The development of digital agriculture in France

Introduction

In the aftermath of the Second World War, mechanisation was a key lever in the modernisation policy pursued by the State, inspired by the American model¹, to increase French agricultural production. The structuring of the French agricultural machinery sector and measures to support investment in equipment led to a rapid rise in the use of agricultural machinery and its widespread adoption^{2, 3}.

Digital technology is part of this long-term dynamic. It is embedded in an increasing number of ever more effective equipment, capable of integrating a wide range of economic, agronomic and environmental parameters, thereby increasing farm productivity. However, the digitalisation is progressing more slowly in agriculture than in other sectors of the economy, notably because of the complexity of agricultural systems, the high investment costs of these tools for farmers and uneven internet and mobile network coverage in rural areas⁴.

The concept of “digital agriculture” encompasses a broad range of technologies that can be used in farming: computer hardware and software, sensors, connected devices, robotic systems, etc. They can be used for management, production, marketing or knowledge-sharing purposes. This note focuses on technologies deployed at farm level and does not address those used upstream of agricultural production (plant breeding⁵, etc.) or downstream, notably for food processing, distribution and marketing.

The first section retraces the main stages in the development of digital technologies in French agriculture and describes the range of solutions available to farmers, as well as the key players in this market. The following section takes stock of trends in their use according to farm profile and production sector. The final section explores the prospects and challenges associated with this digital transformation of agriculture.

Digital technology in the context of French agriculture

The development of digital agriculture

In France, in the 1990s, the use of two technologies fitted to combine harvesters made it possible to generate yield maps: satellite positioning systems (GNSS, Global Navigation Satellite System) and harvested-quantity sensors. By geolocating variations in production within a single field, these maps provide farmers with information enabling them to adjust input application rates, such as seed or fertiliser, for the following growing season. This gave rise to the concept of “precision agriculture”⁶. Digital agriculture extends this initial objective of rationalising inputs and maximising production towards a more comprehensive optimisation of farm management, whether from an agronomic, economic or environmental perspective. Digital tools and services then developed to collect data, process it and present it back to farmers.

During the 2010s, several technologies reached maturity and became more widely accessible (smartphones, sensors, connected devices, satellite imagery, cloud computing, etc.), leading to a proliferation of digital tools and services. Between 2012 and 2022, the number of international patent applications in digital agriculture grew by an average of 9.4% per year, three times faster than in the previous decade⁷. Some of these innovations have struggled to develop a viable business model, such as handheld sensors for measuring plant stress. Others are still little used despite their potential, for example blockchain for the traceability of agricultural production⁸.

In response to this momentum, public authorities are seeking to prevent data generated by French farms from becoming concentratedin the hands of a few multinationals and are pursuing several types of action⁹. The Convergences Institute for Digital Agriculture was established in 2016, followed in 2017 by the digital and data delegation within the Ministry of agriculture¹⁰. Research programmes were launched, such as the Agroecology and Digital programme, and training in digital technology was introduced in higher agricultural education institutions (Institut Agro Montpellier, Bordeaux Sciences Agro, etc.). In addition, French State investment plans (PIA4, France 2030) were implemented to support innovation combining productivity and sustainability objectives. Their aim is to make digital technology one of the main pillars of the “third agricultural revolution”¹¹ designed to provide healthy and sustainable food for a growing population, while reducing negative environmental impacts. At the same time, in Europe, applications for support under the Common Agricultural Policy (CAP) have been digitalised since 2016, while the digital recording of plant-protection interventions and seed treatments (digital plant-protection register) is due to become compulsory from January 1st 2027.

Digital tools

Nearly 2,000 digital tools for farmers are listed¹² on the French market in 2026. This broad technological range can be classified into five main functions¹³ : “observe and measure”, “organise, manage and trade”, “advise and support”, “exchange, share and collaborate”, and “act and apply in the field” (Figure 1).

