-
[hal-05599703] PADI-Location-AR-EN: A normalized Arabic–English spatial entity dataset for epidemiological surveillance
The location of events in multilingual texts, particularly in Arabic, represents a challenge for epidemiological monitoring. Systems such as PADI-web rely on English translation to extract spatial entities, but the scarcity of annotated spatial entities in Arabic can hamper the reliability of translations and extraction. In this context, PADI-Location-AR-EN, which is a dataset of 328 spatial entities that were manually extracted from 96 Arabic-language news articles collected by the PADI-web epidemiological monitoring system, is presented in this paper. Each entity was manually translated into English, normalized using the GeoNames database, and then classified according to its type and spatial category. The dataset can be used to evaluate the translation quality of three machine translation systems (DeepL, Microsoft Azure and Reverso) as well as the performance of named entity recognition models on the translated texts.
ano.nymous@ccsd.cnrs.fr.invalid (Fatima Ezzahra El Houbri) 22 Apr 2026
https://hal.inrae.fr/hal-05599703v1
-
[hal-05602030] Do farmers trust digital decision support tools on pesticide use?
Les outils d'aide à la décision (OAD) sont de plus en plus mobilisés en agriculture pour raisonner l'usage des pesticides en fournissant des recommandations de traitement, souvent au jour près. Malgré les performances reconnues de ces outils, on constate que les agriculteurs suivent relativement peu les recommandations qu'ils fournissent. Cet article cherche à expliquer ces décisions à travers le prisme de la confiance en l'outil et des biais de perception du risque par les agriculteurs. Nous étudions le cas d'un OAD français destiné à la gestion du mildiou de la pomme de terre. À partir d'une base de données originale contenant les recommandations émises par l'OAD et les décisions de traitement des agriculteurs abonnés, nous analysons le suivi des recommandations de l'OAD par les agriculteurs ainsi que l'évolution de ce suivi à la suite d'épisodes de contamination par le mildiou. Nos résultats montrent que les agriculteurs suivent moins les recommandations de non traitement que celles qui recommandent de traiter. Ce résultat s'accentue pour les agriculteurs ayant subi une contamination dans le passé. On montre qu'ils choisissent de moins suivre les recommandations que l'OAD ait fait une erreur de prédiction dans le passé, ou non. Ces résultats suggèrent que le suivi ou non des recommandations est moins lié au niveau de confiance des agriculteurs dans l'OAD, que dans l'évolution de la perception du risque de contamination future, plus forte chez l'agriculteur ayant déjà subi une contamination dans le passé.
ano.nymous@ccsd.cnrs.fr.invalid (Alban Cornier) 24 Apr 2026
https://hal.inrae.fr/hal-05602030v1
-
[hal-05616230] Does digital technology use lead to sustainable business model innovation? The case of the legume value chain in Occitanie, France
The expectations of digital technologies in sustainable agricultural development to deal with several social and environmental issues in our current food system are considerable (Dagoudou et al., 2023). The co-evolution of digital and agro- ecological trajectories could present potential in changing practices and relationships between actors in agri-food value chains (Klerkx et al., 2019). However, applying these technologies in combination with agro-ecology is yet barely studied in agri-food business models and value chains (Slot et al., 2025). There is a need to understand how the use of various digital technologies affects current and leads to new sustainable, agro-ecological and circular business models in agri-food systems. This study is based on the case of the regional legume value chain in Occitanie (France), because this region is a pioneer in regional policies and projects to support the agro-ecological transition (Occitanie Green Pact 2020 (FEDER-FSE+, 2021)) and local legume production is part of climate-resilient activities in EU agri- food systems (Ferreira et al., 2021). The case study includes ten exploratory interviews with regional experts (researchers, consultants and public actors) of 30 minutes to one hour (during May-June 2024), and seven interviews with different business managers in the legume value chain of about one hour (November 2024-March 2025), site visits and on-farm observations (October 2025), and reading academic and grey literature and company websites. Results show that a gap exists between regional experts and agri-food businesses in regards to expectations of digital technologies. Regional experts have high expectations of the transformative power of digitalisation for sustainability, especially in the case of digital platforms and block chains. Nevertheless, in practice, digital technologies are mainly used for convenience, and digitalisation is often not the main path in sustainable transformation of agri-food business models. Small local agri-food businesses in Occtianie do not seem to be at the forefront of digital development. Digital platforms and other communication technologies that connect different actors together do not include all actors, but only the ones who are following the right requirements and certifications. To conclude, digital technologies make things more efficient in agri-food businesses, but that does not mean that they lead to a sustainable transformation. In practice, human interactions between actors remain essential, while digital tools are useful but they are restricted by the structure in which agri- food businesses have to operate.
ano.nymous@ccsd.cnrs.fr.invalid (Laura Eline Slot) 07 May 2026
https://hal.inrae.fr/hal-05616230v1
-
[hal-05616254] Can digital tools improve farmers' commercial conditions? Analysis based on sales strategie
The integration of digital technology into food supply chains came with high expectations, particularly for improving farmers' sales conditions. In theory, digital tools are expected to provide access to commercial information and enable more direct relationships with consumers, thereby mitigating two issues that threaten farm profitability and even survival: transaction costs (Transaction Cost Theory) and commercial uncertainty (Resource Dependency Theory). However, solid empirical evidence is currently scarce. Therefore, the goal of this thesis is to assess the actual impact of digital technologies on the costs and uncertainty burdening farmers. To achieve this, we focus on soft, everyday technologies used by almost all farmers: social media, software, and online sales platforms. We conducted representative quantitative surveys of French market gardeners, who are highly concerned about costs and uncertainty, as well as semi-structured interviews to gain deeper insight into their strategies. We also surveyed managers of online sales platforms to better understand how they interact with farmers. The resulting data are analysed using a range of methods, including propensity score matching and Qualitative Comparative Analysis (QCA). The results demonstrate that digital tools can reduce both transaction costs and market uncertainty. Their impact varies depending on the farmer's sales strategy, specifically whether they sell directly or indirectly to consumers. Transaction costs are reduced when farmers use digital tools for marketing and communication. For those selling directly, this improves sales conditions; for those selling indirectly, it enhances the customer relationship. However, information-sharing tools benefit direct sellers but increase transaction costs for indirect sellers because they create greater information asymmetry. Regarding platforms, we observed that some use strategies that limit interaction between farmers and customers, thereby driving up transaction costs. Market uncertainty can be reduced by using digital tools to organise sales, especially for order-taking, both for direct and indirect sales. Platforms offer two effective strategies for reducing uncertainty, based on how they operate. Platforms that include farmers in their governance and encourage exchange with consumers provide a reliable, long-term outlet to stabilise farmers' sales. Platforms requiring low farmer involvement can serve as an 'adjustment' outlet for dealing with surpluses or issues with farmers' regular sales channels. This research shows that digital tools are one lever for farmers to improve their sales conditions. However, it is not a one-size-fits-all solution. It is therefore crucial to clarify these effects so that farmers can select the tools best suited to their own strategy and objectives.
ano.nymous@ccsd.cnrs.fr.invalid (Romane Guillot-Pelliet) 07 May 2026
https://hal.inrae.fr/hal-05616254v1
-
[hal-05616213] Indicating Interdisciplinarity as Situated Practice: Lessons from the Interdisciplinarity Observatory at DigitAg
This presentation draws on the experience of running the Interdisciplinarity Observatory within the French Convergence Institute DigitAg, a national program dedicated to digital agriculture. As part of a broader reflection in organizational science and science and technology studies on how knowledge integration is governed and made visible, the paper discusses how interdisciplinarity can be indicated—rather than evaluated—as a situated practice. The aim is to share insights from observing, documenting, and participating in organizational attempts to make interdisciplinarity legible inside a large-scale research program. From this embedded perspective, interdisciplinarity appears through the interplay of political, organizational, and practical dynamics. Politically, it is framed as a strategic solution to societal challenges such as agroecological transition and digitalization. Yet policy narratives remain deliberately open, leaving research organizations to translate abstract ideals into measurable signs of achievement. Organizationally, this translation work involves developing commensuration devices and promotional initiatives that materialize what funders would call scientific and interdisciplinary performance. In DigitAg, for instance, matrix organization or co-supervision became key ways to indicate interdisciplinarity. These indicators do not capture practice directly but perform selective visibility, privileging what can be counted and communicated upward. At the level of practice, interdisciplinarity surfaces in researchers' day-to-day efforts to align heterogeneous methods, vocabularies, and timelines. Observing such interactions reveals that interdisciplinarity is sustained less by formal structures than by repeated acts of negotiation, translation, and repair. Taken together, these observations show that indicating interdisciplinarity differs fundamentally from evaluating it. Indication involves tracing how visibility is constructed, by whom, and to what effect, rather than measuring compliance against predefined criteria. It is an epistemic and organizational practice that can reveal where tensions, learning opportunities, and potential leverage points lie within interdisciplinary infrastructures. The presentation contributes to ongoing debates on research evaluation and governance by treating observation itself as a reflexive intervention. It invites further discussion on how to cultivate institutional conditions that allow interdisciplinarity to be seen, questioned, and supported as an evolving, situated practice rather than a fixed policy objective or an institutional design.
