Smart agricultural technology最新文献

{"title":"Fields2Benchmark: An open-source benchmark for coverage path planning methods in agriculture","authors":"Gonzalo Mier ,&nbsp;Ana María Casado Faulí ,&nbsp;João Valente ,&nbsp;Sytze de Bruin","doi":"10.1016/j.atech.2025.101156","DOIUrl":"10.1016/j.atech.2025.101156","url":null,"abstract":"<div><div>The agricultural coverage path planning problem focuses on optimizing coverage paths for agricultural operations. Despite its importance, existing agricultural coverage path planning solutions are highly application-specific, limiting their generalizability and reproducibility. Fields2Cover offers a modular open source software library for coverage path planning, yet it lacks built-in means to benchmark its algorithms and to configure the workflow to specific tasks. It also fails on non-convex fields and operations that require reload trips. This paper introduces Fields2Benchmark, an open-source, modular benchmark designed to standardize the evaluation of agricultural coverage path planning algorithms. Fields2Benchmark includes a dataset with 350 real-world fields, featuring non-convex fields and in-field obstacles. The benchmark decomposes the problem into five units, i.e., field decomposition, headland generation, swath generation, route planning, and path planning— allowing researchers to evaluate and compare algorithms modularly. Each module supports interchangeable algorithms and objective functions, enabling customization for diverse use cases. Fields2Benchmark extends the existing Fields2Cover library by allowing task-specific choices, adding a module that splits a field into simpler parts, introducing new algorithms to shorten routes and handle refill needs, and adding tools to compare algorithms in isolation. Outputs are recorded as structured numerical data and visual representations to facilitate detailed analysis. The capabilities of the benchmark were validated on three use cases concerning field arrangement and route and path planning with and without capacity constraints. Results demonstrate its ability to handle complex field geometries, compare algorithms effectively, and evaluate computational performance. Fields2Benchmark is computationally efficient, with planning times suitable for real-time applications. It is supported by publicly available datasets and code. By standardizing agricultural coverage path planning evaluation, Fields2Benchmark aims to improve the reproducibility in this field, accelerating the research in agricultural robotics and field operations. For industrial robots, this benchmark cuts deployment costs, and shortens time to market, by simplifying the development process.</div></div>","PeriodicalId":74813,"journal":{"name":"Smart agricultural technology","volume":"12 ","pages":"Article 101156"},"PeriodicalIF":6.3,"publicationDate":"2025-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144587775","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}

{"title":"Visual servoing of an underwater robotics arm for automatic sorting of crustaceans","authors":"Giacomo Picardi ,&nbsp;Anna Astolfi ,&nbsp;Karthik Seemakurthy ,&nbsp;Bradley Hurst ,&nbsp;Elena Piana ,&nbsp;Petra Bosilj ,&nbsp;Marcello Calisti","doi":"10.1016/j.atech.2025.101168","DOIUrl":"10.1016/j.atech.2025.101168","url":null,"abstract":"<div><div>Crustaceans aquaculture is a rapidly growing sector, and automatising specific tasks can contribute to increasing its economic return as well as its sustainability. In this paper, we present an autonomous robotic system for sorting crustaceans by size, applied to crayfish farming. The system was designed in close collaboration with a crayfish farming company following a systematic user-driven approach. It consists of a waterproof robotic arm with a custom caging gripper lodging a camera, and a vision system to detect crayfish and sort them by size. All aspects of system design are presented: from manipulator and gripper design and control, to the development of the vision system, and the system integration. The system is evaluated in a controlled laboratory environment using synthetic crayfish models, and in a tank with live crayfish on a farm. Our evaluation shows that the presented system is capable of recognizing and selecting crayfish based on their size, and safely entrapping them in the caging gripper without causing any damage.</div></div>","PeriodicalId":74813,"journal":{"name":"Smart agricultural technology","volume":"12 ","pages":"Article 101168"},"PeriodicalIF":6.3,"publicationDate":"2025-07-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144571559","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}