Figure 1 – Classification of digital tools used on farms according to their function

^{Source: Aspexit}

At the heart of digital agriculture are field observation data. These may relate to the condition of soils, plants and animals (weight, movement, heart rate, etc.), weather conditions or the environment inside livestock buildings (temperature, humidity, ventilation, etc.). Several technologies can be used to collect this information: fixed sensors installed in fields or in livestock buildings, sensors mounted on agricultural machinery or animals (connected collars, electronic ear tags, etc.), drones, satellites, smartphones, etc. Once collected, these data are stored, processed and returned to farmers in various forms: crop or herd monitoring software, dashboards, mobile applications, online services, management platforms or decision support tools. Their analysis can help monitor production (varieties sown, treatment dates, veterinary information, animal movements, etc.), identify risks at an early stage (water stress, disease emergence, presence of pests, etc.), plan interventions or formulate recommendations on fertilisation, irrigation, plant-protection treatments or animal feed. These data can also be transmitted to precision-agriculture equipment installed on agricultural machinery to adjust specific operations, such as input application rates, section control, guidance assistance, spot weeding, etc. They can also feed automated systems in livestock buildings for feed distribution or environmental control. Farmers also make regular use of the internet in their professional activities. According to a survey¹⁴ carried out in 2024, 86% of farmers use it daily to complete administrative procedures, monitor market prices, purchase equipment or inputs, market their products and exchange information on professional forums. The uptake of digital technologies is particularly pronounced among younger generations of farmers. In 2026, nearly three-quarters of new entrants¹⁵ use digital media, mainly social networks, to seek information that could be useful for managing their farm.

Market actors

Two main types of commercial actors operate in the agricultural digital market. The first category is made of traditional players, including agricultural equipment manufacturers and seed companies, which are gradually integrating digital technologies into their products (GPS, sensors, telemetry, etc.) or supplementing their business models with digital platforms. These large agro-industrial companies (Bayer, John Deere, BASF, etc.) partner with digital giants, responsible for the technical infrastructure to provide online services linked to product sales: machinery fleet monitoring, online diagnostics, personalised input recommendations, etc. These strategic partnerships between multinationals strengthen their already dominant positions in the agricultural market and increase the risk of farmers becoming dependent on a limited number of actors¹⁶.

The second major type of commercial actor consists of young AgriTech companies, which offer innovative services and products based on new technologies at different points along the value chain. These may include hardware solutions dedicated to agricultural production (drones, sensors, robots, connected devices, etc.), software (decision support tools, management, accounting, etc.) or upstream services (marketplaces, social networks or dedicated exchange platforms, etc.)¹⁷. To support the emergence of these French start-ups and structure this innovation ecosystem, the Minister of Agriculture and the Secretary of State for Digital Affairs launched the French AgriTech label¹⁸ in 2021. This initiative was followed by the publication of a roadmap¹⁹ for the development of digital technology in French agriculture, with seven strategic priorities, including support and assistance for these early-stage companies. Despite this, investment in these start-ups began to fall sharply in 2023, mirroring the broader downturn in the global AgriTech market²⁰. Funding fell by 74% over the 2023–2025 period, against an unfavourable macroeconomic backdrop (higher interest rates, slower global growth) and following several failures in the insect-farming sector, one of the flagship segments of French AgriTech (Figure 2). In addition, unlike other economic sectors, where product development cycles and time-to-market are short, AgriTech innovations are constrained by the long timeframes of biological processes²¹. Funding is now concentrating on less disruptive innovations, where growth and profitability prospects are considered more robust (biological inputs, management software, agricultural robotics).

Figure 2 – Trends in investment in French AgriTech start-ups

Trends in digital and robotic equipment on farms

Farmers’ gradual adoption of digital technology

Surveys conducted by the Statistical and Foresight Service (SSP) of the Ministry of Agriculture, Agrifood and Food Sovereignty provide information, in particular, on the increasing adoption of digital tools by farmers. This topic is also taking an increasingly prominent place in the questionnaires sent to them.