ano.nymous@ccsd.cnrs.fr.invalid (Jongheon Kim) 07 May 2026
https://hal.inrae.fr/hal-05616213v1
-
[hal-05616202] Media Representations of Digital Tools for Agroecology: Insights from the French Agricultural Press
Digitalisation has become prominent in French agricultural policy and innovation narratives, as new expectations arise for more sustainable production models. Professional agricultural media play a critical role in shaping how farmers perceive both these transitions, acting simultaneously as intermediaries of innovation diffusion and mirrors of sectoral identities. Yet their influence on the representation of digital tools in relation to agroecology remains understudied. This communication addresses the theme “Impacts of digital technology on agriculture” by examining how professional agricultural media construct and mediate the effects attributed to digital tools in connection with agroecology practices. The initial focus on the vegetable sector offers a particularly relevant entry point. Due to its high exposure to environmental challenges, its labour- intensive nature, its frequent under-equipment, and strong expectations for healthy and sustainable food, vegetable production has become a showcase for both agroecological ambitions and digital experimentation. Across a diversified sample of national outlets, we examine which visions of agroecology and digitalisation are deemed legitimate, which actors are granted authority, and which promises and risks are framed as central. The research builds upon initial insights obtained in a master's study focused on a single major outlet, which highlighted a strong valorization of efficiency-driven arguments and the increasing integration of environmental concerns within narratives on digital farming. Expanding to the broader national media landscape allows for a more comprehensive comparison of editorial and discursive framings. Methodologically, the work combines lexical statistics and qualitative discourse analysis. To build the corpus, we employ a structured keyword-based search strategy articulated around three conceptual groups capturing discourses at the intersection of agricultural digitalisation, ecological transition and sector-specific issues:(1) terms referring to digital technologies;(2) terms relating to agroecology and ecological transition, acknowledging ongoing definitional tensions between institutional eco-efficiency framings and diversification-based approaches;(3) production-related descriptors enabling adequate coverage of the market gardening subsector. This keyword structure ensures both inclusion of relevant articles and consistency across outlets. The corpus will cover approximately fifteen years of publications to detect the emergence and evolution of discursive patterns. The corpus will be analyzed using quantitative text-mining (lexical statistics, co-occurrence networks, temporal trends) combined with qualitative examination of framing devices, actor representation and normative orientations. Attention will also be given to the role of media as innovation brokers, in line with current research recognizing professional journalism's strategic involvement in both mediating and shaping the digital futures of French agriculture.
ano.nymous@ccsd.cnrs.fr.invalid (Romain Blancaneaux) 07 May 2026
https://hal.inrae.fr/hal-05616202v1
-
[hal-05616242] Integration of blockchain into agri-Food supply chains: an analysis through the Lens of the “inscription techniques” used by blockchain service providers
Health, nutritional, and environmental crises, reflecting, since the late 20th century, the tensions at work in so-called “concerned” markets (Geiger, Harrison, Kjellberg et al., 2014), have highlighted the growing exposure of agri-food supply chains (SC) to disruptions and uncertainties (Simangunsong, Hendry and Stevenson, 2012). In response, actors within food chains are called upon to improve the traceability of their products (Bosona and Gebresenbet, 2013). Blockchain technology (Nakamoto, 2008), defined “as a digital, decentralized and distributed ledger in which transactions are logged and added in chronological order with the goal of creating permanent and tamper-proof records” (Treiblmaier, 2018, p. 547), has been presented as an asset for improving the performance and sustainability of agri-food SCs (Difrancesco, Meena and Kumar, 2023; Giganti, Borrello, Falcone et al., 2024). Empirical studies have established both the applications of blockchain technology (Cozzio, Viglia, Lemarie et al., 2023; Li, Lee and Gharehgozli, 2023; Vern, Panghal, Mor et al., 2025) and adoption barriers, whether related to psychological factors (Komulainen and Nätti, 2023), trust (Zheng et al., 2017), competitive contexts (Jovanovic et al., 2022), and so on. Above all, the inventory of barriers and the limits of existing analyses (Vu et al., 2023) point to a lack of understanding regarding the specific obstacles encountered during the implementation of blockchain technology within food supply chains. In order to contribute to this needed enrichment of the literature, we will adopt a pragmatic sociology follow-up of the inalienable rights of materiality, considering the articulation between blockchain and supply chain as a networked configuration of both human and non-human actors, whose stability will determine the adoption of an innovation (Callon, 1986). Our approach is structured around two components: (1) semi-structured interviews conducted with 4 blockchain service providers and 6 companies using these services (cooperatives, distributors, etc.), and (2) semantic analyses using Tropes software of videos and LinkedIn posts published by the interviewed service providers. We will more specifically examine inscription techniques as objects that are “mobile, immutable, presentable, readable and combinable” (Latour, 1987, p. 94), which, as “faithful and disciplined allies” (Latour, 1987, p. 84), become resources mobilized with the aim of persuading.
ano.nymous@ccsd.cnrs.fr.invalid (Jan Smolinski) 07 May 2026
https://hal.inrae.fr/hal-05616242v1
-
[hal-05616185] Behind the Screens: Farmers' Lived Experiences of the Social Costs and Benefits of Digitalisation
Digitalisation is changing farming in ways that go well beyond technology or economics. It is reshaping how farmers work, connect with others, and see themselves within their communities. While research highlights gains in efficiency and sustainability (Klerkx et al., 2019), the social consequences of this shift remain less understood. Studies show that farmers do not share a single vision of what improvement means, reflecting diverse understandings of value and impact (Herrera et al., 2016). In a similar vein, research on digitalisation points to uneven access to knowledge, support, and agency across rural contexts, shaping how farmers adopt and benefit from new tools (Townsend & Noble, 2022). Iliopoulos et al. (2025) found that farmers perceive the costs and benefits of digitalisation in diverse and sometimes contradictory ways. Yet most studies still focus on broad trends, overlooking how these changes affect everyday life, wellbeing, and belonging (Rose & Chilvers, 2018). Understanding these experiences is key to ensuring that the digital transition is fair, inclusive, and centred on farmers. This study examines how digital tools,Input Management Technologies (IMT) and Information and Communication Technologies (ICT), influence farmers' wellbeing, work organisation, and social relations. An interdisciplinary workshop with researchers in sociology, economics, ergonomics, management sciences, and agricultural engineering identified key social dimensions potentially affected by digitalisation. Using this framework, we conducted semi-structured interviews with 24 farmers across 16 European Living Labs within the CODECS project, exploring their lived experiences of both the costs and benefits of digital tools. Findings reveal a dual reality. Farmers described greater comfort, less physical strain, and better decision-making. IMT supported sustainable practices and environmental care, often linked to higher job satisfaction and a sense of purpose. ICT enhanced networks and communities, promoting collaboration, knowledge exchange, and visibility in local markets. Farmers also said digitalisation improved farming's image, helping attract younger generations and modernise the profession. However, these gains carried social costs. Farmers reported the mental toll of data overload, financial pressures, and a constant need to update skills. Some noted work tensions between generations, reduced autonomy as technology replaced traditional know-how, and growing isolation as digital engagement replaced face-to-face contact. Overall, digitalisation emerges not only as a technical change but as a social transition reshaping wellbeing, identity, and community life (Bellon-Maurel et al., 2022; Rose & Chilvers, 2018). To make this transformation sustainable, policies and innovations must support farmers' agency, inclusion, and emotional wellbeing at the heart of Europe's digital farming future.
ano.nymous@ccsd.cnrs.fr.invalid (Mauro Florez) 07 May 2026
https://hal.inrae.fr/hal-05616185v1
-
[hal-05561219] Appréhender l'agriculture numérique. 10 ans de recherche interdisciplinaire au sein de l’Institut #DigitAg [Introduction générale]
[...]
