CNR Author
operator:
and / or
Typology
operator:
and / or
Language
operator:
and / or
Date
operator:
and / or
Rights
operator:
and / or
2026
Journal article
Open Access
Balancing accuracy and cost in precision agriculture: a few-shot learning approach for efficient weed − crop segmentation
Catalano Nico, Luglio Sofia Matilde, Chiatti Agnese, Sportelli Mino, Frasconi Christian, Facchinetti Davide, Matteucci Matteo
Autonomous weeding, a task requiring expertise at the intersection of Computer Vision and Agronomy depends on accurate segmentation of crops and weeds from robot-collected images. Traditional segmentation models (i.e. YOLO) require large, densely annotated datasets, whose creation is costly and labor-intensive. In contrast, Few-Shot Learning (FSL) methods can learn from minimal annotated examples and significantly reduce the costs of dataset creation. This study evaluates the ability of a FSL architecture, HDMNet, to perform crop and weed segmentation using only a single annotated support image. Its performance retains 73–80% of the accuracy compared with widely used, annotation-intensive detectors designed for large datasets such as YOLOv5 and YOLOv8 when detecting bean and corn plants. Because reliable estimates of annotation effort are lacking in agriculture, we provide a quantitative assessment of the labor required to produce pixel-level labels. Preparing the 2,069-images ‘Early’ dataset required approximately 181 h, while 102-images ‘Refined’ dataset still required approximately 186 h. Labeling accounted for approximately 25 and 30 h, respectively. These findings show that increasing annotation granularity sharply raises effort without proportional accuracy gains, making dataset scale more beneficial than mask detail for YOLO-based models. In contrast, few-shot methods achieve competitive performance while eliminating most annotation labor. The study is further supported by the release of a new dataset from the 2023 ACRE field competition, including the ‘Early’ and ‘Refined’ versions. Overall, the findings offer practical guidance for designing efficient datasets for agricultural image analysis and demonstrate that FSL can substantially reduce autonomous weeding systems deployment costs.Source: COMPUTERS AND ELECTRONICS IN AGRICULTURE, vol. 245
DOI: 10.1016/j.compag.2026.111524
Metrics:
Catalano Nico, Luglio Sofia Matilde, Chiatti Agnese, Sportelli Mino, Frasconi Christian, Facchinetti Davide, Matteucci Matteo
Autonomous weeding, a task requiring expertise at the intersection of Computer Vision and Agronomy depends on accurate segmentation of crops and weeds from robot-collected images. Traditional segmentation models (i.e. YOLO) require large, densely annotated datasets, whose creation is costly and labor-intensive. In contrast, Few-Shot Learning (FSL) methods can learn from minimal annotated examples and significantly reduce the costs of dataset creation. This study evaluates the ability of a FSL architecture, HDMNet, to perform crop and weed segmentation using only a single annotated support image. Its performance retains 73–80% of the accuracy compared with widely used, annotation-intensive detectors designed for large datasets such as YOLOv5 and YOLOv8 when detecting bean and corn plants. Because reliable estimates of annotation effort are lacking in agriculture, we provide a quantitative assessment of the labor required to produce pixel-level labels. Preparing the 2,069-images ‘Early’ dataset required approximately 181 h, while 102-images ‘Refined’ dataset still required approximately 186 h. Labeling accounted for approximately 25 and 30 h, respectively. These findings show that increasing annotation granularity sharply raises effort without proportional accuracy gains, making dataset scale more beneficial than mask detail for YOLO-based models. In contrast, few-shot methods achieve competitive performance while eliminating most annotation labor. The study is further supported by the release of a new dataset from the 2023 ACRE field competition, including the ‘Early’ and ‘Refined’ versions. Overall, the findings offer practical guidance for designing efficient datasets for agricultural image analysis and demonstrate that FSL can substantially reduce autonomous weeding systems deployment costs.Source: COMPUTERS AND ELECTRONICS IN AGRICULTURE, vol. 245
DOI: 10.1016/j.compag.2026.111524
Metrics:
See at:
Computers and Electronics in Agriculture 
| CNR IRIS | www.sciencedirect.com | CNR IRIS 
2025
Journal article
Open Access
Beyond prototypes: what is missing to fill the gaps in IoT-enabled hydroponics platforms
Sportelli M., La Rosa D., Crivello A., Pineda-Medina Dunia, Bacco M., Barsocchi P.