The first question on farms’ digital equipment appeared in the 1988 agricultural census (RA). Farmers were asked whether they used information technology (computer or Minitel) for farm purposes. Over successive surveys, the questions became more detailed in order to better characterise equipment and uses. The Farm Structure Survey (ESEA) thus devoted a specific section to IT equipment, robotics and precision-agriculture technologies used during the 2022–2023 agricultural year. The results of these surveys show steadily increasing adoption, reflecting a gradual transformation of agriculture rather than a “digital revolution”. Between 1988 and 2023, their use increased regardless of the economic size of farms (PBS), although substantial differences in equipment rates remain between small and large holdings (Figure 3).

Figure 3 – Trends in the adoption of digital tools based on data from the agricultural census and the ESEA

^{Source: Agreste, Agricultural census and ESEA 2023}

Different uses of digital technology across production sectors

In 2023, more than half of farms were equipped with specialised software, mainly for accounting management or production monitoring. Nearly one-third used decision support tools to obtain recommendations, particularly for the application of plant-protection products or soil fertilisation. Uses (Figure 4) and equipment levels differ according to production sector.

Figure 4 – Breakdown of uses of different types of technology in crop and livestock sectors

^{Source: Agreste, ESEA 2023}

Nearly 40% of farms specialised in arable crops, viticulture, or mixed cropping and livestock farming used decision support tools in 2023. Through the recommendations they provide, these tools make it possible to rationalise treatments and can help reduce pesticide use. Multi-year experimental projects carried out in horticultural production under the national DEPHY EXPE programme showed that the treatment frequency index (TFI) fell significantly when a decision support tool was used alongside biocontrol products²². In viticulture, the VitiREV programme in Nouvelle-Aquitaine tested a green insurance scheme guaranteeing growers financial compensation (paid by a private insurer and supplemented by a public subsidy) in the event of yield losses over a season, provided they had followed the recommendations of a decision support tool for fungicide treatments²³.

In addition, 40% of dairy cattle farms and 32% of granivore farms (pig and poultry) use decision support tool to manage land devoted to animal feed (cereals and forage crops). For grazing management, remote-sensing data from satellite images or drone overflights can be used to monitor and estimate pasture forage growth²⁴.

In crop sectors, three-quarters of precision equipment is used to optimise the application of inputs to fields: high-precision guidance systems, ground speed sensors, automated section control (automatic opening and closing of sprayer nozzles based on GPS positioning), etc. In livestock sectors, by contrast, 70% of the precision equipment used is dedicated to monitoring animals, continuously tracking their health status or automatically controlling livestock buildings (temperature, ventilation, misting)²⁵.

Cost of investment in digital equipment

Technological development of equipment, combined with the surge in raw material prices, electronic components and energy following recent crises (Covid, Ukraine and Iran wars), contributed to the sharp increase in agricultural machinery prices between 2005 and 2026 (+46%)²⁶. In addition, beyond the initial purchase price of equipment, its use entails additional expenditure: maintenance, updates, service subscriptions, support, training, etc. Recent work²⁷ has assessed all this expenditure on digital tools and uses by French farmers through the “total cost of ownership” indicator. It estimates that one-third of farmers who use them spend more than €10,000 per year overall to ensure their proper use. Inequalities between farmers may arise from differences in access to these technologies, financial capacity, training and the quality of network coverage on their farms²⁸.

Contrasting deployment of agricultural robotics

The deployment of robots in French agriculture began in the 1990s with the automation of milking on dairy farms. These robots remain by far the most widespread in agriculture, with an estimated 14,000 units in 2023, particularly on farms with more than 50 dairy cows, nearly one-quarter of which report being equipped with one (ESEA 2023). More than 4,000 other robots²⁹ are used in livestock sectors to reduce the burden of certain repetitive tasks: cleaning floors, bedding litter or feeding animals. An analysis by the French Livestock Institute (Idele)³⁰, comparing the economic profitability of dairy farms, shows that farms equipped with a milking robot and those using a conventional milking parlour have comparable profitability when assessed over a long period. When economic conditions are favourable, robotic farms generate higher incomes thanks to greater productivity per cow. Conversely, this strategy of achieving productivity gains through investment makes them more economically vulnerable in unfavourable conditions, as during the 2016 milk crisis.

