Pasar al contenido principal

Modification of livestock housing systems to cope with climate change effects

JDC2022-049782-I

•
Nombre agencia financiadora Agencia Estatal de Investigación
Acrónimo agencia financiadora AEI
Programa Programa Estatal para Desarrollar, Atraer y Retener Talento
Subprograma Subprograma Estatal de Formación
Convocatoria Juan de la Cierva
Año convocatoria 2022
Plan Estatal 2021-2023 Plan Estatal de Investigación Científica y Técnica y de Innovación
Centro beneficiario UNIVERSITAT POLITÈCNICA DE VALÈNCIA

Resultados relacionados

Resultados totales (Incluyendo duplicados): 2
Encontrada(s) 1 página(s)

RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Evaluation of Multi-Object Detection models for automated goat behavior identification in intensive farming facilities

  • Méndez-Reyes, Daniel Alexander|||0000-0001-6414-305X
  • Fajardo-Viloria, Blanca
  • Gil-Vidal, Eliseo
  • Calvet, S.|||0000-0001-6639-9892
  • Costantino, Andrea|||0000-0003-4298-432X
[EN] Computer vision offers significant potential for the continuous, stress-free, and cost-effective monitoring of animal behavior, yet its application in goat farming remains limited. In this study, a Multi-Object Detection (MOD) model was developed to classify goat behavior into four categories: eating, standing, drinking, and lying. Zenithal videos were recorded under both light and dark conditions on an experimental goat farm, resulting in 10740 labelled annotations used to train, validate, and test 13 models leveraging Transformer- and CNN-based pretrained architectures. YOLO-based models (YOLOv8 and YOLOX) achieved the highest overall performances across both large and lightweight versions, demonstrating high detection capability and potential suitability for hardwareconstrained scenarios. YOLOX-based MOD model is identified as the most suitable for goat behavior detection due to its superior classification accuracy, fast inference speed, and fully open-source license, enabling flexible customization, deployment, and reproducibility. Other models, particularly DAB-DETR and H-DINO, underperformed, especially in detecting drinking behavior, which represents the most challenging class due to its visual similarity with standing, class imbalance, and fisheye distortion effects that affect the frame regions where drinkers are located. Mitigation strategies, including focal loss and distortion correction, improved detection accuracy for this class and reduced performance variability. The developed MOD model can be deployed for continuous group-level monitoring of goats, paving the way for scalable and efficient solutions for advanced behavioral analyses. Future works will focus on integrating tracking algorithms for animal-level insights, as well as on evaluating model generalizability across different farming conditions and goat breeds., This work was supported by the "Ayuda a Primeros Proyectos de Investigacion" program [grant number PAID-06-23] , Vice-rectorate for Research of Universitat Politecnica de Valencia (UPV) . Andrea Costantino was supported by the grant JDC2022-049782-I funded by MICIU/AEI/10.13039/501100011033 and by the European Union NextGenerationEU/PRTR. Daniel Alexander Mendez was supported by R+D+i project TED2021-130759B-C31, funded by MCIN/AEI/10.13039/501100011033/and by the "European Union NextGenerationEU/PRTR" and by the grant JDC2023-050534-I funded by MICIU/AEI/10.13039/501100011033 and for FSE+. Eliseo Gil Vidal was supported by PLF4ALL project (CIAICO/2021/212, Generalitat Valenciana).
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Advancing Digital Twin development for Controlled-Environment Agriculture: Comparative evaluation of two greenhouse energy models and their readiness for digital twin implementation

  • Costantino, Andrea|||0000-0003-4298-432X
  • Ghiberti, Alberto Lodovico
  • Bibbiani, Carlo
  • Barbaresi, Alberto
  • Fabrizio, Enrico
[EN] Robust energy modelling is essential for optimizing climate control, energy use, and productivity in greenhouses, particularly under the highly variable weather conditions of Mediterranean areas. This study develops and compares two dynamic modelling approaches: a detailed white-box simulation model with EnergyPlus and a reduced-order model based on a 5R1C resistance-capacitance scheme. Both models were calibrated and validated against experimental data acquired from a monitored Venlo-type greenhouse in central Italy, achieving good agreement with measured indoor air temperatures (RMSE 1.52 degrees C for EnergyPlus and 1.57 degrees C for 5R1C). Beyond predictive accuracy, the models were also evaluated for their readiness for integration into Digital Twin (DT) architectures, based on a structured framework specifically developed in this work. The results show that EnergyPlus offers superior interoperability and compatibility with advanced control systems, while the 5R1C model, despite its simplified structure, ensures computational agility suitable for rapid simulations and hybrid real-time applications. The comparative analysis also reveals that both models can play complementary roles within a DT environment, combining high-fidelity simulation with lightweight surrogate modelling. Overall, this study provides methodological guidance for selecting robust and scalable energy modelling approaches according to design and operational objectives. By bridging the gap between energy simulation and DT implementation, the proposed framework contributes to advancing digitalized climate control strategies for Controlled-Environment Agriculture, supporting future developments in smart greenhouse management., This work was funded by European Union-Next Generation EU, Mission 4 Component 1-Bando PRIN 2022 n. 104 del 2 febbraio 2022 "DiAGreen-Digital twin of Agricultural Greenhouses: a multi-domain tool for energy efficiency, decarbonization, enhanced production and cost reduction of intensive greenhouse cropping systems", CUP E53D23003250006. Andrea Costantino was supported by the grant JDC2022-049782-I funded by MICIU/AEI/10.13039/501100011033 and by the European Union NextGenerationEU/PRTR.