Hydroponic agriculture, when combined with Internet of Things (IoT) technologies, provides a promising pathway to sustainable and efficient food production. This paper aims to systematically review and analyze recent advancements in IoT-based management for hydroponic systems, with a particular focus on assessing the technological maturity of current solutions, identifying existing gaps, and outlining promising directions for future research and development. Based on a review of 74 recent studies, the findings reveal a fragmented landscape characterized by custom-built solutions, predominantly relying on open-source microcontrollers and WiFi connectivity, but with limited adoption of standardized protocols and interoperable platforms. The majority of applications emphasize monitoring of core hydroponic parameters such as pH, EC, and temperature, while emerging uses of machine learning remain at an early stage. Few systems demonstrate readiness for commercial deployment or integration within broader smart agriculture ecosystems. By clarifying the current state of IoT-enabled hydroponics, this review highlights both the opportunities and the challenges in advancing from isolated prototypes toward robust, scalable systems capable of real-world application.Source: HORTICULTURAE, vol. 11 (issue 11)
DOI: 10.3390/horticulturae11111322
Metrics:
Sportelli M., La Rosa D., Crivello A., Pineda-Medina Dunia, Bacco M., Barsocchi P.
Hydroponic agriculture, when combined with Internet of Things (IoT) technologies, provides a promising pathway to sustainable and efficient food production. This paper aims to systematically review and analyze recent advancements in IoT-based management for hydroponic systems, with a particular focus on assessing the technological maturity of current solutions, identifying existing gaps, and outlining promising directions for future research and development. Based on a review of 74 recent studies, the findings reveal a fragmented landscape characterized by custom-built solutions, predominantly relying on open-source microcontrollers and WiFi connectivity, but with limited adoption of standardized protocols and interoperable platforms. The majority of applications emphasize monitoring of core hydroponic parameters such as pH, EC, and temperature, while emerging uses of machine learning remain at an early stage. Few systems demonstrate readiness for commercial deployment or integration within broader smart agriculture ecosystems. By clarifying the current state of IoT-enabled hydroponics, this review highlights both the opportunities and the challenges in advancing from isolated prototypes toward robust, scalable systems capable of real-world application.Source: HORTICULTURAE, vol. 11 (issue 11)
DOI: 10.3390/horticulturae11111322
Metrics:
See at:
Horticulturae | CNR IRIS | www.mdpi.com | CNR IRIS
2025
Conference article
Open Access
End-user requirements modelling: an experience report from digital agriculture
Mannari C., Sportelli M., Meesala H., Okoye O. F., Lepore F., Bacco M., Brunori G., Malizia A., Ferrari A.
Context and motivation: End-user development focuses on enabling non-professional programmers to create or extend software applications on their own. However, before beginning the development process, software engineering best practices recommend performing requirements engineering (RE) activities, including requirements modelling.Question/problem: There is limited research on how end-users can model system requirements. Principal ideas/results: In this experience report, we investigate the problem of end-user requirements modelling in an EU-funded project about agricultural digitalisation. Specifically, a team of agronomists was directly involved in the creation of UML, iStar, and BPMN diagrams to model the transformation of socio-technical processes in four different concrete scenarios. They followed a formalisation procedure proposed within an RE method designed to help stakeholders evaluate the impact of agricultural digitalisation. Starting from textual reports including a description of the process as-is and the process-to-be, they followed step-by-step guidelines for model creation. Contribution: This paper reports insights from the experience from the viewpoint of the agronomists and software engineers involved. We identify nine key lessons that highlight the added value of end-user requirements modelling for achieving a shared and in-depth understanding of the socio-technical processes under analysis.Source: LECTURE NOTES IN COMPUTER SCIENCE, vol. 15588, pp. 304-316. Barcellona, Spain, 2025
DOI: 10.1007/978-3-031-88531-0_22
Project(s): CODECS
Metrics:
Mannari C., Sportelli M., Meesala H., Okoye O. F., Lepore F., Bacco M., Brunori G., Malizia A., Ferrari A.
Context and motivation: End-user development focuses on enabling non-professional programmers to create or extend software applications on their own. However, before beginning the development process, software engineering best practices recommend performing requirements engineering (RE) activities, including requirements modelling.Question/problem: There is limited research on how end-users can model system requirements. Principal ideas/results: In this experience report, we investigate the problem of end-user requirements modelling in an EU-funded project about agricultural digitalisation. Specifically, a team of agronomists was directly involved in the creation of UML, iStar, and BPMN diagrams to model the transformation of socio-technical processes in four different concrete scenarios. They followed a formalisation procedure proposed within an RE method designed to help stakeholders evaluate the impact of agricultural digitalisation. Starting from textual reports including a description of the process as-is and the process-to-be, they followed step-by-step guidelines for model creation. Contribution: This paper reports insights from the experience from the viewpoint of the agronomists and software engineers involved. We identify nine key lessons that highlight the added value of end-user requirements modelling for achieving a shared and in-depth understanding of the socio-technical processes under analysis.Source: LECTURE NOTES IN COMPUTER SCIENCE, vol. 15588, pp. 304-316. Barcellona, Spain, 2025
DOI: 10.1007/978-3-031-88531-0_22
Project(s): CODECS
Metrics:
See at:
CNR IRIS | link.springer.com | doi.org | CNR IRIS | CNR IRIS
2025
Journal article
Open Access
Web components for late blight (Phytophthora infestans (Mont.) De Bary) and early blight (Alternaria solani Sor.) outbreaks forecast on Solanum tuberosum L. in Cuba under future climate scenarios
Pineda Medina D., Crivello A., Sportelli M., Bianchini M., Miranda Cabrera I.