On crop farms, robot use is increasing but remains marginal: the number in operation rose from around 100 units in 2018 to 600 in 2023. Most are used in viticulture and market gardening for physically arduous or labour-intensive tasks, such as soil cultivation and weeding. Several harvesting-robot prototypes are under development, but only a few farms, particularly in fruit production, already use them. Despite a recent increase in supply, from five models marketed in 2018 to 25 in 2023, the crop-robot segment is still emerging. Unlike robots used in livestock buildings or in other economic sectors, crop robots must operate in fields (non-standardised terrain, uneven ground, bad weather, etc.), distinguish crops from weeds, navigate accurately and intervene without damaging plants. Their operation therefore relies on a combination of technologies: geolocation, autonomous navigation, computer vision, sensors, end effectors (precise and sensitive when handling plants), etc. Their large-scale deployment³¹ is constrained by their high cost and by performance subject to several constraints: the conditions in which they operate (weather, field layout, soil bearing capacity, etc.), work rates that are still insufficiently advantageous compared with conventional equipment, and regulations that, for many years, were poorly suited to these new types of machinery.

From January 2027, European Regulation 2023/1230 will replace the Machinery Directive that currently governs robotic solutions. Directly applicable in all Member States, it defines the “autonomous mobile machine”³², harmonises safety standards and the conditions for non-continuous remote supervision³³,and clarifies responsibilities in the event of accidents. By changing the legal nature of the instrument, this new regulation therefore harmonises, across all Member States, the conformity rules imposed on manufacturers while incorporating new requirements relating to technological developments, such as connectivity, artificial intelligence, cybersecurity, etc.

In France, another current regulatory obstacle is the ban on autonomous machines travelling on public roads or in any area accessible from them. Farmers must therefore transport the robot on a trailer to their fields or redesign their farm to avoid crossing public roads. Changes to the French Highway Code could nevertheless be considered, following exemptions granted for on-road driving trials carried out as part of the Agricultural Robotics Grand Challenge. Launched by France 2030, this programs aims at accelerating the development of this sector. Other countries, such as Germany, allow autonomous vehicles to travel on roads under precisely defined regulations.

Prospects and challenges of digital technology in agriculture

Future contributions of artificial intelligence

Artificial intelligence (AI) is opening up new prospects in several technological areas of agriculture (Figure 5).

Figure 5 – Examples of AI applications in agricultural production

^{Source: author}

In agricultural robotics, AI techniques (Computer Vision, etc.) enhance machine autonomy: navigation, adaptability, visual identification, etc. For automated weeding, for example, several types of robots are now commercially available, using AI techniques to carry out successive operations autonomously: crop identification, weed detection, interpretation, decision-making and weeding. Chemical-weeding robots can therefore target spraying precisely and adjust the quantity according to the size of the weeds detected. In response to growing health and environmental concerns, and to tighter regulation of plant-protection products, non-chemical robotic solutions (mechanical or laser weeding) are increasingly favoured in recent research³⁴. In fruit growing, robotic solutions using AI are being developed for harvesting³⁵. They consist of a manipulator arm for mechanical movement, a vision system to locate the fruit and assess its degree of ripeness, and an end effector for harvesting (cutting, suction or gripping). Here, computer vision offers solutions to the perception challenges posed by automated picking: irregular tree shapes, variable fruit positions, occlusions caused by foliage, etc.