ano.nymous@ccsd.cnrs.fr.invalid (Karine Gauche) 20 Mar 2026
https://hal.science/hal-05561219v1
-
[hal-05562802] L’émergence de nouveaux modèles de numérique en agriculture : vers un numérique agricole accessible et responsable
Chapitre 13 de la partie 2 : Processus, pratiques et instruments : comprendre les multiples dimensions de la digitalisation agricole
ano.nymous@ccsd.cnrs.fr.invalid (Véronique Bellon-Maurel) 23 Mar 2026
https://hal.science/hal-05562802v1
-
[hal-05585651] Acquérir des données de terrain de qualité en améliorant les capteurs et leur mise en oeuvre
[...]
ano.nymous@ccsd.cnrs.fr.invalid (Véronique Bellon-Maurel) 09 Apr 2026
https://hal.science/hal-05585651v1
-
[hal-05583908] Regard sur le challenge Sud de #DigitAg : le numérique agricole dans les pays du Sud
Partie 4 : Retour réflexif et perspectives de l’Institut Convergences. Chapitre 11
ano.nymous@ccsd.cnrs.fr.invalid (Nicolas Paget) 08 Apr 2026
https://hal.inrae.fr/hal-05583908v1
-
[hal-05601208] Observer, prédire et expliquer la sécurité alimentaire
[...]
ano.nymous@ccsd.cnrs.fr.invalid (Agnès Bégué) 24 Apr 2026
https://hal.science/hal-05601208v1
-
[hal-05562761] Outils numériques comme instruments d’intermédiation : des attentes multiples aux réalités contrastées
Chapitre 7 de la partie 2 : Processus, pratiques et instruments : comprendre les multiples dimensions de la digitalisation agricole
ano.nymous@ccsd.cnrs.fr.invalid (Sophie Mignon) 23 Mar 2026
https://hal.science/hal-05562761v1
-
[hal-05561313] L’architecture de l’interdisciplinarité : une expérience d’institutionnalisation d’une communauté scientifique interdisciplinaire à #DigitAg
[...]
ano.nymous@ccsd.cnrs.fr.invalid (Karine Gauche) 20 Mar 2026
https://hal.science/hal-05561313v1
-
[hal-05583828] Systèmes d’information, modélisation sémantique et gestion des connaissances pour l’agriculture
[...]
ano.nymous@ccsd.cnrs.fr.invalid (Sandrine Auzoux) 08 Apr 2026
https://hal.inrae.fr/hal-05583828v1
-
[hal-05561297] La digitalisation agricole à l’épreuve des transitions : processus, acteurs et dynamiques systémiques
Chapitre 1 de la partie 2 : Processus, pratiques et instruments : comprendre les multiples dimensions de la digitalisation agricole
ano.nymous@ccsd.cnrs.fr.invalid (Karine Gauche) 20 Mar 2026
https://hal.science/hal-05561297v1
-
[hal-05577964] Chapitre 4 - Aider à la décision avec le numérique via la modélisation, la simulation et l’intégration de données
Cela a été montré dans les trois premiers chapitres de cette partie, les données, qui se diversifient et se massifient, sont un nouvel intrant, transformateur de l'agriculture. Pour que ces données constituent un apport valorisable sur le plan agronomique et pour la transition écologique, il est nécessaire de les structurer et de les transformer en informations mobilisables pour l'action. Le chapitre précédent concernait principalement les systèmes d'information pour les données, les connaissances et le support décisionnel. Le présent chapitre aborde résolument la modélisation, que l'on énonce au singulier en tant qu'activité, mais qui recouvre un ensemble très varié de méthodes et d'objectifs. Ces modélisations, donc, sont élaborées pour répondre aux différentes fonctions cognitives, décisionnelles et opérationnelles qui relèvent de l'aide à la décision.<p>Ce chapitre abordera les différentes étapes et les différents acteurs de la décision, depuis le diagnostic et le conseil agricole jusqu'à la mise en oeuvre au champ. Le choix a été fait ici de regrouper les questions scientifiques abordées dans #DigitAg autour de l'usage de la modélisation, de la simulation et de l'optimisation pour l'aide à la décision en différentes parties portant sur (1) le suivi de l'état des cultures et des agroécosystèmes en général, (2) la prédiction et la gestion des risques, (3) l'optimisation des systèmes au stade de la conception comme au stade de la conduite de production, et enfin (4) l'évaluation du fonctionnement de ces systèmes.</p>
ano.nymous@ccsd.cnrs.fr.invalid (Frédérick Garcia) 02 Apr 2026
https://hal.inrae.fr/hal-05577964v1
-
[hal-05585811] Systèmes d'élevage durables
[...]
ano.nymous@ccsd.cnrs.fr.invalid (Charlotte Gaillard) 09 Apr 2026
https://hal.science/hal-05585811v1
-
[hal-05562537] Appréhender l'agriculture numérique. 10 ans de recherche interdisciplinaire au sein de l’Institut #DigitAg [Préface]
[...]
ano.nymous@ccsd.cnrs.fr.invalid (Vincent Martin) 23 Mar 2026
https://hal.science/hal-05562537v1
-
[hal-05561320] Appréhender l'agriculture numérique. 10 ans de recherche interdisciplinaire au sein de l’Institut #DigitAg [Conclusion]
[...]
ano.nymous@ccsd.cnrs.fr.invalid (Karine Gauche) 20 Mar 2026
https://hal.science/hal-05561320v1
-
[hal-05601207] Chapitre 2. Traitement des données massives et multisources
Ce chapitre présente le traitement des données massives et multisources en agriculture numérique. Il explique comment les données issues de capteurs, satellites, drones, images, séries temporelles et sources ouvertes sont analysées par apprentissage automatique et apprentissage profond. Les applications concernent le suivi des cultures, la cartographie agricole, l’élevage de précision, les maladies végétales et la sécurité alimentaire.
ano.nymous@ccsd.cnrs.fr.invalid (Dino Ienco) 24 Apr 2026
https://hal.science/hal-05601207v1
-
[hal-05561157] Appréhender l'agriculture numérique. 10 ans de recherche interdisciplinaire au sein de l’Institut #DigitAg
Pendant 10 ans (2017-2026), l’Institut Convergences #DigitAg a mené des recherches sur l’agriculture numérique, fruit de la transformation de l’agriculture par la digitalisation, et a réuni à cette occasion une communauté interdisciplinaire de chercheurs issus de divers organismes de recherche et d’enseignement supérieur français (INRAE, Cirad, Institut Agro, université de Montpellier, Inria, Acta, AgroParisTech). Cet ouvrage décrit les outils et méthodes mis au point, couvrant toute la chaîne de la donnée : des capteurs aux images satellitaires, en passant par le traitement des données, l’intelligence artificielle ou les systèmes d’aide à la décision… Leurs usages pour le phénotypage, les élevages durables et la sécurité alimentaire y sont détaillés. L’ouvrage explore aussi les impacts de la digitalisation sur les pratiques des agriculteurs, les écosystèmes et les modèles agricoles en France et dans les Suds. L’interdisciplinarité est ici au cœur de la réflexion : comment cette communauté a-t-elle réussi à faire converger des expertises variées ? Quel a été l’apport de cette approche pour repenser l’agriculture de demain ? Autant de questions auxquelles cet ouvrage apporte des réponses concrètes, nourries par une décennie de recherche collaborative. Cet ouvrage s’adresse aux étudiants et aux chercheurs, et à toute personne intéressée par les avancées méthodologiques des sciences et technologies du numérique en agriculture.
ano.nymous@ccsd.cnrs.fr.invalid (Véronique Bellon-Maurel) 20 Mar 2026
https://hal.science/hal-05561157v1
-
[hal-05562659] Usages et pratiques du numérique par les agriculteurs : quels enseignements pour les liens entre digitalisation et écologisation de l’agriculture ?
