chevron-down Artboard chevron-up Page 1



Farming does not escape from the ICT-revolution. The digitalization of agriculture is based on a number of technologies coming from outside the agricultural sector, like global positioning systems, cloud computing, drones, Internet of Things (IoT). In essence these technologies support very detailed data capturing that in principle can easily be shared (cloud technology) and interpreted with big-data techniques.


Agriculture and the food chain become more and more data-driven by the use of ICT. Precision farming is a good example. The digitalisation of agriculture will have effects on the organisation and functioning of markets for products. Not only on existing markets for products, but also on markets for new products and services, and markets for new software and data as well as on the labour market. Digitalisation stimulated by precision agriculture can be a key contributor to solving problems in agriculture with regards to the environment, the working conditions, food safety and other public interests. However there will also be effects on the organisation of the food chain including the family farm, potentially shifting decision making from the farm and encouraging farm enlargement and chain integration. This could have negative effects on rural employment.

As for the moment the balance seems to be positive, the development and adoption of ICT could be encouraged through the European Innovation Partnership, a public policy of open data and using up to date ICT by the Managing Authorities (Paying Agency) that integrate data exchange with platforms in the food industry. Environmental policy and Food Safety policy could also benefit from ICT and stimulate it.

Special attention is needed for establishing an open data exchange in and around the food chain, with adequate standards and platforms for data exchange that have a governance structure that prevents misuse of natural monopolies or lock-in effects. Making farmers owner of their data (although judicially speaking that is a difficult concept) and providing opportunities to control the flow of their data to stakeholders by authorisations should build trust with farmers for exchanging data and harvest the fruits of the analysis of big data.

Rural development policy and regional policy should guarantee access to band wide in the internet (4G/5G) and help to find new forms of employment in case agriculture becomes less labour intensive.


Following a mechanical and a ‘green’ (genetics and chemicals) revolution, agriculture is now confronted with an ICT-revolution. The digitalization of agriculture is based on a number of technologies coming from outside the agricultural sector, like global positioning systems, cloud computing, drones, Internet of Things (IoT) etc. In essence these technologies support very detailed data capturing that in principle can easily be shared (cloud technology) and interpreted with big-data techniques. The European satellite programs Galileo (navigation) and Copernicus (earth observation) contribute to this. For agriculture the Copernicus program is great news because two satellites provide detailed field imagery to enable precision agriculture applications and one satellite provides imaging capacity at 250 m for global agricultural monitoring. And the Galileo program delivers Europe’s own satellite navigation system which is unique in its civil governance compared to the military systems. Although signal receivers claim to be ready for Galileo, its delay in launching spacecrafts results in reduced applicability. Nevertheless, Global Navigation Satellite Systems (GNSS) are a key enabler for many precision agriculture techniques. Augmented by ground stations, it enables farmers to navigate their machines at 2 cm accuracies, even when revisiting the field a week later. This enables auto-guidance of tractors and implements and facilitates precision seeding, precision weeding, precision fertilising and precision harvesting (amongst other applications). This ICT-revolution results in what is known as smart farming or precision agriculture (see briefing paper 3 in this series (Kempenaar and Lokhorst, 2016) for an overview).

In this briefing paper we describe the effects of digitalisation of precision agriculture on management and business models, the organisation of food supply chains, the markets, governance issues, and current relevant government policies, as these effects influence the future of farming, and should be taken into account in a foresight technology assessment, to which this briefing paper aims to contribute.

We base our description on scientific literature, as well as documents like the strategic agenda of the ERANet ICT-AGRI (Lötscher, 2012), work from the EU’s Future Internet PPP, several technology scans / foresights and a briefing paper we prepared for the OECD (Van der Wal et al., 2015).

Select Topic


Management and business models

Effects on management and the family farm

This briefing paper concentrates on smart or precision farming, adapted from the description given by JRC (Zarco-Tejada et al., 2014): Precision Agriculture (PA) is a farming management concept based upon observing, measuring and responding to inter and intra-field variability and needs in crops and to variability and needs of individual animals with the use of digital techniques. Precision Agriculture has the potential to increase yield, reduce production costs, improve sustainability of production, animal welfare and environmental stewardship. Thus, PA contributes to the wider goal concerning sustainability of agricultural production.

This description underlines a number of capabilities that Smart Farming performs. The technology leads to much more and better data capturing: nature is being digitalised. That leads to better control of the biological productions processes that take place under unpredictable influences like the weather. This makes the process better manageable. In essence this is a process of industrialisation that is not new and already strongly present in indoor agricultural sectors like glasshouse horticulture and the pigs and poultry industry. Better control can of course lead to better optimalisation, or even to self-managing of processes as the software algorithm can deal itself with variations in conditions. The self-driving machinery is the most extreme example of this. More precise this development can be sketched as:

  1. Monitoring: understanding the location, ownership, history, destination, quality conditions, and other functional properties of products and other objects by means of sensors and external data sources.
  2. Controlling: intelligence is added in order to take corrective measures i.e. specific rules prescribe how the objects should respond to certain events. And the situation or environment of the object can be adjusted remotely by actuators.
  3. Optimizing: the performances of the food supply chain are improved by applying advanced algorithms and analytics for simulation and support of the decision making based on optimisation models and predictive analytics.
  4. Self-managing: through combining monitoring, control and optimization objects can operate independently (autonomous) during their way through the food supply chain without human intervention, either on the spot or remotely. Autonomous objects can also become self-adaptive systems which are able to learn about their environment, make a diagnosis of their needs, and adapt to the preferences of the users.
Capabilities and enablers in Smart Farming
Source: LEI, T&U Board, 2016

This development is based on a number of enablers, important conditions for smart agriculture and smart food supply chains. Some of them will be discussed later in this paper, but they are:

  • Disruptive technologies: hardware for mechanisation (robotics), RFID, sensors, wireless networks including broad band in rural areas, web service technology, cloud computing, big data, predictive analytics tools.
  • Standardisation: fast, error-free and efficient exchange of digital data within and between companies based on information standards.
  • Competencies: awareness, adoption and knowledge of digital information systems and standards and the skills to use them.
  • Governance: organisational implementation and business models, including agreements on ownership rights and decision rights, remuneration, risk management etc. (see section below on data ownership).

This effect of ICT on management has important consequences. It means that some forms of labour on the farm are replaced by machines and software. Farming becomes even more capital intensive and less labour intensive.