AI algorithms can also improve the recommendations of decision support tools through their ability to analyse, in real time, massive quantities of heterogeneous information, both in nature (agronomic, meteorological, etc.) and format (images, text, data, etc.). For plant-disease detection, several scientific studies³⁶ are exploring what Deep Learning approaches applied to leaf images could deliver. Building a training dataset is a prerequisite for satisfactory performance of these models. The annotated image database must be sufficiently large and varied to represent different plant diseases at various stages of development and across plant varieties. In livestock sectors, AI is used to process various types of information collected on a farm (movements, sounds, weight, perspiration, etc.) in order to individualise animal monitoring (behaviour, health, feeding, etc.)³⁷. In Canada, a research chair has launched several projects aimed at providing dairy farmers with tools and indicators to detect early signs of poor welfare in cows using AI. The objective is to improve their longevity and reduce the occurrence of disease³⁸. Italian researchers, for their part, have tested an experimental system mounted on a robot dog, using several artificial-intelligence algorithms in different livestock contexts: monitoring hens in a poultry house, observing cows in a barn and monitoring animals at pasture. Designed to be integrated into various robotic platforms, the system can track animals and analyse their behaviour in real time. AI applications are also being developed for grazing management. One example is a prototype combining autonomous robotics, digital twins³⁹ and artificial intelligence⁴⁰. It collects field data in real time, analyses them and guides the herd towards optimal grazing areas. These are determined according to the availability and quality of grass, as well as the location of animals and their nutritional status.

More recently, specialised conversational agents for agriculture have been developing, based on large language models (LLMs)⁴¹. They provide agronomic advice (e.g. E.L.Y) or livestock-production advice (e.g. FarmLife), analyses of farm accounts (e.g. TerraGrow), information on products and equipment and on current regulations (e.g. Lexagro), or support for administrative tasks (e.g. Hectar.ai). Since 2024, a hackathon (a programming marathon) has been organised at the Paris International Agricultural Show by the association La Ferme Digitale⁴², during which interdisciplinary teams have 30 hours to develop functional solutions based on generative AI that address agricultural challenges. To support the development and widespread adoption of these various AI-based innovations while protecting farmers from the associated risks (algorithmic bias, opaque decision-making, AI hallucinations), a public report⁴³ by the CGAAER recommends adapting the government’s AI strategy to the agricultural and agrifood sectors.

Potentially ambivalent effects of digital technology on agriculture in the future

Like any technological innovation, digital technology has and will have potentially ambivalent effects on French agriculture in many respects (generational renewal, food sovereignty, climate change, farm profitability, international competitiveness, etc.). Digital technology and robotics will help reduce physical and mental strain, automate certain repetitive tasks and thereby make the occupation more attractive. According to a survey in the ruminant sector, livestock farmers identify time savings and improved working comfort as the main benefits of these technologies⁴⁴. However, some farmers consider that these tools distance them from direct contact with living organisms and “distort” their conception of their occupation⁴⁵. For others, they are a source of cognitive overload linked to the continuous flows of information they generate. Moreover, in farming systems increasingly driven by data, information collected by sensors is often processed directly by algorithms to formulate recommendations, without the underlying reasoning being explained. This algorithmic opacity and the farmer’s reduced involvement in observing and understanding agronomic processes may raise fears of a gradual loss of knowledge and a weakening of decision-making autonomy⁴⁶. Finally, from an economic perspective, the productivity gains associated with equipment modernization must be weighed against the financial costs involved.

Agriculture is increasingly confronted with environmental challenges, including persistent pollution, intensifying climate change and water stress. Through its ability to capture large volumes of heterogeneous information, analyse them and transmit them in order to adapt operations⁴⁷, digital technology can provide solutions to support farmers in the agroecological transition, which involves more complex production systems in both crop and livestock sectors⁴⁸. However, these technologies could also support a form of “weak agroecology”, focused on optimisation and efficiency without more fundamentally revisiting current dominant models, whose sustainability is being questioned⁴⁹. The manufacture and use of digital tools also have an environmental impact, which varies according to the technological complexity of the devices. Different devices are available on the market to perform the same agricultural task. For example, dairy cows can be identified using electronic ear tags based on radio-frequency identification (RFID). They can also be identified by a camera system coupled with computer-vision algorithms which, in addition, provides information on animal behaviour (feeding and resting phases, distance travelled, likelihood of oestrus, etc.). The environmental impact (manufacturing and energy consumption) of the second system is substantially greater⁵⁰.