Chapitre 6 de la partie 2 : Processus, pratiques et instruments : comprendre les multiples dimensions de la digitalisation agricole
ano.nymous@ccsd.cnrs.fr.invalid (Chloé Alexandre) 23 Mar 2026
https://hal.science/hal-05562659v1
-
[hal-05548661] Wheat growth model capturing growth-defense trade-off
Improving crop productivity in agroecological systems subject to multiple abiotic and biotic stresses requires a comprehensive integration of physiological mechanisms into plant growth models. In this article, we analyze the structure, components and limitations of current process-based models (PBMs) and Functional-Structural Plant Models (FSPMs) used to simulate wheat (Triticum spp.) growth. Although these models are well adapted to represent light interception, carbon assimilation and biomass allocation, they remain mostly oriented toward yield or growth prediction and usually neglect biotic and abiotic stress factors, which are crucial under agricultural conditions. In this article, we review the main physiological concepts of growth, including photosynthesis, nitrogen uptake, source-sink relationships and respiration costs, with an emphasis on resource allocation trade-offs. These trade-offs, particularly between growth and defense, are rarely explicitly integrated into current modeling frameworks, despite their decisive role on yield and growth under stresses. To fill these gaps, we propose a conceptual model that explicitly integrates physiological trade-offs between growth and defense, as well as hormonal signaling networks. By adopting a more explanatory and integrative approach, this work aims to improve the ability of models to facilitate the transition towards a stronger integration of agroecological principles.
ano.nymous@ccsd.cnrs.fr.invalid (Pauline Dusfour-Castan) 26 Mar 2026
https://institut-agro-rennes-angers.hal.science/hal-05548661v2
-
[hal-05560736] A Hybrid LiDAR–PROSAIL Approach for LAI Estimation in Vineyards
Leaf Area Index (LAI) is a key indicator of canopy development and productivity in sustainable viticulture. However, accurately estimating LAI in vineyards remains challenging due to complex canopy architectures, which often lead radiative transfer models (RTMs), such as PROSAIL, to underestimate LAI. This study proposes a novel hybrid approach tailored to vertically trained vineyards that uses a PROSAIL-based spectral LAI baseline with LiDAR-derived structural information through a data-driven correction strategy. The study was conducted during the 2023 season in a Mediterranean vineyard comprising three grapevine varieties. A structurally weighted index (LLI_w) was derived using multiple linear regression to assign global, phenology- and variety-specific weights accounting for the relative importance of structural parameters across growth stages. The PROSAIL-derived LAI and LLI_w were subsequently integrated using Support Vector Machine regression. The hybrid model showed strong agreement with ground-measured LAI (R² = 0.88) and clearly outperformed PROSAIL alone (R² = 0.42). The hybrid model captured spatio-temporal and varietal differences in canopy development, correcting seasonal underestimations and revealing structural dynamics not detected by spectral models alone. By integrating spectral and structural data, this hybrid approach provides physiologically informed, temporally consistent LAI estimates, supporting adaptive, variety-specific vineyard monitoring.
ano.nymous@ccsd.cnrs.fr.invalid (Leire Sandonís-Pozo) 20 Mar 2026
https://hal.inrae.fr/hal-05560736v1
-
[hal-05518104] Les radios locales peuvent-elles contribuer à la surveillance épidémiologique ?
[...]
ano.nymous@ccsd.cnrs.fr.invalid (Rémy Decoupes) 19 Feb 2026
https://hal.science/hal-05518104v1
-
[hal-05524078] Robustness of high‐throughput prediction of leaf ecophysiological traits using near infrared spectroscopy and poro‐fluorometry
Abstract: Water scarcity is a major threat to crop production and quality. Improving drought tolerance through variety selection requires a deeper understanding of plant ecophysiological responses, but large-scale phenotyping remains a bottleneck. This study assessed the potential of high-throughput tools (spectroscopy and poro-fluorometry) to predict leaf morphological and ecophysiological traits in a grapevine diversity panel grown in pots under well-watered outdoor conditions and under three contrasting soil water treatments in a greenhouse. We found a certain complementarity between measuring devices. Spectrometers could accurately predict leaf mass per area, water content, and water quantity (R2 &gt; 0.58), while the poro-fluorometer was efficient for predicting net CO2 assimilation (R2 &gt; 0.72), regardless of the water treatment. The prediction of leaf mass per area using spectrometers appeared to be quite robust across both outdoor and greenhouse experiments, while the prediction of water use efficiency was dependent on the water treatment, with much better predictions under moderate (R2 &gt; 0.73) than severe water deficit. Calibrated models were then applied to the full diversity panel using only high-throughput measurements to estimate trait values and their broad-sense heritability. Leaf mass per area, also measured directly, showed similar heritability whether based on observed or predicted data. Heritability estimates for predicted traits reached up to 0.5. Overall, our findings support the use of spectroscopy and poro-fluorometry as reliable, nondestructive tools for high-throughput phenotyping, enabling genetic studies on drought-related traits in grapevine. Plain Language Summary: Drought is a major challenge for crop production. To breed grapevines that can better tolerate dry conditions, it is crucial to evaluate how plants respond to water stress using quick phenotyping tools. In this study, we tested two fast, nondestructive tools, near-infrared spectroscopy and poro-fluorometry, on grapevines grown with different water levels. Spectroscopy was accurate for evaluating leaf thickness and water content, while poro-fluorometry was better at predicting photosynthesis. These tools were effective even across different growing environments. The results demonstrate that these methods can aid in estimating genetic variability in drought-related traits, facilitating the selection and improvement of drought-tolerant grapevines.
ano.nymous@ccsd.cnrs.fr.invalid (Eva Coindre) 23 Feb 2026
https://hal.inrae.fr/hal-05524078v1
-
[hal-05525643] Quelles configurations de travail en élevage laitier breton pour la transition et la collaboration agroécologiques ?
Des formes de collaboration inter‑exploitations présentes depuis l’après‑guerre se recomposent aujourd’hui sous l’effet du développement de pratiques agroécologiques, afin de coproduire et de cogérer les ressources stratégiques nécessaires à cette transition. Les coopératives d’utilisation de matériel agricole (Cuma), ainsi que les groupes d’échanges et de formation collective entre pairs, sont particulièrement mobilisés en élevage laitier, notamment dans les exploitations herbagères économes en intrants. Toutefois, l’élevage laitier demeure le secteur agricole où la charge de travail par actif est la plus élevée, ce qui interroge la capacité des éleveurs à dégager du temps pour participer à ces activités collaboratives. Dans ce contexte, une question centrale émerge : Existe‑t‑il des configurations de travail favorables à la fois à l’orientation herbagère des élevages bovins laitiers et à l’implication dans les collaborations facilitant la gestion de systèmes herbagers ? Pour y répondre, une recherche exploratoire interdisciplinaire associant zootechnie et sociologie a été menée en 2022 en Bretagne, reposant sur des entretiens avec une trentaine d’éleveurs laitiers. Après avoir précisé l’approche et la méthode d’enquête, ce chapitre analyse d’abord le rôle de l’échelle de l’exploitation — entendue comme le ratio entre dimension économique et nombre d’actifs — dans l’orientation herbagère. Il examine ensuite les autres facteurs conditionnant l’engagement dans les collaborations inter‑exploitations. Enfin, il propose des pistes de recherche complémentaires ainsi que des orientations pour l’action publique et le développement.
ano.nymous@ccsd.cnrs.fr.invalid (Anne-Lise Jacquot) 24 Feb 2026
https://hal.science/hal-05525643v1
-
[hal-05601679] Mapping the adoption of digital technologies in agriculture and its implications for SDG 2: A bibliometric review
Agriculture and food systems are central to achieving the Sustainable Development Goals (SDGs), yet progress remains insufficient. Digital technologies are often presented as part of the solution; however, research in Information Systems and digital agriculture largely prioritises productivity and efficiency outcomes. Adoption is frequently treated as a discrete decision rather than as a multi-stage process, with limited attention to how sustainability dimensions are addressed across contexts. This fragmentation constrains understanding of how digital adoption relates to SDG objectives. This study addresses this gap through a bibliometric and structured coding review of the scientific literature on digital technology adoption in agriculture. It examines how adoption is framed across stages, influencing factors, and heterogeneous technology clusters, including data-driven production, digital market connectivity, digital knowledge services, and digital government platforms, alongside sustainability dimensions. Findings show that adoption is mainly conceptualised as a readiness-based decision moment, with limited differentiation between pre-adoption, adoption, and post-adoption stages. Individual determinants dominate, while organisational and institutional dynamics remain under- theorised. Sustainability is mainly framed through productivity and resource-efficiency, with less attention to governance and social dimensions. By clarifying these patterns, the study reframes adoption as a multi-stage, multi-level process and supports more sustainability-oriented research on digital transformation in agriculture.
ano.nymous@ccsd.cnrs.fr.invalid (Mauro Florez) 24 Apr 2026
https://hal.inrae.fr/hal-05601679v1
-
[hal-04506794] Optimal structures of crop irrigation strategies with state constraints
We investigate an optimal control problem of crop irrigation with non-autonomous and non-smooth dynamics. Depending on contexts and objectives, several formulations associated to different constraints and criteria can be derived. Our work aims at providing optimal feedback solutions for these problems by deriving and analyzing the optimality necessary conditions. To this end, we assemble the different problems into a common formulation, and we carry out a dedicated way of handling state constraints. We show that all optimal irrigation strategies belong to a family of simple parameterized time-varying feedback controls, independently of the context and objective, and suitable for computational purposes.