{"title":"TinyML and IoT-enabled system for automated chicken egg quality analysis and monitoring","authors":"Omoy Kombe Hélène ,&nbsp;Martin Kuradusenge ,&nbsp;Louis Sibomana ,&nbsp;Ipyana Issah Mwaisekwa","doi":"10.1016/j.atech.2025.101162","DOIUrl":"10.1016/j.atech.2025.101162","url":null,"abstract":"<div><div>The poultry industry grapples with challenges in maintaining optimal egg quality during incubation, with environmental factors such as temperature, humidity, and lighting playing crucial roles. Traditional methods of egg quality assessment often lack precision and can be time-consuming and costly. This study addresses these challenges by introducing an innovative solution that combines Artificial Intelligence (AI) and Internet of Things (IoT) technologies, offering a transformative approach to automating the egg mirage process and improving overall egg quality analysis. The web-based program provides real-time feedback on egg quality, utilizing a Convolutional Neural Network (CNN) algorithm. Our system, implemented on Arduino Nano 33 BLE Sense, demonstrated remarkable performance with a TinyML classification F1-Score of 97.4 % and an accuracy rate of 95.79 %, paving the way for a more precise and efficient method of egg quality monitoring. The success of this research not only revolutionizes egg quality monitoring in the poultry industry but also sets a precedent for the integration of AI and IoT in addressing complex challenges in agricultural practices, including power management.</div></div>","PeriodicalId":74813,"journal":{"name":"Smart agricultural technology","volume":"12 ","pages":"Article 101162"},"PeriodicalIF":6.3,"publicationDate":"2025-07-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144595393","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}

{"title":"Precision irrigation with AI-driven optimization of plant electrophysiology","authors":"Yiting Chen ,&nbsp;Devon Scott ,&nbsp;Hieu Trung Tran ,&nbsp;Yan Sum Shirley Yip ,&nbsp;Soomin Shin ,&nbsp;Woo Soo Kim","doi":"10.1016/j.atech.2025.101169","DOIUrl":"10.1016/j.atech.2025.101169","url":null,"abstract":"<div><div>As global water scarcity intensifies and agricultural demands rise, there is a critical need for efficient irrigation management systems. Traditional autonomous irrigation solutions often depend on soil moisture and environmental sensors that indirectly reflect plant water status, leading to suboptimal irrigation practices. In this study, we introduce an innovative AI-powered autonomous irrigation system that leverages plant electrophysiological (EP) signals to directly monitor real-time plant water status for the first time. Our system integrates EP sensors, real-time signal acquisition and processing, and a convolutional neural network (CNN)-based predictive model to optimize irrigation conditions. Results indicate that EP signals can effectively differentiate between various irrigation levels with a temporal resolution of seconds, significantly enhancing water-use efficiency through real-time feedback. By optimizing water consumption using the AI algorithm, our approach can achieve at least a 10 % reduction in water use while maintaining optimal water conditions for crops. This method represents a promising advancement for precision agriculture and sustainable water management.</div></div>","PeriodicalId":74813,"journal":{"name":"Smart agricultural technology","volume":"12 ","pages":"Article 101169"},"PeriodicalIF":6.3,"publicationDate":"2025-07-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144581010","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}