In France, the desire to promote responsible digital agriculture, combining economic and environmental objectives, was reflected in particular in the launch, in 2023 for an eight-year period, of the national research programme on agroecology and digital technology, under France 2030. This programme has four scientific strands: responsible research and innovation; the potential of genetic resources for agroecology; new generations of agricultural equipment; and modelling and data-processing tools and methods for decision support. In 2026, around twenty research projects had been launched under the programme and 56 PhD projects had been funded.

Digital technology also brings risks about maintaining control over the sovereignty of French agriculture. The growing volume of information now produced by farms – yields, cropping practices, plant-protection treatments, economic performance – constitutes a strategic resource for research, innovation and public expertise. Privileged access to these data by a few private actors could strengthen their dominant position, notably by enabling them to anticipate developments in agricultural markets⁵¹. Digital agriculture also increases the risks of farmers becoming technically dependent on “proprietary” equipment or software. This may result in asymmetries in relations between suppliers of digital solutions and farmers, as well as greater vulnerability to failures and cyberattacks. These problems are reinforced by the rise of AI, which plays a central role in agricultural decision-making processes.

From a legal perspective, the large volumes of data collected raise issues concerning the protection of farmers’ economic and personal data, the conditions under which they are shared, their portability and the purposes for which they are used. Accordingly, several recent EU regulations (Data Act⁵², Cyber Resilience Act⁵³, AI Act⁵⁴, Machinery Regulation⁵⁵) aim to guarantee farmers, among others, effective control over access to use, sharing and portability of their data, thereby limiting the risk of this information becoming concentrated in the hands of a few multinationals.

Conclusion

Digital technologies have gradually become embedded in farming with a range of tools and services to support farm management, optimise technical operations, automate certain tasks and reduce physically demanding work. However, levels of digital equipment vary according to production sector, farm size and investment capacity. Farmers’ age and level of education, together with the quality of internet and mobile coverage across the territory, have been identified as factors that may hinder farmer’s adoption of digital technology⁵⁶. According to a study based on Eurostat’s digital-skills indicator, French farmers appear to perform above the European average⁵⁷.

Nevertheless, the spread of digital technology is a source of significant tensions for farmers. Expected gains in productivity, precision or working conditions must be weighed against all the costs involved (purchase, maintenance, subscriptions, updates, etc.), as well as the risks of information overload or loss of decision-making autonomy. The development of “proprietary” solutions, the concentration of agricultural data and increased exposure to cyberattacks create new sovereignty challenges for French agriculture.

Technology alone cannot provide a solution to the sector’s structural challenges, whether generational renewal, adaptation to climate change, farm competitiveness or the reduction of environmental impacts. Responsible use of innovations geared towards improving working conditions, production quality and the resilience of sustainable farming systems can, however, provide levers for progress, provided that they remain subordinate to agriculture’s agronomic, economic, social and environmental objectives.