ano.nymous@ccsd.cnrs.fr.invalid (Ruben Chenevat) 28 Mar 2025
https://hal.science/hal-04506794v2
-
[hal-05521017] How data and the digital technologies are shaping the data economy for agrifood systems
The agri-food sector faces urgent challenges such as food security, environmental sustainability, and equitable access to resources. The Data Economy for AgriFood Systems (DE4AFS) offers a transformative solution, leveraging data and digital technologies to address these issues while driving innovation and efficiency. By fostering interconnected ecosystems through federated data spaces, DE4AFS enables secure and trustworthy data sharing, promoting collaboration among stakeholders. Artificial intelligence (AI) further amplifies its potential by generating actionable insights, enhancing decision-making, and supporting sustainable practices like smart agriculture and supply chain optimization. However, challenges such as the digital divide, governance complexities, and environmental impacts remain critical barriers. This chapter outlines the technical and organizational dimensions of the DE4AFS, emphasizing the role of AI and federated data spaces in creating resilient, equitable, and sustainable agri-food systems. It highlights pathways for addressing these challenges and provides a roadmap for leveraging the digital transformation to achieve global sustainability goals in the agri-food sector.
ano.nymous@ccsd.cnrs.fr.invalid (Sjaak Wolfert) 20 Feb 2026
https://hal.inrae.fr/hal-05521017v1
-
[hal-05611381] Convergence Of Research in Digital Agriculture Leading Labs (CORDiALL) - Book of abstracts
Book of abstracts of the CORDiALL conference. Digital agriculture stands for the digitalisation of agriculture. It has been the focus of a growing number of research and innovation projects over the last 10 years. In France, #DigitAg is the first structuring large research project ‘about and for’ digital agriculture. Other research projects &amp; programmes have followed, at regional level, such as OccitANum (Occitanie Digital AgroEcology), at national level such as the ‘Agroecology &amp; Digital Tech’ program and its flagship projects (CoEditAg, Wait4, Pl@ntAgroEco, LINDDA), and at European level such as the CODECS and QUANTIFARM (tbc) projects. #DigitAg is also connected to other vibrant communities devoted to research in Digital Agriculture emerging worldwide, in Africa or South America. The year 2026 will mark the end of several of these projects, eg #DigitAg, CODECS, QUANTIFARM but also the ramp-up of newly-developed projects on digital agriculture worldwide, especially in Africa - as part of the TSARA initiative - and South America, in relationship with CGIAR and with Brasilian ecosystem. This pivotal moment is the opportunity to organise an International Conference, in order to build convergences between continents, between disciplines, between pioneering and newcomers in digital agriculture research.
ano.nymous@ccsd.cnrs.fr.invalid (Véronique Bellon-Maurel) 05 May 2026
https://hal.inrae.fr/hal-05611381v1
-
[hal-05592002] Chapitre 4. Aider à la décision avec le numérique via la modélisation, la simulation et l'intégration de données
[...]
ano.nymous@ccsd.cnrs.fr.invalid (Frédérick Garcia) 15 Apr 2026
https://hal.science/hal-05592002v1
-
[lirmm-05596142] Combining Symbolic and Generative AI to Explore Knowledge Base and Control Cabbage Pests in West-Africa
Suggesting agroecological solutions based on local resources for crop protection to farmers in the South is a challenge. To this end, the Knomana initiative gathered knowledge on plant-based solutions derived from local flora to replace synthetic pesticides and antimicrobial products. It relies on data collected in scientific literature which describes protection systems, mainly composed of a plant which controls a pest on a crop, and includes additional information such as the location, the used plant part, and the plant extract type. To facilitate dataset exploration and produce recommendations, knowledge is formally represented and then analyzed using a symbolic AI method (Formal Concept Analysis, i.e. FCA) and its relational extension (RCA). FCA and RCA enable the extraction of implication rules (the Duquenne-Guigues Basis of Implications) that reveal relevant dependencies and generalizations in the data. As a symbolic method, FCA and RCA produce robust and reliable results, but that are not directly intelligible by domain experts. This is where Generative AI (LLM) could help, by presenting the results in plain language that is expected to be understandable and provide an adequate level of detail for domain experts. As a use case, we focus on the protection of Brassicaceae in West Africa, a major crop of interest affected by many pests causing significant losses. In addition to data from Knomana, the dataset also indicates, for each crop and plant, whether it is used for human care, consumed, cultivated, or spontaneous. After extracting the rules using RCA, rules are reformulated using three LLMs (ChatGPT, Claude, and Gemini) and submitted to domain experts (entomologists). They evaluate to what extent this reformulation is complete, correct (i.e., free of misinterpretation), non-redundant, adds information coming from LLM general knowledge which is relevant, and provides useful insights. These results are the first step in developing AI-based recommendation systems for the protection of agricultural crops in the South.
ano.nymous@ccsd.cnrs.fr.invalid (Alain Gutierrez) 18 Apr 2026
https://hal-lirmm.ccsd.cnrs.fr/lirmm-05596142v1
-
[tel-05489364] Faire-commun par les modèles : une géographie située des futurs possibles de la forêt en Zone Sylvo-Pastorale du Sénégal
Cette thèse s’inscrit dans le cadre des efforts de lutte contre la désertification au Sahel, notamment à travers le projet Grande Muraille Verte. Elle analyse la gouvernance des ressources sylvo-pastorales en combinant les notions de Communs et la modélisation d’accompagnement (ComMod), afin de développer une géographie située des futurs possibles. En appui au programme Dundi Ferlo, qui rassemble un consortium composé d’ONG spécialisées dans le pastoralisme et la reforestation, ainsi que des instituts de recherche, des collectivités locales, des autorités coutumières (chefs de village) et les services forestiers sénégalais, cette recherche répond à la question suivante : comment l’articulation des Communs et la modélisation d’accompagnement (ComMod) permet-elle de concevoir, tester et négocier des dispositifs de reforestation compatibles avec les dynamiques du pastoralisme mobile ? La démarche vise à réduire les inégalités de pouvoir entre les participants. Elle propose de "faire-commun par les modèles" en accompagnant la gouvernance des ressources sylvo-pastorales dans le Ferlo (Sénégal), un territoire marqué par des reconfigurations foncières et institutionnelles liées au pastoralisme mobile. Le cadre théorique associe la sociologie de la traduction à l’approche par les communs, en envisageant les Communs comme des « réseaux de traductions normatives ». La méthodologie repose sur ComMod conçue comme une recherche-action participative, visant à co-produire des connaissances utiles dans un contexte précis avec une perspective de géographie située des futurs. Au cœur du dispositif, le modèle à base d’agents DundiModel a été co-construit lors d’ateliers sur le terrain. Ce modèle agit comme un objet-frontière permettant d’expliciter hypothèses et controverses. Il formalise et explore les conséquences d'hypothèses issues de savoirs divers. Deux contributions principales structurent la démarche : (i) une ingénierie stigmergique qui cumule et rend traçables les empreintes du processus (schémas conceptuels, versions du code, paramètres, comptes rendus) et facilite la révision informée des objectifs et des règles, (ii) une exploration massive et organisée du modèle (analyses de sensibilité, profils de calibration, Pattern Space Exploration) qui délimite l’espace des futurs possibles et renforce la plausibilité des scénarios. Les résultats montrent que la variabilité climatique et la démographie pastorale dominent les trajectoires futures. Les pratiques pastorales ont un impact limité sur la déforestation mais une influence plus forte sur la régénération des arbres ligneux. Pour des conditions climatiques et démographiques données, les stratégies pastorales (calendrier, formations) peuvent favoriser une densité d’arbres allant jusqu’à 110 arbres/ha dans un scénario favorable, comparé à 80 arbres/ha dans un scénario défavorable (contre 70 arbres/ha aujourd’hui). En revanche, ces stratégies coûtent au bien-être pastoral (formation des bergers, structure des troupeaux, état corporel des animaux). Dans tous les cas, le modèle montre que les gains forestiers restent coûteux pour l’économie pastorale. Ces résultats invitent à reconsidérer les controverses : plutôt que de stigmatiser le pastoralisme mobile, il s’agit de négocier les mobilités saisonnières, les règles d’accès et les clôtures de reforestation pour permettre à la fois reforestation et bien-être pastoral. Les politiques doivent tenir compte des régimes climatiques et des dynamiques pastorales, penser à l’échelle des systèmes de mobilité et adopter une planification polycentrique, en s’appuyant sur des indicateurs fonctionnels. Le modèle proposé est volontairement limité et ne représente pas toutes les dimensions du système agro-sylvo-pastoral du Ferlo. Il sert surtout d’espace de dialogue informé plutôt que d’outil prescriptif. Méthodologiquement, la thèse propose une grammaire opératoire du "faire-commun par les modèles" : enregistrer et cumuler les traces des apprentissages, de la co-construction du modèle jusqu'à l'exploration des scénarios pour cartographier des futurs situés et négociables des communs sylvo-pastoraux dans le Ferlo.