This could have implications for rural development: it implies less work in agriculture, but also that some of the work and decisions move from the farm to experts that provide their services remotely. Where in the past a farmer had a look at his cows to see which ones are in heat and can be inseminated, that is now checked by a pedometer tagged to the cow. Originally that service sent an SMS to a farmer with a suggestion from the breeding organisation on the selection of the bull. As most farmers took the first choice, a breeding organisation like the Flemish-Dutch cooperative CRV has now turned that into a service where a remote computer in Arnhem takes the decision on the selection of the bull. Such services are typically provided from metropolitan regions with an attractive business environment.

Another effect on management is that the risk of moral hazard that exists in agriculture can be reduced or disappear. This risk of moral hazard occurs because the production process is not easy to monitor. That means that an investor or a manager (in economic theory labelled as ‘the principal’) cannot easily control the worker (‘the agent’). How should an investor learn if his farm manager is doing his utmost best, as if he was the owner himself, and is not incorrectly blaming his mediocre results to bad luck in weather or diseases? How should the manager know that the farm labourer at a far-away field is not shirking? Such agency problems lead to transaction costs (for the principal to control the agent). ICT reduces this risk and associated costs considerably. This could have big implications for the future of the family farm.

There is a rival theory that explains the fact that in Europe family farms dominate over latifundia and investor-owned plantations. That is based on the observation that farm families employ their capital and own labour at relatively low incomes without much labour from outside due to other frictions in markets. Outside labour in weekends, at nights and in overtime is often expensive due to labour market regulations, where at the same time tax and social security rules are often favouring the self-employed in farming.

1 Latifundia are very extensive parcels of land, historically owned by the upper class

As working outside the family farm often includes travelling or leaving the farm, and total farm income (including a capital remuneration) is more an objective for the family than the income per hour worked, the outflow of labour from agriculture is often slow and takes place at the generational transfer when potential successors vote with their feet. These low incomes makes investment in agriculture for investor-owned firms like food companies unattractive as it implies low rewards for a high risk. Also this underpinning of the family farming system is effected by ICT, as smart farming makes risks better manageable and perhaps even tradeable in insurance or capital markets.
In conclusion, and in combination with other factors as the developments in the labour market (demography with less high skilled farmers and migration in Europe of low-trained labour), the capital market (with low interest rates making investment in land more attractive) and product markets (with higher prices due to global scarcity of food and biomass), ICT could have an effect on the organisation of the family farm, giving way to medium sized enterprises as we know them in other sectors of the economy.

Effects on business models

Before turning to business models, that describes with which value proposition to clients money is earned by firms and farms, it should be noted that the ICT-revolution has wider implications for farming. First of all ICT has big effects in scientific developments that also influence agriculture. Genetics, based on computer power, is a good example. Second, and probably even more important if it comes to the consequences of ICT on the organisation of the agriculture sector, developments like the Internet of Things are not only adopted in smart farming techniques, but also in the agri-food supply chain (for tracing and tracking, supply management, compliance auditing etc.) and even at consumer level (starting to link food consumption, lifestyle behaviour data like in wearable tech and health data). Such data trends have the potential to be disruptive for current food chains (e.g. more online sales in short supply chains, more prescriptive farming) and lead to new forms of organisation and new business models (Figure 5). That implies that a discussion on business models in farming should not concentrate on the current boundaries of the family farm, but look a bit broader (including data exchange in the food chain).

Effects on ICT on business models in the food chain
Source : LEI, Wageningen UR

Digitalisation of farming processes continues to expand and intensify. The supply and demand of farming data is rapidly growing. There is a surge of data-tools in the market and even more in the making. Data-driven initiatives are steadily increasing in agrifood chains. Farming is becoming a booming data harvesting business where many players are taking bites into data generated by farming. Many data-driven initiatives are still exploring viable business models to capture the value of data. A variety of business models are being used and develop with different value propositions to different stakeholders (Ge and Bogaardt, 2015).

The variety of business models of those data-driven initiatives (e.g. data exchange platforms) present in the agrifood sector, can be illustrated by examples of the five typical business models according to Spijker (2014) in which value is created from data: by selling data, by innovating products through data, by swapping commodity offerings into value-added services, by creating interaction in the value chain, and by creating a network of value based on data exchange. Below examples are described of each of these business models.

Basic data sales

An example is Farmobile in the US that sells a simple data collection tool that centralises grower’s agronomic data from multiple systems in one electronic farm record. Farmobile standardizes the data and makes it easily searchable for customers who want to purchase data. The data management system of Farmobile originates with a $1,250 annual subscription fee. If farmers opt to share their data through Farmobile, they will get 50 percent of the revenue derived from selling the data. At Farmobile the electronic farm record (ERF) is owned by the farmer. Data is stored as long as the subscription remains active. The farmer’s data is housed on cloud servers of Farmobile. Farmers have the power to authorize or deny access. By the end of 2015 Farmobile has raised $5.5 million in equity financing from Anterra Capital, for developing new modules on top of its database platform. Anterra Capital is a growth capital fund, jointly funded by proprietary investment of FIL and Rabo Private Equity.

Another example is Farmers Business Network (FBN), with investment by Google Ventures. Its mission is to make data useful for farmers to select the optimal seeding grade for their variety and their field in order to reach the maximum potential. In 2015 FBN has aggregated data from 7 million acres of farm land across 17 states in the USA. FBN is able to assess the performance of 500 seeds and 16 different crops. No data is shared with other parties. Access is by payment: 500 USD per year. FBN is not linked to any company. They are a community of farmers and independent persons (Ge and Bogaardt, 2015).

Product innovation

Equipment manufacturers such as the manufacturers of tractors, combine harvesters and milking robots, collect data from the agricultural machinery of the farmer such as location of the machinery, engine hours, operational data (e.g. amount of fuel used) and diagnostic data of the machinery. This is first of all done to support smooth operation of the equipment and have feedback for further innovation of the product. In some cases there is perhaps also a strategy to develop new services based on the collected data. Examples of equipment manufacturers that collect farm data are e.g. Lely Industries (data from milking robots) and John Deere, that stores the data in the web portal as a service for the farmer and if so requested combines it with historical and real-time data regarding weather prediction, soil conditions, crop features etc. in order to help farmers to run and manage all their operations. The exchange of farmers’ machinery data is limited to the subsidiaries of John Deere, and authorized dealers and suppliers in order to monitor safety and equipment performance and enhance customer support. The platform itself, however, is open to everyone free of charge (Ge and Bogaardt, 2015).