Jérôme Lerbourg
Centre for studies and strategic foresight

Footnotes

¹ R. Dumont, Les leçons de l’agriculture américaine, Paris, Flammarion, 1949.
² Brunier S., Pinaud S., 2025, « La montée en puissance de l’agriculture française : infrastructure machinique et engrenage de la production », Revue d’anthropologie des connaissances, 19-4.
³ Mirouse B., 2026, Une mutualisation croissante des machines de récolte entre agricultures, Agreste Primeur, n° 2026-3.
⁴ Council's in-house research service (ART), 2025, From screens to fields : how digitalization is transforming agriculture.
⁵Zhang Y. et al., 2025, « Revolutionizing Crop Breeding: Next-Generation Artificial Intelligence and Big Data-Driven Intelligent Design », Engineering, vol. 44, p. 245-255.
⁶ Oui J., 2024, « Agriculture de précision et tournant environnemental », Réseaux, 244(2), p. 117-149.
⁷ Office européen des brevets, 2025, Digital agriculture. Towards sustainable food security, rapport.
⁸ Seminar replay, « Regards croisés sur 10 ans d’innovations numériques en agriculture », organized by the AgroTIC Chair in December 2025.
⁹ Bournigal J.-M. et al., 2015, Agriculture et innovations 2025, report to the ministers responsible for agriculture and research.
¹⁰ From June 2023, the duties of this delegation were transferred to a new position of “Senior Digital Officer”, responsible for coordinating the ministry’s actions in the development of digital technology.
¹¹ Plan d’investissement France 2030, 2021, Objectif 6.
¹² WIKI AGRI TECH (participatory platform for digital tools for farmers)
¹³ Leroux, C. et Touraine, A., 2023, Classification des outils numériques en agriculture, Aspexit et Chaire AgroTIC, 19 p.
¹⁴ Étude Agrinautes 2024 by ADquation for NGPA.
¹⁵ Ipsos BVA poll commissioned by ACTA on les pratiques d’information des nouveaux agriculteurs, 2026.
¹⁶ Sauvagerd M. et al., 2024, Big Data & Society.
¹⁷ Lerbourg J., 2021, Les grands enjeux de l’agriculture numérique : équipements, modèles agricoles, big data, Analyse, n° 172, Centre for studies and strategic foresight, French Ministry of Agriculture, Agrifood and Food Sovereignty.
¹⁸ French government, 2021, Agriculture et Innovation : lancement de la French AgriTech, press kit.
¹⁹ French government, 2022, Agriculture et numérique, feuille de route.
²⁰ DigitalFoodLab, 2025, FoodTech en France : rapport sur les investissements et l’état de l’écosystème, rapport.
²¹ Fernandez-Vidal J., Alarcon S., 2025, « Financing agricultural innovation: Challenges and alternatives to venture capital in the AgTech sector », Food Policy, vol.136.
²² Paris B. et al., 2026, « 2.ZERHO : Des combinaisons de leviers et un outil numérique en horticulture pour approcher le pilotage en zéro pesticide », Innovations Agronomiques, 109, p. 44-57.
²³ Lefebvre M. et al., 2025, « Green Insurance for Pesticide Reduction: Acceptability and Impact for French Viticulture », European Review of Agricultural Economics, Vol. 51, p. 1201-1272.
²⁴ Jennewein J. et al., 2025, Multi-sensor proximal remote sensing for cover crop biomass estimation at high and moderate spatial resolutions.
²⁵ Lerbourg J., 2025, « Agriculture et outils numériques », Économie et société à l’ère du numérique, Insee Références, édition 2025, p. 104-105.
²⁶ Agreste, Insee, Indice des prix d’achat des moyens de production agricole (IPAMPA)
²⁷ Jaballah B. et al., 2026, « Les investissements numériques dans les exploitations agricoles françaises », Économie rurale, 395(1), p. 35-49.
²⁸ Académie des technologies, 2019, Big Data. Questions éthiques.
²⁹ Ruiz V., 2023, Usages des robots en agriculture, Observatoire des usages du numérique en agriculture.
³⁰ Idele, 2024, Robot de traite : au-delà d’un simple équipement, quels impacts sur les systèmes ?