{"title":"A method for herder sheep ownership identification based on an improved Mask2Former","authors":"Yaosheng Han ,&nbsp;Chunmei Li ,&nbsp;Xiangjie Huang ,&nbsp;Hao Wang ,&nbsp;Qing Dong ,&nbsp;Qihua Li ,&nbsp;Shiping Zhang","doi":"10.1016/j.atech.2025.101135","DOIUrl":"10.1016/j.atech.2025.101135","url":null,"abstract":"<div><div>Overgrazing is one of the primary causes of ecological degradation on the Qinghai Plateau, which severely restricts both the sustainable use of grasslands and the effective management of sheep herds. To address this challenge and achieve a balance between livestock and grassland resources, a semantic segmentation model based on the color features of sheep backs is proposed to accurately identify individual sheep and determine their herder affiliation. To support model training, a dedicated dataset was constructed for sheep back color segmentation and herder classification in the Qinghai Plateau region, offering a rich and diverse set of samples. In terms of model improvement, the proposed method builds upon the original Mask2Former network by introducing a Feature Pyramid Network (FPN), Haar wavelet transform, and a Convolutional Block Attention Module (CBAM). These components enhance the model's performance in complex backgrounds and fine-grained segmentation tasks by optimizing multi-scale feature fusion, improving local feature extraction, and focusing on key regions. Experimental results show that, compared with the original Mask2Former, the improved model achieves increases of 1.89%, 1.26%, 1.19%, and 1.22% in mIoU, Precision, Recall, and F1-score, respectively. These enhancements significantly improve the model's accuracy in fine-grained color segmentation. They also demonstrate its robustness and broad applicability in complex environments. This study provides an innovative solution for sheep management and herder attribution on the Qinghai-Tibet Plateau and opens a new research direction for color-feature-based image segmentation tasks.</div></div>","PeriodicalId":74813,"journal":{"name":"Smart agricultural technology","volume":"12 ","pages":"Article 101135"},"PeriodicalIF":6.3,"publicationDate":"2025-07-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144570507","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}

{"title":"ESLC-YOLOv8: Advancing real-time pineapple recognition with lightweight deep learning","authors":"Weihua Shen ,&nbsp;Mengyao Dong ,&nbsp;Zhaoxin Zhang ,&nbsp;Xiaying Hao ,&nbsp;Yuzhen Su ,&nbsp;Zhong Xue","doi":"10.1016/j.atech.2025.101139","DOIUrl":"10.1016/j.atech.2025.101139","url":null,"abstract":"<div><div>Given the current limitations of intelligent pineapple harvesting machinery and the complexity of field environments, significant challenges arise, including the color similarity between pineapples and the background, as well as substantial occlusion and overlap among plants and leaves. This study introduces an enhanced object detection algorithm, EIEStem-v7DS'Sample-LSCD-CA-YOLOv8n (ESLC-YOLOv8), designed for real-time pineapple detection in complex agricultural environments. First, we propose the EIEStem module to enhance the backbone network's convolutional layers, significantly improving edge feature extraction and spatial information preservation. Second, we introduce the v7DS (YOLOv7 DownSample) module to replace conventional downsampling operators, effectively minimizing feature information loss during resolution reduction. Finally, we design a Lightweight Shared Convolutional Detection Head (LSCD) that dramatically reduces model parameters while maintaining detection accuracy, coupled with a Coordinate Attention (CA) mechanism to reinforce critical feature representation. Extensive experimental evaluations indicate that the proposed model achieves the Recall of 0.904 and the mean average precision of 0.945, while reducing the model size to 4.5 MB. Moreover, the number of parameters and floating-point operations decrease by 8.87×10⁵ and 1.6 G, respectively, compared to the original model. The results indicated that the proposed model exhibits superior detection performance for pineapples in complex environments, striking an effective balance between detection accuracy and real-time processing.</div></div>","PeriodicalId":74813,"journal":{"name":"Smart agricultural technology","volume":"12 ","pages":"Article 101139"},"PeriodicalIF":6.3,"publicationDate":"2025-07-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144570506","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}