Commodity swap

In a commodity swap, data is exchanged between (for instance) farmers and food-manufacturers to increase the service-component of the transaction. Examples are software made available by processors.

Value chain integration

Multinational agricultural biotechnology corporations like Monsanto and Dupont have adopted the strategy of acquiring start-ups to strengthen their existing position. For example in 2013 Monsanto acquired weather and agronomic data modelling start-up Climate Corporation that provides planting advice to farmers based on data science. And Monsanto’s primary software product FieldScripts helps farmers to maximize productivity, minimize risks and realize higher yields. This move into prescriptive farming originally also included an investment in a machine manufacturer, but this business has now been sold to Deere. Monsanto charges $10 per acre for FieldScripts. Monsanto states that they share farmer’s business (and personal) data only with subsidiaries and business partners of FieldScripts. No data is shared, traded or sold with marketers. Monsanto may publish data related to FieldScripts but only with expressing written consent of the farmer and without disclosing the name and field location of the farmer. Furthermore the agreement between FieldScripts and the farmer states that in no event Monsanto and seed dealer agents are liable for any incidental, consequential, special or punitive damages resulting from the use of FieldScripts (Ge and Bogaardt, 2015).

Value net creation

In the Netherlands Farm Digital, a public-private partnership programme within the Dutch top sector policy, is standardising and digitalising farm data about food safety and sustainability, and developing and implementing an independent digital platform that will enable users to freely exchange certification data. Farm Digital has set up a proof of concept with AgriPlace, a start-up by a Dutch NGO with a sustainability compliance objective, for farmers to share their certification data like GlobalGAP, with the auditing organisations and the food processors (or, in the case of farmers in developing countries, with Dutch importers).

Another example is the EU project FISpace (Future Internet Business Collaboration Network) which is now available for commercial exploiting by offering a business-to-business collaboration platform that could link platforms like, 365Farmnet, Akkerweb, Agriplace and others via a Linux-like Open Source model. Several EU FI-PPP accelerator projects are using the platform. For example SmartAgriFood (see its conceptual architecture in the figure below) which is supporting SMEs in the development of smart services and apps for the agri-food sector, and Finish which focus on projects realizing software applications for complex supply chains and networks of perishable food and flowers (see chapter 5 for more information on platforms).

Architecture of Smart Agri Food
Source: Wolfert et al, 2014.

Food supply chains

In this section we elaborate the point that was introduced above, the fact that ICT influences the governance of global value chains. As in the discussion on the effect of ICT on the family farm also here the fact is that ICT lowers transaction costs. It becomes cheaper to contact others (in social media as well as in Alibaba and other trading platforms). For an economist this suggests that markets will work better and that activities that are done within companies can sometimes be better outsourced to independent suppliers or even self-employed persons. This is a process that has been observed in the economy over the last decades.

When it comes to agriculture, where self-employed working in family farms are already the norm, the effect of ICT could be a contrary movement. As explained before, ICT leads to better control of the production process, including a better planning. Production can be better programmed, and developments in genetics work in the same direction. A good example is the “plant factories” as operated by Green Sense Farms (USA: nearly 5700 m3 of growing vegetables under red and blue LED lights 22 hours per day) or Aerofarms. This company uses similar technology and states on its website: “We disrupt traditional supply chains by building farms on major distribution routes and near population centers. We defy traditional growing seasons by enabling local farming at commercial scale all-year round. We set a new standard for traceability by managing our greens from seed to package. And we do it all while using 95 per cent less water than field farmed-food and with yields 75 times higher per square foot annually”. They do this by building, owning and operating indoor vertical farms that grow safe, nutritious food with a claim of better taste as the vegetables don’t have to be washed, and where the harvest moment is totally predictable from the moment of seeding.
Such programmability, in addition to other trends in the markets (like the big competition between retailers that want to differentiate themselves with specific products), can lead to asset specifity: the fact that a supplier invests in machinery, production methods or production know-how that is very specific for one buyer. This creates an interdependency in the chain, that asks for another governance mechanism, to reduce the dependency and the risk of power-play (ex-post haggling in the economic jargon). Spot markets are then replaced by contracts or even joint ventures. Competition gives way to cooperation, to maximize the value in the chain in addition to deals to share the value. Figure 7, based on (Boehlje, 1999), illustrates this, including the aspect that the optimal governance structure is also influenced by the question if the contribution to this value maximization process can be objectively shared in the market. ICT contributes to asset specifity as it generates much more data that can be added to the transaction, as explained in a section within Management and business models. The large array of food safety and sustainability programs in the food chains are a good example. More data on food safety and sustainability can be measured, audited and accompany the sales transaction in schemes like GlobalGap, BRC, Fair Trade, Organic production etc. Retailers, under pressure of consumers, NGOs and government ask for such schemes and link them to different segments in the consumer market. This leads to investments of farmers (and other partners in the food chain) into such schemes. Such investments are fixed costs that are easier for larger farms and contribute to an increase in farm size (although small farms can benefit because the schemes also transfer know how of desired production methods to farmers). The example of Farm Digital given in the section on ‘Management and business models’ shows how this process is strengthened once that these data start to be exchanged by ICT in platforms like Agriplace. On one hand that lowers the effect as it becomes cheaper to exchange data, but on the other hand it makes it also easier to exchange even more data to support certification and auditing, and perhaps even establish new schemes for new market segments.

ICT provides many options to codify the transaction, but the improved programmability and other aspects discussed above lead to more complexity in the transaction and more coordination (as information of the consumer or retailer or food processor is needed to take on-farm decisions). That moves the governance from the Market arch type towards a Modular type in which the supplier becomes a turn key supplier that more closely work with the food processor or retailer. This is a trend that is clearly observable in for instance the growing of fresh tomatoes.