ano.nymous@ccsd.cnrs.fr.invalid (François Vendel) 02 Feb 2026
https://hal.science/tel-05489364v1
-
[tel-05539177] Petits jeux de données et prédiction en Intelligence Artificielle, vers une meilleure cohabitation : application à la gestion durable de l'enherbement des systèmes agricoles à La Réunion
Dans les parcelles agricoles, les adventices plus communément appelées mauvaises herbes sont considérées comme les bioagresseurs les plus nuisibles. Sans désherbage rapide, elles peuvent entraîner des pertes de rendement considérables. En régions tropicales, leur nuisibilité est particulièrement forte, car le climat chaud et humide favorise leur développement toute l’année.À La Réunion, la canne à sucre occupe plus de la moitié de la surface agricole utile et constitue le premier produit d’exportation. Dans ces parcelles, la diversité d’espèces adventices est importante et peut provoquer des pertes de plusieurs centaines de kilos de canne par hectare et par jour de retard de désherbage, du fait de la compétition en eau, lumière et nutriments. Pouvoir anticiper quelles espèces apparaîtront dans une parcelle et à quel niveau d’enherbement permettrait de cibler plus efficacement les méthodes de désherbages quelles soient chimiques ou alternatives et ainsi, réduire les pertes de rendements et le temps consacré au désherbage. Toutefois, cette prédiction est complexe, car la composition floristique des adventices sur la parcelle dépend d’un ensemble de facteurs environnementaux, géographiques et historiques. L'utilisation du machine learning est une solution où en apprenant à partir de données issues de parcelles, les algorithmes peuvent estimer la flore adventice attendue sur une nouvelle parcelle. Cependant, la collecte de ces données rencontre deux obstacles majeurs, l’hétérogénéité des études existantes et le coût humain et temporel élevé lié à l’expertise nécessaire pour identifier les espèces. Cela entraîne un manque d’observations, des valeurs manquantes et du bruit dans les bases de données, réduisant la précision des modèles. Cette thèse s’intéresse à l'amélioration de ces prédictions avec ce genre de contraintes. Elle explore l’adaptation d’algorithmes à de faibles quantités de données (Few Shot Learning), l’utilisation de techniques de génération de données adaptées à la topographie et au climat de l'île ainsi que l’imputation des valeurs manquantes lorsque leur taux est grand. Ces approches sont appliquées à la prédiction des communautés adventices de la canne à sucre à La Réunion en utilisant des algorithmes dans un cadre multi-labels. Les résultats montrent une amélioration notable des performances des modèles, même si la mise en place d’une démarche standardisée de collecte de données reste indispensable pour obtenir des prédictions plus fiables et robustes.
ano.nymous@ccsd.cnrs.fr.invalid (Frédérick Fabre-Ferber) 06 Mar 2026
https://theses.hal.science/tel-05539177v1
-
[hal-05493267] Family coordination and adaptation to crises: the new role of digital technology in the resilience of pastoral livestock systems in Senegal
Background: Multiple crises are threatening Sahelian pastoral systems. Adopting digital tools could help improve the resilience of pastoralists. However, little is known about the role of digital tools in family-based herd management in times of crisis. Aim: This study examines the impact of digital tools on family coordination and adaptive crisis responses. Methods: The study combines quantitative analyses of a database on digital equipment, covering 1,260 household leaders and members, and around 50 qualitative interviews exploring their phone usage during day-today tasks and when dealing with crises in Senegal. Results: Findings reveal a high rate of mobile phone ownership among household leaders, who typically make herd-related decisions. We observed an increase in use among women and sons involved in production – particularly in systems where market access depends on digital technology ownership. On a daily basis, mobile phones facilitate coordination among family members. They are used to transfer information and money within a spatially dispersed family group. Although intra-family negotiations occur, decision-making remains centralized. During crises, response strategies can be implemented rapidly, by mobilizing information networks and improving access to markets for herd destocking. Conclusions: The study highlights the collective organization of pastoral livestock systems, and underlines the importance of internal dynamics when it comes to resilience and adaptation strategies.
ano.nymous@ccsd.cnrs.fr.invalid (Margot Moniot) 04 Feb 2026
https://hal.inrae.fr/hal-05493267v1
-
[hal-05465253] Coherent Augmentation of Controversial Comments
Self-sufficiency has grown in popularity over the past decades as a response to ongoing societal crises. Considering the immediacy and gravity of the challenges at hand, it raises considerable debate within the public sphere, notably on widely used media such as YouTube. We propose to study this phenomenon by performing a sociological analysis of the technical knowledge around self-sufficiency in YouTube comment sections. Such study requires large corpora of annotated data, which are difficult to produce and typically scarce, especially in languages other than English. To overcome the problem, we implement in this work two independent methods to augment a dataset of labeled French-language YouTube comments. A first method, AugArg, relies on the use of argumentative structure as a controversy marker in the original corpus, while the other method, AugLLM, proposes to prompt an LLM to generate controversial comments. The fine-tuning of a CamemBERT model is performed using both the original and augmentated data, enabling us to compare the efficiency of both methods in representing controversiality in the context of self-sufficiency. Our results indicate that utilizing argumentative structure in the corpus provides a more controversy-representative set of comments in comparison with LLM-oriented methods.
ano.nymous@ccsd.cnrs.fr.invalid (Marina Musse) 19 Jan 2026
https://hal.science/hal-05465253v1
-
[tel-05426346] Transition numérique et mise en œuvre de l’agroécologie : une analyse du rôle des coopératives vitivinicoles et de fruits et légumes
La digitalisation est aujourd’hui perçue comme un levier de transformation durable des systèmes agricoles et alimentaires, et de renforcement des dynamiques agroécologiques. Cependant, elle suscite encore des controverses dans la littérature, notamment sur les questions d’adoption et d’usages effectifs des outils numériques en agriculture, mais aussi sur son lien avec l’agroécologie. Ce travail de recherche analyse le rôle des coopératives dans la transition numérique de l’agriculture dans une perspective de mise en œuvre de l’agroécologie. Il vise d’une part, à comprendre les mécanismes de diffusion des outils numériques en agriculture à travers l’analyse du rôle des coopératives en tant qu’intermédiaires, d’autre part, à examiner le lien entre digitalisation et agroécologie, ainsi que le rôle des coopératives dans l’articulation de ces deux dynamiques afin de parvenir à une transformation durable de l’agriculture. Sur le plan théorique, cette recherche mobilise les travaux sur les intermédiaires d’innovation pour comprendre le rôle des coopératives dans le processus de digitalisation de l’agriculture. Cette approche s’articule avec la littérature sur l’adoption des technologies, combinant le modèle UTAUT et le cadre TOE, afin d’expliquer le comportement d’adoption des agriculteurs vis-à-vis des outils numériques et analyser comment ces outils sont mobilisés dans la mise en œuvre des pratiques agroécologiques à travers leurs usages. D’un point de vue méthodologique, cette thèse étudie le cas des coopératives vitivinicoles et de fruits et légumes localisées principalement dans la région Occitanie. Elle repose sur 44 entretiens individuels semi- directifs, dont 20 entretiens réalisés auprès de dirigeants et d’agents techniques de coopérative, et 24 auprès d’agriculteurs coopérateurs. Les entretiens ont été analysés à partir d’une méthode d’analyse de contenu thématique, complétée par une classification ascendante hiérarchique (CAH) sur de l’ACM. Nos résultats révèlent que les coopératives agissent effectivement comme des intermédiaires de la transition numérique, en assurant des fonctions d’intermédiation de connaissances (ou cognitives) et d’innovation numérique pour favoriser la diffusion et l’adoption des outils numériques auprès des agriculteurs et accompagner les usages. Les analyses des entretiens auprès des agriculteurs permettent d’identifier les freins et les motivations à cette adoption. En outre, ces analyses montrent que le lien entre la digitalisation et l’agroécologie reste à ce stade peu perceptible, car la majorité des offres d’outils numériques est jusqu’ici destinée à la mise en œuvre d’une agroécologie « faible » impliquant des changements mineurs et moins profonds dans les pratiques agricoles. Ce travail met également en relief l’éventuel rôle d’arbitrage des coopératives dans la mise en œuvre des transitions numérique et agroécologique pour une agriculture durable. Cette thèse contribue à mettre en lumière le rôle central des organisations intermédiaires, notamment des coopératives, dans le développement et la mise en œuvre des transitions numérique et agroécologique, tout en relevant l’émergence de nouvelles fonctions et stratégies d’intermédiation associées. Elle invite également à porter un regard réflexif sur les mécanismes de gouvernance coopérative dans l’étude des processus d’innovation au sein des coopératives.