{"title":"A smart IoT-based hydroponics system for small-scale household in Bangladesh","authors":"Md. Abdul Awal,&nbsp;Aditi Saha Pio,&nbsp;Mushfaka Jannat Mim,&nbsp;Pronab Kumar Paul Partha,&nbsp;Md. Abdullah Al Kafi,&nbsp;Shareen Farha","doi":"10.1016/j.atech.2025.101163","DOIUrl":"10.1016/j.atech.2025.101163","url":null,"abstract":"<div><div>In Bangladesh, intensified growth in populations, diminishing arable area, and traditional farming strategies have stunted soil fertility and crop yield. Hydroponic systems become more preferred as a viable alternative for beneficial production of agriculture. Precise environmental monitoring is essential for maximizing yields; yet, existing data collection techniques are laborious and time-consuming. Real-time monitoring of essential factors in hydroponic systems may successfully mitigate these substantial obstacles to improving crop performance. This study developed an IoT-based automated monitoring and control system for the continuous measurement of important hydroponic parameters, including pH, temperature, total dissolved solids (TDS), and electrical conductivity (EC) and conducted it with the spinach (<em>Spinacia oleracea</em>) plant for field tests. The system incorporated sensors with a microcontroller and relied on Wi-Fi connection for real-time data processing through a developed mobile application. Performance testing showed good precision, with percentages of errors of 0.001 % (pH), 0.306 % (temperature), 0.062 % (TDS), and 0.002 % (EC) compared to the observed and measured data. The system automatically activated relays to regulate acid and nutrient control pumps when parameters fell below thresholds, notifying users through the mobile app for prompt intervention. After implementing this system for spinach cultivation, ideal conditions were maintained consistently, resulting in an average value of pH 6.08 ± 0.15, temperature 26 ± 1.2 °C, TDS 1150 ± 198 ppm, and EC 1.7 ± 0.45 mS/cm, approximately. Hence, this developed technology offered a feasible solution for hydroponics management issues and could be adopted extensively to improve agricultural productivity.</div></div>","PeriodicalId":74813,"journal":{"name":"Smart agricultural technology","volume":"12 ","pages":"Article 101163"},"PeriodicalIF":6.3,"publicationDate":"2025-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144587773","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}

{"title":"Optimized Yolov5s-Im for real-time apple flower detection in drone-based pollination","authors":"Shahram Hamza Manzoor ,&nbsp;Zhao Zhang ,&nbsp;Hongwen Li ,&nbsp;Qu Zhang ,&nbsp;Kuifan Chen ,&nbsp;C. Igathinathane ,&nbsp;Tianzhong Li ,&nbsp;Wei Li ,&nbsp;Muhammad Naveed Tahir ,&nbsp;Nabil Mustafa ,&nbsp;Mustafa Mhamed","doi":"10.1016/j.atech.2025.101150","DOIUrl":"10.1016/j.atech.2025.101150","url":null,"abstract":"<div><div>As traditional pollinators face increasing threats from climate change, the development of robotic pollination technology has become imperative, with apple flower detection emerging as a critical component of the technology. Deep learning (DL) advancements present novel methods in enhancing apple flower detection efficiency. However, deploying in real time on resource-constrained drone platforms demands a balance between computational efficiency and accuracy. To address this challenge, this study introduces an improved you-only-look-once version 5 small (YOLOv5s-Im) model by improving the original YOLOv5s architecture, using MobileNet version 3 as the backbone and GhostNet as the neck. This study then validated the YOLOv5s-Im performance by deploying it in real time on a drone platform designed for apple flower pollination. YOLOv5s-Im achieved an 88 % detection accuracy and averaged 41.6 pollination attempts per 3-minute flight across five tests, significantly outperforming YOLOv5s and YOLOv5s with Transformers (YOLOv5s-T) as backbone (fewer than 10 attempts), due to its 2 FPS inference speed versus their 0.05 FPS. Control tests with lightweight models YOLOv5s with ShuffleNet version 2 (YOLOv5-Sh-V2) and YOLOv5s with MobileNet version 2 (YOLOv5s-Mb-V2) as backbones, averaged 37.8 and 30.6 attempts per flight, respectively, with accuracies of 80 % and 82 % mAP and detection speeds of 1.0 FPS and 0.7 FPS, further confirming YOLOv5s-Im’s superior balance of accuracy and efficiency. Its robust accuracy (84 %-88 %) across diverse conditions—clear light (88 %), afternoon settings (86 %), angled views (87 %), and low-light shadows (84 %)—demonstrates reliability in varied orchard environments. Compared to YOLOv5s, YOLOv5s-T, YOLOv7, YOLOv8, and Faster-R-CNN, YOLOv5s-Im excels with precision (90.6 %), recall (87.7 %), mAP50 (91.2 %), and F1-score (89.42 %), while reducing GFLOPS by 89 % and model size by 85 %, achieving high frame rates (227 FPS on NVIDIA RTX 4060 Ti, 22 FPS on Jetson Xavier, 4.56 FPS on Intel NUC11TNKi3). These results make YOLOv5s-Im an effective solution for real-time apple flower detection under natural lighting conditions in drone-based pollination systems.</div></div>","PeriodicalId":74813,"journal":{"name":"Smart agricultural technology","volume":"12 ","pages":"Article 101150"},"PeriodicalIF":6.3,"publicationDate":"2025-07-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144587776","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}