The invasive nature of late blight (Phytophthora infestans [Mont.] De Bary) and early blight (Alternaria solani Sor.) has caused important losses in the potato crop, and studies point to climatic variability as one of the most significant causes. The objective of this work was to predict the probability of outbreak of late blight and early blight epiphytotic diseases through weather patterns during the potato harvest season. Future meteorological data for the years 2024 to 2075 obtained from the National Institute of Meteorology in Cuba were used. For the late blight forecasting model development, disease behavior rules and a Random Forest model were used and validated on a small real-filed dataset. For early blight prediction model a framework based on disease behavior rules and phenological age of the crop (P-Days) was developed. The web system was implemented using the Python programming language, the Flask and Bootstrap frameworks, and the necessary libraries, and PyCharm as the development environment. Likewise, the PostgreSQL manager and PgAdmin were used for data management and as a tool for information administration. A web system was obtained that alerts on the probability of outbreak of late blight and early blight in the provinces of Mayabeque, Villa Clara and Ciego de Ávila, important potato-producing regions in Cuba. The forecast for the year 2025 was analyzed and it was found that Ciego de Avila and Mayabeque show a higher number of critical days for each of the diseases. In the month of March, late blight proliferation reached its peak, while early blight outbreak appeared to be more intense in January. The web system is an intelligent tool to guide farmers in making the necessary decisions and prevent an epidemic outbreak.Source: SMART AGRICULTURAL TECHNOLOGY, vol. 11 (issue 101003)
DOI: 10.1016/j.atech.2025.101003
Metrics:
Pineda Medina D., Crivello A., Sportelli M., Bianchini M., Miranda Cabrera I.
The invasive nature of late blight (Phytophthora infestans [Mont.] De Bary) and early blight (Alternaria solani Sor.) has caused important losses in the potato crop, and studies point to climatic variability as one of the most significant causes. The objective of this work was to predict the probability of outbreak of late blight and early blight epiphytotic diseases through weather patterns during the potato harvest season. Future meteorological data for the years 2024 to 2075 obtained from the National Institute of Meteorology in Cuba were used. For the late blight forecasting model development, disease behavior rules and a Random Forest model were used and validated on a small real-filed dataset. For early blight prediction model a framework based on disease behavior rules and phenological age of the crop (P-Days) was developed. The web system was implemented using the Python programming language, the Flask and Bootstrap frameworks, and the necessary libraries, and PyCharm as the development environment. Likewise, the PostgreSQL manager and PgAdmin were used for data management and as a tool for information administration. A web system was obtained that alerts on the probability of outbreak of late blight and early blight in the provinces of Mayabeque, Villa Clara and Ciego de Ávila, important potato-producing regions in Cuba. The forecast for the year 2025 was analyzed and it was found that Ciego de Avila and Mayabeque show a higher number of critical days for each of the diseases. In the month of March, late blight proliferation reached its peak, while early blight outbreak appeared to be more intense in January. The web system is an intelligent tool to guide farmers in making the necessary decisions and prevent an epidemic outbreak.Source: SMART AGRICULTURAL TECHNOLOGY, vol. 11 (issue 101003)
DOI: 10.1016/j.atech.2025.101003
Metrics:
See at:
Smart Agricultural Technology | Smart Agricultural Technology | Usiena air - Università di Siena | CNR IRIS | www.sciencedirect.com | Usiena air - Università di Siena | GitHub | CNR IRIS
2024
Conference article
Open Access
An IoT Platform for smart hydroponics: building blocks and open challenges
Sportelli M., Crivello A., La Rosa D., Bacco M., Incrocci L., Barsocchi P.
