (24 models). In all, 182 evaluations were performed for each pathogen. For the prediction of the presence of <em>C. burnetii</em> in the bulk-tank milk, Random Forest and XGBoost were the two tools that produced the best measures of quality: the median value for four measures of quality of the predictions (accuracy, precision, recall, F1-score) evaluated was over 0.942 for each of these two tools. The results of the analysis for SHapley Additive exPlanations values for the impact of the independent variables in the prediction of the presence of the pathogen indicated that (a) the management system applied on the farm and (b) the average wind speed at the location of the farm during the 15 days prior to sampling, were the variables that mostly influenced the prediction outcome. In the case of <em>Listeria</em> spp., meaningful classification could not be achieved, due to the highly imbalanced and limited nature of the dataset. For this pathogen, the models heavily favoured the majority class and produced predictions that defaulted almost entirely to ‘no presence’ of the pathogen; for the minority class, recall and precision were as low as 0.300. The findings of this study indicate that machine learning algorithms can be employed for the prediction of the presence of <em>C. burnetii</em> in the milk produced in dairy sheep and goat farms. These computational prediction models were developed using field data collected during countrywide studies. The models showed excellent performance and thus can be employed as adjunct tools to support clinical decisions in the health management of dairy small ruminant farms and the production of saf
{"title":"The contribution of explainable machine learning in the detection of two zoonotic pathogens: Coxiella burnetii and Listeria spp., in the milk of dairy small ruminants","authors":"Eleni I. Katsarou , Yiannis Kiouvrekis , Daphne T. Lianou , Themistoklis Giannoulis , Efthymia Petinaki , Charalambia K. Michael , Natalia G.C. Vasileiou , Anargyros Skoulakis , Angeliki I. Katsafadou , Nikolaos Solomakos , Dimitriοs C. Chatzopoulos , Maria V. Bourganou , Dimitra V. Liagka , George C. Fthenakis , Dimitris Kalatzis , Vasia S. Mavrogianni","doi":"10.1016/j.atech.2025.101767","DOIUrl":"10.1016/j.atech.2025.101767","url":null,"abstract":"<div><div>Among the zoonotic pathogens, <em>Coxiella burnetii</em> and <em>Listeria</em> spp. are of particular significance, as they may cause relevant infections in people. These often arise as the consequence of consumption of dairy products contaminated with these pathogens. Hence, there is a scope for developing a tool through the application of machine learning methodologies, by means of which one may produce an accurate prediction regarding the presence of the above pathogens in the milk produced at small ruminant (sheep, goat) farms. The objective of the present study was the development of computational models for predictions for the presence of these pathogens in the bulk-tank milk in dairy sheep and goat farms. The presence of <em>C. burnetii</em> or <em>Listeria</em> spp. therein was defined as the target value and computational models were constructed for their prediction, through the use of independent variables (<em>n</em> = 11 for <em>C. burnetii</em> and <em>n</em> = 8 for <em>Listeria</em> spp.) related to factors prevailing at the farms. Seven machine learning tools were employed: Decision Trees (36 different models evaluated for each pathogen), Random Forest (9 models), LightGBM (24 models), XGBoost (54 models), k-Nearest Neighbours (32 models), Support Vector Machines (3 models) and Neural Networks (24 models). In all, 182 evaluations were performed for each pathogen. For the prediction of the presence of <em>C. burnetii</em> in the bulk-tank milk, Random Forest and XGBoost were the two tools that produced the best measures of quality: the median value for four measures of quality of the predictions (accuracy, precision, recall, F1-score) evaluated was over 0.942 for each of these two tools. The results of the analysis for SHapley Additive exPlanations values for the impact of the independent variables in the prediction of the presence of the pathogen indicated that (a) the management system applied on the farm and (b) the average wind speed at the location of the farm during the 15 days prior to sampling, were the variables that mostly influenced the prediction outcome. In the case of <em>Listeria</em> spp., meaningful classification could not be achieved, due to the highly imbalanced and limited nature of the dataset. For this pathogen, the models heavily favoured the majority class and produced predictions that defaulted almost entirely to ‘no presence’ of the pathogen; for the minority class, recall and precision were as low as 0.300. The findings of this study indicate that machine learning algorithms can be employed for the prediction of the presence of <em>C. burnetii</em> in the milk produced in dairy sheep and goat farms. These computational prediction models were developed using field data collected during countrywide studies. The models showed excellent performance and thus can be employed as adjunct tools to support clinical decisions in the health management of dairy small ruminant farms and the production of saf","PeriodicalId":74813,"journal":{"name":"Smart agricultural technology","volume":"13 ","pages":"Article 101767"},"PeriodicalIF":5.7,"publicationDate":"2025-12-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145925633","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号