{"title":"“Smart support for fruit farm business decision-making: A framework for digital controlling adoption”","authors":"Luis Müller ,&nbsp;Robert Luer ,&nbsp;Wolfgang Lentz","doi":"10.1016/j.atech.2025.101157","DOIUrl":"10.1016/j.atech.2025.101157","url":null,"abstract":"<div><div>Decision-makers in horticultural enterprises face significant challenges, including structural changes toward fewer and larger enterprises, increasing weather extremes, skilled labor shortages, and increasing sustainability demands. Addressing these challenges requires the use of effective business management instruments. However, initial research suggests that controlling is practiced infrequently and with limited intensity in the sector. This study investigates barriers to and drivers of the adoption of controlling, focusing on the role of digital transformation in enhancing its use. Nineteen semi-structured interviews and farm inspections were conducted with 28 current and prospective farm managers in the fruit-growing sector in Germany and Luxembourg. Results show that decision-making is primarily driven by experience and intuition. Lack of experience with controlling hinders the visibility of its benefits, thus preventing the acquisition of controlling expertise. In the absence of such expertise, controlling methods cannot be effectively implemented, rendering the benefits elusive. Additionally, the limited availability of processed data restricts meaningful business analyses. Farm management and information systems (FMISs) provide minimal controlling functionalities. Reporting obligations relating to crop protection are the primary drivers for FMIS adoption. Therefore, reporting obligations in external accounting and data requirements for funding and loan procurement serve as entry points for controlling-focused software solutions. Information and controlling systems thus hold potential to enhance routine decision-making, rendering the benefits of controlling more tangible. This study contributes a conceptual framework to broaden the understanding of the often negatively perceived cost-benefit ratio of controlling in small and medium-sized enterprises and identifies strategies to strengthen its practical relevance.</div></div>","PeriodicalId":74813,"journal":{"name":"Smart agricultural technology","volume":"12 ","pages":"Article 101157"},"PeriodicalIF":6.3,"publicationDate":"2025-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144581186","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}

{"title":"YOLOv9-GSSA model for efficient soybean seedlings and weeds detection","authors":"Baihe Liang ,&nbsp;Liangchen Hu ,&nbsp;Guangxing Liu ,&nbsp;Peng Hu ,&nbsp;Shaosheng Xu ,&nbsp;Biao Jie","doi":"10.1016/j.atech.2025.101134","DOIUrl":"10.1016/j.atech.2025.101134","url":null,"abstract":"<div><div>To monitor soybean seedlings growth in real time, an effective method for accurately identifying seedlings and removing weeds is essential. Challenges include the small size and morphological similarity of seedlings and weeds, complicating conventional detection methods. To tackle these issues, we propose a real-time detection algorithm called YOLOv9-GSSA. The improved Mosaic-Dense algorithm increases object density at the model's input layer, enhancing its ability to capture detailed features. Additionally, the GSSA neck optimization module, combining GSConv and Gated Self-Attention, supports key information extraction and multi-scale feature interaction. The Swin-GSSA prediction head further utilizes spatial positional information, improving small object detection and handling overlapping occlusion. Experimental results show our model achieves a mAP of 47.5% with a detection speed of 23.42 ms per image, suitable for real-time monitoring. The enhanced model significantly improves the detection of soybean seedlings and weeds, making it a valuable tool for managing farmland effectively. This ultimately aids in precise yield estimation and decision-making in precision agriculture.</div></div>","PeriodicalId":74813,"journal":{"name":"Smart agricultural technology","volume":"12 ","pages":"Article 101134"},"PeriodicalIF":6.3,"publicationDate":"2025-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144548891","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}