Governance of Global Value Chains
Source: adapted from Gereffi et al., 2005

Where the modular governance form still asks for relatively high competences of the suppliers, as they process the information (including that of their lead company clients) the Captive model arises when most of the decisions are taken centrally. Either because decisions with the supplier have been automated in algorithms, or if there is an information advantage at the central level due to e.g. the aggregation of data from the different suppliers with up to date ICT.
Contract design between farmers and food processors illustrates their position in such value chains. A nice example is the difference between a contract for canning peas and one of sugar beet (Bogetoft et al., 2002). Where in sugar beet growing there is a lot of autonomy for farmers to decide on sowing and harvesting dates, choice of variety etc., this is not the case in contracts for growing vining peas for the canning industry, where most of the decisions are taken centrally. The main reason is that in sugar beets the harvested product can be stored a few weeks, either at the farm or with the factory, without too much costs and loss of quality. Therefore decisions on harvesting moments can be left to the farmer who can take local circumstances (soil, weather forecasts, farm planning) into account. Peas have to be processed within three hours after harvesting and the factory needs a constant stream of raw material. Therefor decisions have to be made centrally. Other aspects play a role (sugar companies are often cooperatives, in the past the farmers owned a lot of the harvesting machines themselves, where viners are owned by contractors), but the point we want to make here is that the organisation of value chains is a result of drivers like the complexity of the transaction and the competences needed with the supplier. ICT changes those drivers, via better programmability, asset specifity and lower transaction cost to provide the data to partners in the chain and take decisions remotely.

Hence ICT in the form of smart farming and smart industry (Industry 4.0) will change the way the global food value chains are organised. Besides competition, that is very dominating in the current discourse on the power balance in the food chain, collaboration to create more value with new business models and new forms of governance will become more important.

Competition between organisational forms

The analysis above shows that in an economy not only products compete, but also organisational forms. For current organisations that implies that organising the data-exchange in the food chain through collaboration can help them to stay competitive. In other words, firms in the food chain (e.g. a feed company, a breeding organisation, farmers and a slaughterhouse or dairy cooperative) can choose to organise the data-exchange between their organisations to gain the advantages of a data-driven food chain and keep more or less the current delineation between the organisations, or they can give way to a more integrated value chain.

Depending on such decisions two different scenarios could become possible: the captive prescriptive farming model, or the open smart farming model. In the captive prescriptive farming model:

  • The farmer becomes part of one integrated supply chain as a franchiser/contractor with limited freedom.
  • Actors in the chain work on one ICT-platform for e.g. a potato breeder, a machinery company, a chemical company, farmers and French fries processors.
  • Some of these actors will merge into the integrated company (compare the situation in some pigs- and poultry integrations)
  • With a weak integration of data exchange with service providers like banks and accountants and with governments as a result
    In the open smart farming model:
    • The current actors of the food chain stay independent, the governance of the food chain is more that of a modular than a captive arch-type
    • Data is shared through common, open platforms with competition between specialised services and apps.
    • That also supports data exchange with service providers and government, and thus reduces administrative burden.

Prescriptive farming in the USA is currently more organised along the first model than the open model, seen the examples given in chapter 4. Perhaps that model is easier to organise than the open smart farming model that asks for higher upfront, common investments with a business model for data exchange platforms that have a bit of an infrastructure / utility character. We come back to this issue below.


Digitalisation in agriculture will affect markets. For example technology lowers the costs of storing, sharing and analysing data. Lower transaction costs mean that new suppliers could come to the market (see Airbnb that makes rooms of home owners available for tourists or Uber that makes it possible for car owners to become taxi-driver). But it could also bring new consumers to the market as a global consumer-base becomes available. We have already seen that digitalisation could differentiate products with credence attributes (e.g. by adding data like in organic farming or fair trade or PDO). Or it could add services to a product. This could also lead to very specialised products for small markets (the so-called ‘long tail effect’). ICT can also change the market place itself, as in the auction example where the brick and mortar auction becomes a virtual one. Institutions in the market place can also change, like the introduction of dynamic pricing (pricing dependent on the characteristics of the buyer or the moment of buying). Last but not least the government faces changes in the way it interferes in in markets: regulation can be out of date and on the other hand ICT can provide new options to fine tune policy – parts of the current Common Agricultural Policy with its field maps to monitor greening would not be possible without ICT. Figure 9 illustrates this framework.

Changes in the markets due to ICT
Source: LEI, Wageningen UR

We use this framework to look into more detail to the (potential) effects of ICT on markets.

Existing product markets

The examples given in the introduction above already illustrate how markets for products change. ICT differentiates products, at least in the short time: machines with IoT sensors and machines without, dairy products where ICT and data help to substantiate a sustainability claim and those not, etc. The changes in product specification are at the moment especially strong in machines and installations bought by farmers. Sensors and other internet of things technology improve their functionality and lead to predictive maintenance. There are not yet many examples of such products that change completely into a service: e.g. by adding inventory management advice to installations (e.g. using IoT on climate control equipment to take over the management of the warehouse with potatoes from a remote control centre). Either the technology is not far enough (or already self-regulating) or such a service oriented business model is very different from the current one for a supplier from the sales of machinery, and includes new risks.

ICT can have important effects on the organisation of the value chain. An aspect not yet discussed is that sometimes value chains are shortened as the function of middleman disappears. Travel agencies and bookshops know this all too well. Another aspect is that new types of pricing and contracts become the standard. We mentioned already online-auctions and dynamic pricing. In the insurance market (including agricultural insurance) and banking risk profiling can be improved, even to the point that it raises ethical issues (in life insurance for instance). An example for better credit risk assessment in agricultural banking is the use of (open) data to calculate a specific benchmark for an individual farm that a credit officer can use in her decision to make an offer for a loan.

In existing markets, the government faces the question if its regulation is still up to date, given new ICT options. The fact that markets change, could lead to new policy challenges or make current regulation outdated – we discuss this in more detail in the last chapter of this paper. It could also lead to new technical options for current regulations. There are several examples in the Common Agricultural Policy. As data is recorded and produced in the frame of precision agriculture (for instance geo-referenced data such as soil characteristics, crop status at land parcel level) these data can (and are more and more often) required for policy monitoring in the Integrated Administration and Control System (IACS). This system supports different administrative and control procedures (Zarco-Tejada et al., 2014): farmers’ declaration document; administrative documents; objective evidence of compliance with legislation. Making such data obligatory also stimulates precision agriculture. Such data are also used for surveys such as the Farm Structure Survey (FSS) or the Land Use and Cover Area frame statistical Survey (LUCAS).

Such data transfer also raises issues to which extent governments should also adopt industry standards (or develop them together with industry) on product coding, definitions of fields and crops and to which extent data sharing platforms should be used that are also used by the industry and its compliance or certification bodies for private standards.