ano.nymous@ccsd.cnrs.fr.invalid (Békanty Ange Kouassi) 19 Dec 2025
https://hal.inrae.fr/tel-05426346v1
-
[hal-05363376] Estimating small-grain cereal plant density at early growth stages using leaf tip density dynamics derived from submillimeter-scale RGB imagery
Plant density is an important variable for management and phenotyping of small-grain cereal crops such as wheat and barley. While many image-based estimation methods exist to replace laborious manual counting, most of them rely on empirical relationships that may not generalize well to different sites, growth stages, species and varieties. In this study, we propose a novel small-grain cereal plant density estimation method that uses leaf tip density dynamics derived from submillimeter-scale images acquired at 45° view zenith angle. This method contained two steps. In the first step, a P2PNet deep learning detection model was trained to estimate leaf tip count in a surface of known area to get the leaf tip density. An occlusion correction method was then applied on this density, leading to an estimation error of about 20% at critical growth stages. In the second step, a wheat leaf dynamic model was used to simulate the evolution of leaf tip density over thermal time as functions of several variables, including mean time of plant emergence, phyllochron and plant density. This model was then inverted using a lookup table approach to estimate plant density from leaf tip density dynamics. The results obtained on three test datasets indicated that two observations performed before the appearance of the second and third leaves could be sufficient to attain a relative plant density estimation error of about 10%. We also discussed that this method should be able to work on other datasets without recalibration, and estimate other variables such as phyllochron at early growth stages. The code will be available at: https://github.com/wdwzytc/WheatPlantDensity.
ano.nymous@ccsd.cnrs.fr.invalid (Tiancheng Yang) 13 Nov 2025
https://hal.inrae.fr/hal-05363376v1
-
[hal-05214356] Comparing the impact of weather forecasts, observation and parameterization uncertainties when predicting water stress in vineyards
We analyze the impact of different uncertainty sources when using of a crop water stress model with weather forecasts in order to make crop water stress predictions. Three uncertainty sources are jointly studied: (i) the observation uncertainty, coming from the use of past weather data not directly located on the field under study, (ii) the input uncertainty, coming from imperfect values of critical crop model parameters, (iii) the forecast uncertainty coming from the difficulty to predict future weather data. These three uncertainty sources are modeled and their impact characterized using a dedicated double-stage sensitivity analysis that allows to tackle two issues: dependencies among input factors and multivariate outputs. The analysis is performed on a large database of 12000 scenarios covering 5 years of weather data observations and forecasts from Integrated Forecast System Ensemble Prediction System (IFS-EPS) associated to 10 different locations in France and various simulated soil conditions. The results show that all uncertainty sources matter, but not in the same ranking depending on lead time. In particular, forecast uncertainty becomes in average the dominant source from the 5th day of forecast, accounting for 64% of the variability at a 7-day lead and 86% at a 15-day lead. Our analysis also highlights the existence of several interpretable dynamic patterns of uncertainty dominance, where each of the two other uncertainty sources (past weather data and cropping system parameterization) can dominate the overall uncertainty for some specific part of the forecasting period, stressing that no uncertainty source can be a priori neglected.
ano.nymous@ccsd.cnrs.fr.invalid (Bachar Tarraf) 19 Aug 2025
https://hal.inrae.fr/hal-05214356v1
-
[hal-05324268] AgroforestAR: A mobile app for visualizing Agroforestry systems in Augmented Reality
Agroforestry is gaining more and more attention from researchers and practitioners in temperate areas, but it remains a vague concept for most of the public. This is because the renewal of interest for agroforestry systems is quite recent, demonstration sites are rare, and trees are still young, and therefore not very visible. The aim of AgroforestAR is to allow visualizing what an agroforestry system could look like on a given piece of land (including in your garden!). It uses the augmented reality capabilities already available in most smartphones, to superimpose, on the view seen by the phone’s or tablet’s camera, trees aligned along a line defined by the user by walking from one side of the piece of land to the other side. The user can then choose the tree species (among 5 available species currently) and size, as well as different distances between tree lines and different distances between trees along the line. The user can choose between four seasons, which will affect sun elevation, and for deciduous species also canopy leafiness, in order to visualize tree shade projection at different times of day. The app is freely available on Apple (https://eneo.fr/agroforestAR_ios) and Android (https://eneo.fr/agroforestAR_android) app stores. To use it, stand at the bottom-left corner of the plot, open the app, “scan” the ground around you to detect the soil surface, and click on the dotted area where you want to plant the first tree. Then walk to the top-left corner of the plot, and click where you want to plant the last tree in the row. Four rows of trees are automatically placed to your right. Beyond the use as an awareness-raising tool for the public, this app could be used in the future to help farmers decide between several possible options for their agroforestry project. Therefore, in the future, we intend to add the possibility to download more complex agroforestry patterns, using the ESSU concept (Rafflegeau et al. 2023) and combinatorial maps (Lemiere et al. 2023) to represent complex agroforestry systems. Thus, an advisor could design one or several alternative systems, send a download code to the farmer, who could then visualize the different options directly in their own fields. The following step will then to link this tool with prediction models to visualize the production of ecosystem services.
ano.nymous@ccsd.cnrs.fr.invalid (Marie Gosme) 21 Oct 2025
https://hal.inrae.fr/hal-05324268v1
-
[hal-05324035] Les coopératives, intermédiaires de la transition numérique de l’agriculture ?
La digitalisation est perçue comme un levier de transformation durable des systèmes agricoles et alimentaires, et une contribution aux dynamiques agroécologiques. Cependant, elle fait l’objet de nombreuses controverses, notamment sur les questions autour de la connaissance et du déploiement effectif des outils numériques dans les exploitations. Cette contribution examine les mécanismes de diffusion du numérique en agriculture à travers l’intermédiation des coopératives. Plus particulièrement, elle explore dans quelle mesure ces organisations peuvent impulser l’innovation numérique en assurant un rôle d’intermédiaire d’innovation. Cette recherche s’appuie sur 20 entretiens semi-directifs réalisés auprès de dirigeants de coopératives vitivinicoles et de fruits et légumes, principalement localisées en Occitanie (France). L’analyse a permis de révéler un engagement des coopératives à différents stades de la transition numérique de l’agriculture à travers une diversité d’actions et de stratégies d’intermédiation, et d’identifier différents profils d’intermédiaires. Le rôle central des coopératives et des techniciens agricoles dans l’articulation entre la digitalisation et les enjeux agroécologiques est identifié.
ano.nymous@ccsd.cnrs.fr.invalid (Békanty Ange Kouassi) 21 Oct 2025
https://hal.inrae.fr/hal-05324035v1
-
[tel-05325824] Optimal control of irrigation: Double mathematical and agronomic modeling towards an application to the Optirrig model
This thesis investigates, with the help of optimal control theory, the topic of optimizing crop irrigation systems in agronomy, at the plot scale, in greenhouse or in open-field conditions, in a context of limited water resources. This specific context gives a crucial role to the tactical irrigation piloting and to the decisions made by the operator over time. Using a mathematical formalism of crop irrigation in terms of an optimal control problem, the goal is to develop theoretical tools in order to provide efficient decision strategies that can be transposed to the operational world. The first part of this thesis focuses on the analysis of an irrigation problem in a context of greenhouse agriculture. We utilize the CCI (Controlled Crop Irrigation) model, which is a generic model written as a controlled (non-smooth) dynamical system, describing order-1 biophysical processes in a simple way, and enabling a study of intrinsic properties related to these central physical phenomena. Our study first leads us to revisit the usual Bang-Bang principle to a version dealing with piecewise affine dynamics under state constraints and a stochastic version, allowing to deduce structural properties of optimal controls and their associated state trajectories. Then we derive the optimality necessary conditions through the Pontryagin Maximum Principle in a non-smooth framework, in order to more precisely characterize the optimal solutions to irrigation problems subject to various types of constraints and criteria. These conditions, combined with the aforementioned structural properties, result in a formulation of a class of optimal policies in terms of parameterized feedback. This means that solving the a priori difficult optimal control problem (in infinite dimension) amounts to performing an optimization in small dimension. This small dimension thus allows us to study, through a global numerical exploration, the properties of optimal solutions, especially the sensitivity of the timing of decision-making with respect to the system parameters (plant, soil, material), and the outcomes (benefits and drawbacks) of sub-optimal decisions. The second part addresses the question of applying these theoretical results in an operational framework closer to real-world conditions. To this end, we employ a double modeling approach, consisting in building connections between the CCI model (simple, theoretical) and the Optirrig model (operational, practical) in order to compare and refine their hypotheses and results. In particular we aim for testing the theoretical scenarios of optimal irrigation taken from the CCI model, by implementing them in the Optirrig model and evaluating their performances in a practical framework. In addition, we take advantage of "averaged cost control" techniques to formalize the irrigation problem under different types of real world uncertainties. In particular, we are interested in finding optimal solutions that are robust regarding the variability of the most sensitive model parameters, as well as finding solutions that optimize the biomass production "on average" with respect to meteorological uncertainties, described as jumps on the state variable.