Hydroponics addresses inefficiencies in traditional soil-based farming by optimizing water, nutrients, and pesticide use. This method has the potential to boost crop yields and significantly reduce water consumption, tackling issues of inefficient irrigation and fertilization. This study presents the development of an IoT-based platform designed to optimize the management of smart hydroponic systems. The proposed platform facilitates real-time monitoring of key environmental parameters for hydroponic farming as well as automatic regulation of greenhouse conditions, such as temperature, humidity and nutrient levels. The platform leverages middleware software to ensure seamless communication and data management, enabling efficient decision-making processes. The primary aim of this work is to enhance the productivity and efficiency of hydroponic systems through a scalable and user-friendly solution by integrating different sensors and actuators that could be accessible through both web and mobile applications. The platform's open and flexible architecture supports the integration of advanced sensing technologies and artificial intelligence, contributing to the digitalization of agriculture and promoting environmental sustainability.DOI: 10.1109/metroagrifor63043.2024.10948780
Project(s): CODECS
Metrics:
Sportelli M., Crivello A., La Rosa D., Bacco M., Incrocci L., Barsocchi P.
Hydroponics addresses inefficiencies in traditional soil-based farming by optimizing water, nutrients, and pesticide use. This method has the potential to boost crop yields and significantly reduce water consumption, tackling issues of inefficient irrigation and fertilization. This study presents the development of an IoT-based platform designed to optimize the management of smart hydroponic systems. The proposed platform facilitates real-time monitoring of key environmental parameters for hydroponic farming as well as automatic regulation of greenhouse conditions, such as temperature, humidity and nutrient levels. The platform leverages middleware software to ensure seamless communication and data management, enabling efficient decision-making processes. The primary aim of this work is to enhance the productivity and efficiency of hydroponic systems through a scalable and user-friendly solution by integrating different sensors and actuators that could be accessible through both web and mobile applications. The platform's open and flexible architecture supports the integration of advanced sensing technologies and artificial intelligence, contributing to the digitalization of agriculture and promoting environmental sustainability.DOI: 10.1109/metroagrifor63043.2024.10948780
Project(s): CODECS
Metrics:
See at:
CNR IRIS | ieeexplore.ieee.org | doi.org | CNR IRIS
2024
Conference article
Open Access
Towards a toolkit for socio-technical process modelling in agriculture: a pilot study
Mannari C., Sportelli M., Okoye O. F., Bacco M., Ferrari A., Malizia A., Brunori G.
Digital technologies are transforming agriculture, affecting social, institutional, economic, environmental, and technological dimensions. To ensure sustainable development, it is essential to anticipate these impacts and create conditions for sustainable change. Living Labs (LLs) concept facilitates this by involving various stakeholders in co-designing solutions. This paper presents a socio-technical process modelling method using Model-driven requirements engineering (MoDRE) techniques. It employs UML class diagrams, iStar diagrams, and BPMN diagrams to model process structures, goals, and flows. The method, part of the Horizon Europe project CODECS, involves data collection, diagram design, and iterative feedback, tested in a precision irrigation pilot study in Tuscany. Preliminary results demonstrate the method's effectiveness in supporting interdisciplinary teams, fostering better communication, and aiding in the analysis of digitalisation impacts on agricultural processes. Furthermore, the discussion with stakeholders allowed the fine-tuning of the models and enriched the method for co-creating the diagrams with a toolkit composed of a set of guidelines for eliciting process-relevant information from LLs, a checklist and a detailed procedure for graphical representation.DOI: 10.1109/metroagrifor63043.2024.10948866
Metrics:
Mannari C., Sportelli M., Okoye O. F., Bacco M., Ferrari A., Malizia A., Brunori G.
Digital technologies are transforming agriculture, affecting social, institutional, economic, environmental, and technological dimensions. To ensure sustainable development, it is essential to anticipate these impacts and create conditions for sustainable change. Living Labs (LLs) concept facilitates this by involving various stakeholders in co-designing solutions. This paper presents a socio-technical process modelling method using Model-driven requirements engineering (MoDRE) techniques. It employs UML class diagrams, iStar diagrams, and BPMN diagrams to model process structures, goals, and flows. The method, part of the Horizon Europe project CODECS, involves data collection, diagram design, and iterative feedback, tested in a precision irrigation pilot study in Tuscany. Preliminary results demonstrate the method's effectiveness in supporting interdisciplinary teams, fostering better communication, and aiding in the analysis of digitalisation impacts on agricultural processes. Furthermore, the discussion with stakeholders allowed the fine-tuning of the models and enriched the method for co-creating the diagrams with a toolkit composed of a set of guidelines for eliciting process-relevant information from LLs, a checklist and a detailed procedure for graphical representation.DOI: 10.1109/metroagrifor63043.2024.10948866
Metrics:
See at:
CNR IRIS | ieeexplore.ieee.org | doi.org | CNR IRIS | CNR IRIS