Questions of industry standards are not trivial as they are very much linked with the important issue of administrative burden. A lot of organisations including the government tend to use their own definitions: is a sow a sow once it is at a certain age, or once it gives birth to piglets, or already when inseminated irrespective if this leads to live piglets? – It can make sense for nutritionists, veterinaries, housing specialists or fiscal accountants to differ in such definitions. But the farmer can only use one and he seldom has a big say in such data standards. Something similar holds for websites and other forms to provide data. For all organisations it is the easiest solution to build its own website where a farmer is requested to punch in the data. But a lot of data has to be provided to different parties and farm management systems seldom serve all these destinations with a simple click.

Another issue that governments face in such situations is the property right issue: who owns this administrative data and should it be open data, at least at aggregated level (Zarco-Tejada et al., 2014). Some EU-countries like The Netherlands and United Kingdom have opened up their field data to the public, implying that everybody can see which crops have been grown on a certain field in the last five years. Like other open data this can create new services, but it can also be controversial.

Market for new products and services

The difference between existing products and new products is a gradual one. But at a certain stage so much smart technology has been added to a product like a tractor that it becomes a new technology with new services. Porter and Heppelmann (2014) gave a nice and much quoted example in the Harvard Business Review of the tractor and how new technology make it a new product with changing industry boundaries.

Source: Porter and Heppelmann, 2014. Reprinted with permission

They demonstrate that smart, connected products require a new technology infrastructure, including a network communications to support connectivity, a product cloud (containing the product-data database, a platform for building software applications, a rules engine and analytics platform), and smart product applications. That technology enables the collection, analysis, and sharing of the large amount of data generated. The smart connected products will expand the (competitive) boundaries of an industry itself and so transforming competition. Smart connected farm equipment such as tractors, tillers and planters will enable better overall equipment performance. The tractor manufacture expands from tractor manufacturing to offering a package of connected equipment and related services that optimize results. So a tractor company finds itself competing in a broader farm automation industry. However industry boundaries are expanding even beyond product systems to systems of systems. In precision agriculture not only farm machinery but also irrigation systems, soil and nutrient sources are connected with data on weather, crop prices, and commodity futures to optimize farm performance.

The emergence of systems of systems raises the question whether a company should seek to provide the platform that connects the related products and data. Or a company can provide open connectivity to related products produced by others. We elaborate that point in the next section.

The ICT revolution brings not only redefined products that cross industry boundaries but also new products: smart phones, apps, drones, milking robots etc. This brings new suppliers with new products in these agricultural input markets. The current market for agricultural equipment and business solutions is still dominated by traditional companies (agricultural original equipment manufacturers and suppliers) but due to the development and adoption of information technology as data analytics and software solutions new, non-traditional players are increasingly entering the market and strengthening their market position. The main market players for precision farming technologies can be segmented into eight categories: traditional agricultural original equipment manufacturers, traditional suppliers, seed companies, large global IT infrastructure providers, high-tech solutions providers (drones, sensors, control systems etc.), start-ups developing smart devices and apps, investment funds/traders, and universities and research centres (Roland Berger, 2015). These crop protection and seed companies, equipment companies, fertilizer companies, retail distributors, and start-ups, are now competing to provide the best precision equipment and digital services. Some companies have already developed an integrated offering of equipment and services for farmers, mainly in the U.S. Corn Belt. Digital start-ups offer only a portion of the full suite of equipment and services that farmers need. The absence of integrated offerings for the overall market creates an opportunity for large companies with more financial resources, whether they are producers of seed, fertilizer, crop protection, or equipment. These companies can gradually build a compelling one-stop solution that will allow them to compete for the lion’s share of the market (Corsini et al., 2015).

The government faces several questions in these markets. There can be issues on regulating the use of these products: the commercial use of drones is a clear example (Van der Wal et al., 2015), where governments are forced to rethink current aviation rules. And there is the question of such innovation should be supported and how that should be done, including the provision of infrastructure like broadband or promoting adoption by setting industry standards (as the EU did in the very different cases of GSM for mobile telephony and organic farming). This could help to create markets.

New software and data markets

Besides attention to new physical products as discussed in the previous section, there is a reason to pay special attention to markets for software and data, with apps (applications) and platforms as special categories of software. If data, apps and platforms are becoming so important, can markets for them be created and how can we value them?

We showed already that one of the main challenges is how to organise the data exchange, in a more integrated value chain and certainly in a more open way. Creating value out of data is based on the combination of data from different sources and on the aggregation of data. And the one that gathers the data is not always the organisation that brings out the optimal use of the data. To give some examples: data from soil analysis, crop monitoring and weather forecast have to be combined with data on pesticide recipes to generate a real time spraying advise and to instruct the spraying machine accordingly for automatic filling. The disease pressure of phytophthora in an area can be measured by aggregating the observations from individual fields of different farmers. Data from online sales of manure or fertilizers have to be shared with the accountant, the inspection services of retail standards like GlobalGap and the government for environmental legislation. Banks need access to accounting data of a farm and to benchmarks. This means that there is a need for Agri-Business Collaboration and Data Exchange Facilities (ABCDEFs) (Poppe et al., 2015).
Industry is trying to solve this bottleneck of data sharing by setting up data platforms that perform such a function. This is often done on a company basis (e.g. This supports a strategy of companies to become more service oriented and earn money from the added value of data. Such platforms develop into eco-systems of applications (apps, as in smart phone apps, but here also used for machine to machine applications) that interact with each other (e.g. 365FarmNet).

As such this works fine, especially if a food chain is very integrated with one dominant company. The Dutch veal industry is probably an example: farmers are under contract with a slaughterhouse that also provides calves and feed as inputs for farmers and provides (e-)tools for administration and management. The integrating slaughterhouse can create a platform where all apps work in an integrated way based on company data standards and data event processing (pushing data from one app to another). In the machinery industry something similar applies if a company (e.g. John Deere) supplies all the machinery to a client (which is in reality not the case, farmers also buy machinery from competing companies and use contractors with machinery, making support for fleet management difficult).

However many European food chains base their strength on specialisation and are not integrated. For instance in dairy there is a need for data exchange between the machines and robots (machinery industry), breeding, feed companies and dairy companies around the farmer, as well as the service industry like veterinaries, inspection services, accountants and the government (environmental issues, agricultural policy, food safety). In this system the data management is traditionally done by the cooperatives and other organisations around the farm: it are the food processors who send an invoice to the farmer on what they bought from them, as that is more efficient than the farmer invoicing his own sales.

In some European countries the industry has until now tackled the management of these data flows in a digital world by investing in EDI-standards based on reference data models, created by industry standard organisations. For international use these standards are incorporated in the UNCEFACT system and collaboration is currently sought from Dutch and German standard organisations with the American standard organisation AgGateway to promote the use in the rest of Europe.