ano.nymous@ccsd.cnrs.fr.invalid (Ruben Chenevat) 22 Oct 2025
https://hal.science/tel-05325824v1
-
[hal-05353348] Agriculture écologique et technologies numériques - Convergence ou contre-sens ?
Le développement des technologies numériques est souvent présenté comme une réponse aux enjeux actuels de l’agriculture, promettant à la fois un avenir meilleur pour les agriculteurs et les agricultrices, et des bénéfices pour la société dans son ensemble. L’agriculture numérique résoudrait enfin l’équation entre nourrir la population et respecter l’environnement. Elle mobilise à ce titre d’importantes ressources publiques et privées. Pourtant, ses effets font l’objet de vives controverses. Agriculture écologique et technologies numériques analyse la manière dont le numérique s’intègre dans les différentes trajectoires d’écologisation de l’agriculture et les influence. Il explore les représentations, les usages et les transformations de pratiques que son déploiement engendre. Il montre que la digitalisation actuelle entre fréquemment en tension avec une écologisation forte de l’agriculture, qu’il s’agisse des techniques, des objectifs, des logiques de raisonnement, des temporalités ou encore des enjeux politiques et sociaux. Des formes d’hybridation apparaissent toutefois possibles, que ce soit avec des modèles d’écologisation industrielle ou dans le cadre d’une transformation plus globale de la digitalisation, repensant ses modèles techniques, économiques et politiques.
ano.nymous@ccsd.cnrs.fr.invalid (Éléonore Schnebelin) 07 Nov 2025
https://hal.inrae.fr/hal-05353348v1
-
[hal-05353446] Public and private support of the AgriTech : what rationale for what innovation pathway ?
AgriTech are “advanced technological products and services that increase productivity and reduce environmental impact”, according to the website of SummitAgro, a Chilean agricultural inputs company. Similarly, on the website of BPI France, the French public investment bank, we can read that AgriTech start-ups “propose breakthrough innovations and provide more efficient, environmentally-friendly solutions”. We observe positive discourses and a variety of support instruments from both public and private actors, as well as a global increase in the interest of venture capital and private equity in the agrifood sector (Lajoie-O’Malley et al., 2020; Sippel and Dolinga, 2023). This support is legitimized in particular by the “missions” that these technological developments would enable to fulfil, related to the “techno-scientific promise” of AgriTech, such as efficiency, productivity, environmental damage but also working conditions, cooperation or participation (Martin and Schnebelin, 2024). Both private and public organisations support the development of AgriTech, with discourses on “mission-oriented” support. Despite this evolution and the growing influence of private funding on the agricultural innovation system (Glenna et al., 2015), the AgriTech ecosystems, and their impact on the agricultural innovation system, remain poorly studied (Klerkx and Villalobos, 2024). It raises questions about what are the rationales of public and private support for AgriTech development, and how it does or does not promote the “missions” of environmental, economic and social improvement. It has not yet been well investigated whether public and private instruments support AgriTech in the same way or if they have specificities in terms of rationales for action. To address these research gaps, we present a comparative study of AgriTech innovation ecosystems in France and Chile. These two countries offer different settings in terms of institutional context, while both are very dynamic in terms of AgriTech development. In France, the development of the AgriTech sector and its start-ups has been a public policy objective for almost a decade, with discourses on the missions that this technological development will fulfill (Bournigal et al., 2015). Chile has been among the first countries in Latin America to have a dedicated public program for generic start-up support (StartUp Chile - Gonzalez-Uribe and Leatherbee, 2018). More recently, it has been identified as a favourable and attractive area for the development of agricultural technologies (Radiografía Agtech -Endeavor, 2022), and an ecosystem of innovation focused on AgriTech is emerging, driven by public support for innovation and the regrouping of sectoral companies, such as the creation, of AgroTech Chile. Through an analysis of the CrunchBase database we describe the landscape of AgriTech funding. This is complemented with analysis of interviews with key public and private stakeholders of AgriTech innovation ecosystems, to describe the rationale of AgriTech support instruments and who and what they support, through their perceived benefits or their assessment (Kerr et al., 2017). We characterize these instruments according to the organisations providing the support, its objectives, the organisations that can benefit from the support, the type of resources it provides or facilitates access to, and the stage of intervention (Audretsch et al., 2020). The results will allow us to discuss the articulation of private and public support for agricultural innovation, the evolution of agricultural innovation policies in a context of neoliberalism and of a regime of entrepreneurial innovation (« the Silicon Valley model of innovation »).
ano.nymous@ccsd.cnrs.fr.invalid (Éléonore Schnebelin) 07 Nov 2025
https://hal.inrae.fr/hal-05353446v1
-
[hal-05271158] Wheat plant density estimation by inverting leaf density dynamics retrieved from submillimeter-scale RGB imagery
The European Plant Phenomics Symposium (EPPS), bringing together scientists from academia and industry as well as stakeholders dedicated to advancing plant phenotyping research and technology in Europe. This symposium provides a collaborative and interactive platform for sharing the latest innovations, methods, and applications in phenotyping with the goal to address the pressing challenges of sustainable agriculture, climate resilience, and food security across Europe and beyond.
ano.nymous@ccsd.cnrs.fr.invalid (Tiancheng Yang) 21 Sep 2025
https://hal.science/hal-05271158v1
-
[hal-05283043] Assessing fruit tree vigor in peach and apple orchards through wood segmentation in ground-based RGBimages
[...]
ano.nymous@ccsd.cnrs.fr.invalid (Khac-Lan Nguyen) 25 Sep 2025
https://hal.science/hal-05283043v1
-
[hal-05256477] Holding Experts Together?
Contemporary socio-technical challenges such as digital transformation and climate change demand collaboration across diverse professional knowledge domains, including academic, technical, and policy-oriented fields. While interdisciplinary and transdisciplinary approaches have long sought to integrate different forms of knowledge, their effectiveness remains uneven. This panel introduces interexpertise as a novel framework for rethinking collaboration, moving beyond conventional notions of expertise as academic, hierarchical and exclusive. Instead, interexpertise conceptualizes professional knowledge more broadly, encompassing distinct yet interdependent forms of specialized traditions and problem-solving approaches (Abbott).Building on interdisciplinarity (Barry &amp; Born) and transdisciplinarity (Dedeurwaerdere), interexpertise shifts the focus from integration to negotiated authority, recognizing that consensus is neither always achievable nor desirable. It engages with STS — notably trading zones (Galison), boundary objects (Star &amp; Griesemer), and boundary work (Gieryn) — while also drawing from the sociology of professions and organizational science. By foregrounding professional expertise beyond academia, this framework interrogates continuity in entrenched knowledge hierarchies, change in expertise negotiation processes, and critique of dominant models of collaboration.Through theoretical analysis and empirical cases in fields such as digital agriculture, environmental governance, and education, the panel explores how reframing collaboration as interexpertise reconfigures epistemic hierarchies and fosters both co-creation and conflict-resilient collaboration. Case studies will highlight procedural tools such as reflexive negotiation protocols and co-design frameworks that promote inclusive and context-sensitive knowledge integration. By examining both successful collaborations and enduring tensions, this panel advances STS debates on “holding together” and offers interexpertise as a framework for fostering resilience in complex socio-technical systems.
ano.nymous@ccsd.cnrs.fr.invalid (Jongheon Kim) 15 Sep 2025
https://hal.science/hal-05256477v1