³¹ Chaire AgroTIC, 2024, « Comment réussir le passage à grande échelle de la robotique en agriculture ? », seminar.
³² Definition of an autonomous robot in European Regulation 2023/1230: “Mobile machinery that has an autonomous mode, in which all essential safety functions of the mobile machinery are ensured in its travel and working area without permanent interaction by an operator”.
³³ The autonomous machine must be equipped with a “supervisory function” so that it can be subject to non-continuous remote supervision by means of a device enabling information or alerts to be received and limited commands to be given to the machine.
³⁴ Lytridis C., Pachidis T., 2024, « Recent Advances in Agricultural Robots for Automated Weeding », AgriEngineering, 6(3), p. 3279-3296.
³⁵ Kaleem A., 2023, « Development Challenges of Fruit-Harvesting Robotic Arms », AgriEngineering, 5(4), p. 2216-2237.
³⁶ Al Kafi A. et al., 2026, « LeafAI: Interpretable plant disease detection for edge computing », PLoS One, 21(1).
³⁷ European Parliament Think Tank, 2025, Transforming animal farming through artificial intelligence.
³⁸ Chaire recherche-innovation en bien-être animal et intelligence artificielle, Projet Well-E.
³⁹ A digital twin is the virtual replica of an entity (an object, process or system), supplied with information flows. The twin digitally represents the state of the entity throughout its life cycle and may even be used to simulate changes to it.
⁴⁰ Sibona F.et al., 2025, « A Framework to Enable Eco-Cyber-Physical Systems for Robotics-Focused Digital Twins in Smart Farming », Journal of Field Robotics, 42, p. 3016-3 037.
⁴¹ Acta, La Chaire AgroTIC, Académie d’agriculture de France, 2026, Intelligence artificielle en générative en agriculture, 70 p.
⁴² La Ferme Digitale is an association promoting innovation and digital technology in the agriculture and food sectors.
⁴³ CGAAER, 2025, L’intelligence artificielle au service de l’agriculture et de l’agroalimentaire, rapport n° 25034.
⁴⁴ Idele, 2023, Technologies numériques : comprendre et accompagner leur essor en élevage ruminant, Dossiers techniques de l’élevage n° 8.
⁴⁵ Mazaud C., 2019, « La conception du métier pour comprendre l’appropriation du numérique par les agriculteurs », Sciences Eaux & Territoires, n° 29, p. 50-51.
⁴⁶ Brault N., Yatskul A. et Dubois M., 2026, « The Algorithm, the Farmer and the Agronomist: Who is Smart in Smart Farming? », Philosophia Scientiæ, 30-1, p. 55-76.
⁴⁷ Rogel Gaillard C., Sainte-Marie J., 2024,Agroécologie et numérique : quelle synergie pour la transition écologique ?
⁴⁸ Anastasiou E. et al., 2025, « Assessing the agroecological impact of digital tools in livestock production: A systematic review », Smart Agricultural Technology, vol.12.
⁴⁹ Leroux C., 2025, Quel(s) numérique(s) pour accompagner la transition agroécologique ?, Aspexit.
⁵⁰ La Rocca P. et al., 2024, « Estimating the carbon footprint of digital agriculture deployment », Journal of Industrial Ecology.
⁵¹ Académie des technologies, 2019,Big Data. Questions éthiques.
⁵² Parlement européen et Conseil de l’Union européenne, 2023, Règlement (UE) 2023/2854 sur les données.
⁵³ Parlement européen et Conseil de l’Union européenne, 2024, Règlement (UE) 2024/2847 sur la cyber-résilience.
⁵⁴ Parlement européen et Conseil de l’Union européenne, 2024, Règlement (UE) 2024/1689 sur l’intelligence artificielle.
⁵⁵ Parlement européen et Conseil de l’Union européenne, 2023, Règlement (UE) 2023/1230 sur les machines.
⁵⁶ Aranguri M., Mera H. et Lucini C., 2025, « Digital Literacy and Technology Adoption in Agriculture: A Systematic Review of Factors and Strategies », AgriEngineering, 7(9), p. 296.
⁵⁷ Radlinska K., 2026, « The Role of Digital Skills in the Digital Transformation of Agriculture—Evidence from the European Union », Sustainability, 18(3), 17 p.