These EDI standards are very important and have to be enlarged to handle IoT data, but that is only the first challenge. A second challenge is to support the actual data exchange between the apps. With the multiplication of data sources and the fragmentation of software into apps (see below), the transaction costs of linking all relevant data points with each other (one to one) becomes cumbersome. A medium sized Dutch agricultural machine company (Kverneland) recently realised that it had to connect the IoT data stream from its machines with over 50 different farm management software programs that its European clients use. Just making that data available with its own or a reference standard in the cloud will not do, as the farm management software packages probably have to link to 40 different machine companies – it looks attractive to replace 50*40=2000 links by one clearing house / hub).

Even if direct links would work (which is doubtful), there is an important third challenge: the market for services in the form of apps. Some farm management software try to develop apps into complete dashboards, but in many cases there is a fragmentation into individual apps, like on the smart phone: an app for weather advice, an app for spraying sugar beet, an app for transport planning etc. Many of the data-platforms mentioned above support this development (e.g. 365 FarmNet, Akkerweb).

This makes software development easier and specialists can work on a specific service in the form of an app. Many of these apps are commissioned by the platform and paid by them; however there are also apps that have a more independent business model. It could be in the interest of users (farmers) and data-platform operators to have a certain competition between apps / services on quality and price and be able to select the most appropriate app for e.g. a spraying advice on phytophthora. And in the (many) cases were a farmer uses more than one data-platform (e.g. of a machine company and of a food processor) it is certainly in his interest to pay only once for e.g. a weather advisory service that can be used in different apps on the different platforms he uses. A situation where he pays for a precise weather forecast from the Dutch meteorological institute KNMI via an app or data platform of a machinery company (directly or indirectly via the price of his machines) and a similar advice from the private weather forecast company Meteo Consult via an app of the sugar beet company to which he delivers his produce is to be prevented. In eco systems of apps, this challenge should be tackled.
A fourth challenge is in the governance of the data exchange. According to the FAIRport conditions data must be findable, accessible, interoperable and reusable, but in reality this is hardly the case. In many cases the ownership of data is quite unclear. Legally ownership of data is hardly a concept, it is mainly based on privacy regulations. New concepts like the right to be forgotten, are introduced. Sometimes ‘primary data’ are seen as owned by the farmer, ‘computed data’ as being owned by the one who did the computing. There is clearly a commercial battle going on as data are seen as a strategic asset, having a value.

Most manufacturers of agricultural machines (tractors, equipment, milk robots etc.) use technological measures, such as passwords or encryption, to protect competitors and third party's from copying, tampering or pirating valuable, reliable software code that controls the vehicle, provides safe operation in accordance with safety standards, and complies with applicable emission regulations. In the U.S., the Digital Millennium Copyright Act generally prohibits circumvention of technological measures to gain access to works protected by copyright. However, the Copyright Office can grant exemptions from this prohibition pursuant to a public rule-making proceeding.

In 2015 the Copyright Office issued a two year exemption for Class 21 under 17 USC § 1201(a)(1) for the vehicle owner (acting alone) to circumvent technological measures that protect vehicle software for repair, diagnosis and lawful modification in limited circumstances. The exemption excludes the owner (or others) from circumventing security measures for: (1) computer programs for control of telematics, (2) computer programs for entertainment systems, (3) circumventions that violate Department of Transportation regulations, (4) circumventions that violate Environmental Protection Agency regulations, or (5) circumventions that violate applicable law, among other things.

The European Union issued Directive 2001/20/EC on May 22, 2001, on Harmonisation of Certain Aspects of Copyright and Related Rights in the Information Society, ("2001 EU Copyright Directive") to enable EU members to implement the WIPO Copyright Treaty with anti-circumvention provisions analogous to the Digital Millennium Copyright Act in the United States. (2001 EU Copyright Directive, available at In contrast to the U.S., the member states of the EU generally do not have regular rule-making procedures on exemptions to the anti-circumvention provisions. 17 U.S.C. § 1201(a)(1)(B)-(E).

Instead, most EU nations appear to have few generally static legislatively-approved exceptions to the anti-circumvention measures that were passed through each of those nation's full legislative or parliamentary process. See, e.g., Act on Copyright and Related Rights (Copyright Act), Article 69(e), Decompilation and Article 95(b) Measures in Respect of Limitations (Germany).

In the U.S. to modify legally the software under the above Class 21 exemption, the vehicle owner might pursue a fair use justification or under the section 117 exception (e.g., for machine maintenance and repair) of the Copyright Act. Under currently existing copyright law and hitherto, the sales of a computer or a vehicle does not transfer ownership of the copyright in the vehicle software to the purchaser of the computer or vehicle. For example, the purchaser of the vehicle does not generally receive the right to copy, modify or distribute the vehicle software, unless authorized under contract or applicable law. Newer equipment of some manufacturers, like John Deere, supports subscriptions to quick, convenient and professional vehicle software updates wirelessly or through a direction connection to a diagnostic port to minimize downtime of vehicles for critical agricultural tasks. Seen these technical developments of being able to update software, the need to have the rights (and be able) to change software in such complex machines like tractors, combine harvesters and milking robots, has not turned out to be a big issue, although this was suggested a few years ago by some. Farmers seem to be more concerned about the flow of data between different software (either at the same time between e.g. software from machines and farm management software, or over time when they change to another brand) than the question to which extent they own the software of their machinery.

Also for algorithms in apps IPR-rules are relevant. In these commercial battles it is not so clear what the scope of the different platforms are, and they are changing: the data platform of a machinery company tends to start with apps on its own machinery or fleet management integrating with other machines. But next is the integration with the input industry, and that brings integration with e.g. farm management software or inspection services. It is on these borders where important innovations can take place but where there is confusion on scope and commercial interests (that differ between the supplying industry, food processors and the farmers themselves).

The expectations of the value of big data have also raised the idea that this value can be allocated in some way to the underlying smaller data sets that contribute to the big data insights. However it is unclear how this can be allocated, and if a reward systems or a market for data could work anyway. The unclear situation on data ownership, together with the high expectations on an unclear value of data hamper the governance of data platforms and data exchange between platforms. Should the data exchange be organised in a commercial way by an independent (ICT) company (like SAP and F4F are doing with SAP’s Hannah software: the Facebook solution), or in a non-profit way as a kind of cooperative organisation between platforms (building on experiences with e.g. standard organisations like AgroConnect, service suppliers like GS1 or EDI-circle and open source software development like LINUX) or is there even a need for a public infrastructure (similar to highways that supported the car industry to move people around). And what is the position of farmers in this game: should they or their farm organisations invest?

For the European machinery, input and food industry it is of utmost importance that this issue of agri-business data exchange is solved. It is of more importance here than in the USA, seen how we have organised our food chains with maximum specialisation between the layers instead of integration, with many specialised products in specific value chains (PDO, organic, specialized products instead of big commodities) and high demands on data for tracing and tracking, food safety and sustainability schemes. The international orientation of the machinery industry as well as the multi-national presence of many of the food processing companies (including cooperatives) are another reason of being confronted with this need to link data from different apps and platforms at a European scale.

The FI-PPP project FIspace (commissioned by DG Connect) designed a solution that could create a business collaboration service based on open source software that would link the different platforms. Farmers could be users of different platforms and communicate between them and with others as e-mail can be exchanged between different mail-programs or websites can be looked at with different browsers. This solution would create a specific layer between the EDI-standards developed by standardisation organisations and certified by UNCEFACT and the commercial platforms.
Such a solution would support the market for platforms that is now an emerging business (often connected to the machinery or input industry), the market for apps, and perhaps even create a market for farmers’ data. Figure 12 provides the value proposition for such a service.

Integrating different ICT-platforms (Eco-systems of apps) with a business collaboration service
Source: FI-PPP project FIspace
Value proposition for a collaboration service that would support the markets for platforms, apps and data

Platforms tend to have high network effects that could turn them into a natural monopoly. They work like a telephone network or a social media service (e.g. Facebook): people have an incentive to join the network that other persons use as this maximizes options for communication with others. Platforms can also create lock-in effects: they want to keep their clients and don’t have an incentive to make it easy to leave and take your data with you. If you lose your own information when you leave, there is incentive to stay. Both effects can make platforms very profitable, and raise competition issues.

Governments are confronted with another issue concerning such platforms. Farmers do no only exchange data in the food chain with input suppliers and food processors, but also with the government, for instance in the CAP (but also in animal registration etc.). Paying agencies and other government agencies can make beneficial use of digital data for control purposes. It would help if they use the same data standards and data exchange platforms as are used in the food chains, to reduce the administrative burden (‘simplification of the CAP’). Another aspect is the role of open data, to trigger innovation. For this reason some paying agencies make field data available to the public. A third aspect is that farmers and other stakeholders could share data in the European Innovation Partnership (EIP) Agri framework of interactive innovation with advisory services and researchers. E-Science implies a shift from data collection and research on one experimental farm resulting in a general advice towards using data from many farms resulting in specific advises for individual farms (EU AKIS, 2015).

Data governance and ownership

Issues at institutional levels

Digitization in agriculture and agri-food sector provides institutional changes. Williamson (2000) distinguishes four levels of institutions as shown in the diagram below.

Digitalisation of agriculture on four levels of institutions
Source: Adapted from Williams (2000)

According to Williamson property rights change much slower (e.g. over a period of 10-100 years on average), but they determine contractual relations, that change faster (e.g. on average in a period between 1 and 10 years) and those contract determine daily trade. The influence is not only from culture (norms and values, e.g. regarding types of meat that can be eaten) to property rights and further down, but also the other way around: if something loses its scarcity (e.g. options to broadcast radio or television) contracts and property rights change (e.g. commercial television instead of the national public channel).

The ICT-revolution in agriculture will continue for some time. The speed of technological developments and digitization in agriculture is much greater than the speed with which institutions change: old institutions can block uptake of ICT and ICT use can lead to new types of property rights (like the right to be forgotten). As such changes take place at different speeds, discussions and frictions are to be expected: see the examples of Airbnb and Uber are sometimes blamed to have a business model that is not legally acceptable. Through changes in property rights and transaction costs, the effects of the ICT-revolution in agriculture could be larger in the coming years than they were until now.

Data ownership

Data collected in precision farming offer opportunities to enhance production but it is not always clear who owns data generated through precision farming: farmers, the agronomists or service providers that create data, or machinery and software producers. Farmers in the US have already voiced concerns about the unregulated use of data from their businesses (POST, 2015).

The ownership of data sets forms another critical aspect related to the massive amount of data that is gathered from sensors that are distributed across the farm. This has implications for data-sharing schemes and privacy issues. Farmers tend to be unwilling to share data about their operations with third parties, which poses serious problems regarding efficient data integration (EIP-AGRI, 2015). There are different aspects in this “ownership” discussion. One is that farmers see some data as privacy sensitive or a business secret that they do not want to see revealed to competitors or (e.g. in case of environmental data) to the public or the government. Agri-businesses (including e.g. equipment manufacturers) sometimes state in their contracts that they treat collected data as personal, offer options to reduce the data flow and honor requests to remove data from their servers. Another aspect is if the farmer has access to the data himself now, and can combine it with other data on his farm, e.g. in his own farm management dashboard, as well as in the future when he changes suppliers.
From a legal perspective there are no rules regulating ownership of a data set. There are no property rights in data as you can only have property rights in a tangible good. Data are also not protected by intellectual property rights, like books, movies and software are protected by copyright against copying (Moerel, 2014). However there are other legal rules that can apply and determine whether data can be used for certain purposes and whether data can be transferred to another party. Non-public data can be protected: (1) under applicable privacy rules (if the data contains personal information), (2) as confidential information (e.g., trade secrets) in many jurisdictions (if the information is not publicly available), or (3) under database rights, which are applicable under EU Directive 91/250/EEC.  The U.S does not have similar protection of database rights.  However, if the data is disclosed (e.g., reported) to a government regulator, publicly disclosed or published, the protection as confidential information may be unavailable or lost. This all means that (at least for the moment) contract law between farms and other businesses determine if a farmer feels himself “owner” of the data gathered on his farm and can use that in other applications. New business models for data management are needed; sharing and open-data sources should be developed to bring Precision Farming to the next level. The recognition of data ownership is crucial. Portals that can facilitate the exchange of data are a prerequisite (EIP-AGRI, 2015).


From 2010 through 2014 there were 5,337 new patent registrations worldwide relating to precision and conventional equipment for agriculture: various sensors, variable-rate/depth components, connectivity of sensors and equipment, automated applications for dairy farms, autonomous vehicles, precision harvesters and mowers, sensors and components for autonomous driving, equipment components, and tractor components. 70 per cent of those new agriculture patents are assigned to North America (the location of the filing company’s main headquarters), 15 per cent in Europe, 8 per cent in China and 7 per cent in other regions in the world (Corsini et al., 2015).
One of the main restrictions for data sharing among institutions, farmers, advisers and researchers is due to non-standard software and data formatting solutions. The challenge is to enable data sets to be shared easily irrespective of the sensor model and brand used. As modern farms are increasingly loaded with all kind of sensors, data management, data storage, data sharing and interconnectivity strategies are urgently needed (EIP-AGRI, 2015).

Two scenarios

n general effects of technologies on society are being overrated on the short term and are being underrated on the long term. With this in mind we envision the possible effects of the digitalisation in agriculture and the agrifood sector on governance issues by applying two future scenarios developed by EU SCAR (2015). One is the High Tech scenario, in which the power and decision making takes place at a higher level away from farmers and the value chain is organised along a Captive model (see chapter 4). The alternative is the Self-organisation scenario, in which farmers, citizens and decentralised governments take the lead and the Modular model of chain management prevails.

High Tech Scenario

The High Tech scenario assumes a world dominated by large multinationals and advanced technology (ICT, robotics, genetics). It is characterised by globalisation, widespread use of unmanned vehicles, contract farming and outsourcing, with a large urban population. European institutions are strong, national governments are weak. In general it is a wealthy society, but inequality creates concern. Sustainability problems are largely solved through technical solutions such as precision farming and genetic modification (EU SCAR, 2015). In this scenario governments are faced with monopolies, privacy and exclusion issues.
An example is John Deere’s Big Data pact with Monsanto. John Deere has granted Monsanto exclusive access to data of its systems that can be used for the Climate Field View, the management platform of Monsanto that farmers use to measure the growth of their crops and the soil conditions. Monsanto is now able to offer their customers a better link to data systems of John Deere.
Possible effects of this scenario are:

  • Scaling-up: large agricultural companies arise from mergers; they are closing contracts with multinationals in supply, processing and retail. NGO’s are one of the checks and balances and they force multinationals to have farmers working in a sustainable way, using such technology.
  • Ownership relations: data owned by multinationals, farmers towards a franchise model, exclusion.
  • The role of national governments decreases. Agreements must be made at the international level (WTO, TTIP) by international operating enterprises. The EU represents European countries.
Self-Organisation Scenario

In the Self-organisation scenario a Europe of regions exists with bottom-up democracy where new ICT technologies with disruptive business models undermine large companies, high-tech and traditional craftsmanship that leads to self-organisation, bottom-up democracy, short-supply chains, and multi-forms of agriculture. European institutions are weak, regions and cities rule and follow quite different pathways for agriculture. Products are traded between regions. There is inequality between regions, depending on endowments. In this scenario the government has to deal with the question how to arrange and guarantee public interests when facing institutional changes. Possible effects of this scenario are:

  • Scale: collaboration at regional scale arises, added value is more than just money but also experience, origin of food products.
  • Ownership relations: farmers remain owner of the data, farmers organize themselves in cooperatives that close contracts; new business models emerge.
  • Sustainability: the availability of knowledge on high-tech systems improves the resource efficiency and decreases the pressure on the environment of agricultural firms; agreements between companies, citizens, NGOs and regional or local government on goals, monitoring and control.

Data Management

The future ownership of data is a major policy concern.

The clear main policy concern identified by the experts stems from the insight that the future of PA will probably be dominated by data exchange, and that platforms will be used for this data exchange. In this development, those who own the data can direct and control the data sets, are in the central position of power, and create the added value and earn a major share of income generated in agriculture. Thus, the most critical issue for the future of PA and farming in Europe lies in future ownership of data and control of these platforms, and, secondarily, in issues concerning privacy. These issues are relevant in every scenario. In 'Scenario 1 – Economic Optimism', big companies are in charge of the data; in 'Scenario 2 – Global Sustainable Development' it is the government; in 'Scenario 3 – Regional Competition', local governments may not own the data, but at least have access to all of the data; and in 'Scenario 4 – Regional Sustainable Development', people and businesses own their data, but also share data easily. This topic was clearly the strongest worry as it concerns power shifts in the sector, and it is listed as the top priority for policy and legislation. It was also stressed by the experts that the specific context of European farming plays a role here: European agriculture is characterised by diversified farming with many high quality products, the value of which depends strongly on data (from food safety, tracing and tracking to brands, organic food, etc.). In addition, Europe has innovative, highly skilled farmers, and a large and leading specialised machinery industry. These characteristics and strengths combined with existing initiatives on e.g. pushing digitalisation in Europe provide a competitive starting point. At the same time, the pressure from developments in Silicon Valley or other leading high-tech regions means that a strong effort is needed in order to ensure that 'control over data' from the European agricultural sector does not lie increasingly outside of Europe.


Precision agriculture and the digitalisation of agriculture will have effects on the organisation and functioning of markets for products. Not only on existing markets for products, but also on markets for new products and services, and markets for new software and data. This concerns price, volume, transaction costs, market structure and behaviour (monopolies, regional exclusion), level of innovation, privacy issues in data, system risks (cyber security) etc. These market outcomes can be assessed with regulatory impact assessments, or societal cost/benefit analysis. Leading, if outcomes are undesirable, to market regulation or to other government interventions (e.g. training in education). This implies that all these issues could have implications for different EU policy domains.

Source: LEI, Wageningen UR

Agriculture serves several interests e.g. interests related to the production of crops and the provision of employment. Agriculture has various public interests which demand for a kind of government interference: production value and employment of agriculture, food security, food safety and animal health, animal welfare, environment and climate, nature and biodiversity, water management, liveable rural area (regional development). Some of these public interests are not automatically guaranteed by the market and ask for government intervention. The digitalisation of agriculture serves some interests such as food safety and less environmental pressure. The (negative) external effects in agriculture can be reduced by ICT in a way that could be more attractive than by regulation. This invites governments to take an active role in promoting ICT. However, ICT also has its negative aspects, the effect on employment. ICT could help the competitive position of the machinery industry, but it could also destroy jobs in rural areas. And there are new issues that could be put on the plate of the government like privacy of farmers, ownership of data, and shift of power balance in the chain.

Against this background we discuss the most important issues from this report for relevant EU policies.