PowerPoint Presentation

[Audio] Welcome to module 6 "Case studies from the field".

[Audio] In the previous modules we built a toolkit of concepts. In this module we use it on real cases. Cases are valuable because they show how technology, data, business models and policy interact in a concrete setting. But a case can also mislead if we read it uncritically, so let me first suggest a method. I suggest six lenses, which map to what we have learned. First, the problem and the actors: who has what problem, and who is involved? Second, the stage of the value chain and the technology used, as in Module 2. Third, the data flows and their governance: who generates the data, who can use it, under which rules, as in Module 3. Fourth, the ecosystem: who orchestrates, who complements, which interfaces matter, as in Modules 1 and 4. Fifth, the business model: who pays and who benefits. And sixth, the evidence: what exactly was measured, against which baseline, over what period, and who reported it. That last lens deserves emphasis. Many digital agriculture results are reported by the project or company that developed the technology, often as headline percentages. Such figures are useful, but they are not the same as independent evidence. We will see an example of the difference later in this module. We will go through five cases. Case one is precision crop management, using results from the large European IoF2020 project and comparing them with independent reviews. Case two is traceability, contrasting a retailer-led blockchain pilot with a Europe-wide assessment of blockchain projects. Case three is interoperability in practice: a Belgian potato harvester manufacturer, a Flemish data platform, a French farm software and a Finnish platform, working together. Case four is the DIVINE project, which tested the value of data sharing in four pilots. And case five is about people: an advisory platform developed in Greece, a European training project on blockchain skills, and the TALLHEDA activities that connect them. As we go, I invite you to keep your product example from Module 1 and your service from Modules 4 and 5 in mind. At the end of the module, we will compare the cases and draw out lessons for your own context..

[Audio] Our first case is precision crop management, and the source is the Internet of Food and Farm 2020 project, IoF2020, which we met in Module 4. It had a consortium of seventy-three partners and thirty-three use cases across five areas: arable crops, dairy, fruit, meat and vegetables, with the aim of demonstrating Internet of Things solutions in operational conditions. Two arable use cases give us numbers. In a comprehensive review of the economic and environmental benefits of digital agricultural technologies, published in 2024 by Papadopoulos and colleagues in Smart Agricultural Technology, which synthesised data from 136 peer-reviewed papers and 28 documents with empirical data from EU projects, the IoF2020 within-field management zoning use case in the Baltics is reported as achieving a 22 to 30 percent reduction in nitrogen fertiliser use and a 43 to 53 percent reduction in herbicide used for haulm killing in potatoes, with a 2 percent yield increase. The precision crop management use case, with IoT sensors and satellite data, is reported to have achieved a 5 percent reduction in nitrogen application and a 5 percent reduction in labour duration. These are impressive results, and they illustrate what site-specific management can do when it is well designed. Now apply the sixth lens: evidence and limits. How do these results compare with the wider literature? A 2026 systematic review of the economics of precision crop production, in the journal Agriculture, found that precision technologies achieve average input savings of 8 to 20 percent and yield increases of 2 to 6 percent, with reported returns on investment typically between 5 and 15 percent. For variable-rate fertilisation specifically, savings of 7 to 12 percent are typical. The same review warns that more complex strategies do not automatically give better economic outcomes: in one study, a sensor-integrated strategy led to excessive nitrogen savings and yield losses of 3 to 5 percent in some years. So the extreme values in project reports, such as savings above 40 percent, should be read as results for specific fields, seasons and baselines, which may include inefficient starting practices. They show potential, not an average expectation. Meanwhile, the Papadopoulos review also found that farm management information systems improved yield by 10 to 15 percent in the studies reviewed, and variable-rate technologies saved 20 to 50 percent of water in vineyards and pear orchards, which again shows the wide range depending on context. The lesson is twofold. For farmers and advisors: look for evidence from conditions similar to yours, and calculate the economics of your own farm. For educators: teach students to read effect sizes critically and to ask about baselines. For discussion: what would you want to know about a farm before predicting its savings from precision nitrogen management?.

[Audio] Our second case is traceability, which we discussed in Module 2. Let me present two contrasting pieces of evidence. The first is the best-known example worldwide: the pilots by Walmart and IBM on blockchain-based food traceability. Walmart's food safety lead tested how long it took his team to trace a package of sliced mangoes back to its farm, and it took seven days. With a blockchain solution built on Hyperledger Fabric, the same trace took 2.2 seconds. A parallel pilot traced pork in China. By September 2018, according to published accounts, Walmart was tracing over twenty-five products from five suppliers, and it later required suppliers of leafy green vegetables to upload their data to the blockchain by September 2019. Apply the ecosystem lens. This was a retailer-led ecosystem with a very strong orchestrator. Walmart had the market power to oblige suppliers to join and to share data. That is why it could achieve a critical mass quickly. Most agri-food chains in Europe, with many small producers and no dominant orchestrator, do not have such a lever. Also note that the headline result compares a manual, paper-based trace across many partners with a purpose-built digital system, so the speed-up reflects digitalisation in general and not blockchain alone. The second piece of evidence is a European reality check. The TRUSTyFOOD project, coordinated by Fraunhofer IML, evaluated studies and projects and found around 320 blockchain projects in the agri-food sector, only a few of which are moving into full operation. It identifies challenges such as complex supply chains, heterogeneous data silos, and the emergence of isolated "blockchain islands": systems that do not connect with each other. Its approach is to build a European roadmap and to focus on use cases with a clear value proposition for supply chain managers. Now the legal context from Module 2. The EU requirement of one step back, one step forward means that every operator already must know its supplier and customer. Digital systems extend this, but they depend on the same fundamentals: shared identifiers, consistent data, and trust that what is entered is true. A blockchain can preserve a record, but it cannot check it. The lessons for our topic are clear. Technology alone does not create a traceability ecosystem. What matters is who orchestrates, whether participants share standards, whether data quality is controlled at the source, and whether the benefits, such as faster recalls, access to premium markets, or compliance with rules like the Deforestation Regulation, justify the cost for each participant. Interoperability between systems is the next challenge, which leads to our third case. For discussion: in your product chain, who could play the role of orchestrator for traceability, and what would make smaller actors willing to join?.

[Audio] Our third case is a very practical illustration of the interoperability issues from Module 3. The starting point is the Belgian company AVR, which makes potato harvesters. During harvest, its machines collect data about the field, potato variety, yield, soil conditions and field operations. But AVR sells machines worldwide, to farmers and contractors who use different software packages, often connected to their own regional data-sharing platform. Imagine a Finnish farmer with an AVR harvester who wants to see the harvest data in the Finnish farm management system. Before the digital bridge, according to AVR's own account, the company had to contact the other platforms and make additional technical developments, even though it already had an application programming interface connection with the Flemish platform DjustConnect. In extreme cases it had to ask Finnish farmers to register on the Flemish platform. French farmers faced a similar problem: many use the French software MyEasyFarm, but without a bridge to DjustConnect some missed important data from their harvesters. The solution came on two levels. On the technical level, in June 2024, the three platforms, DjustConnect from Flanders, Tritom from Finland and Agdatahub from France, signed a memorandum of understanding to link up. On the organisational and legal level, a Finnish partner, 1001 Lakes, helped draft the Potato-X rulebook, an agreement establishing a common framework for the network. A central element of this rulebook is farmer permissioning: all data sharing for data created by the farmer is subject to a consent procedure that the farmer fully controls. Later, MyEasyFarm announced an integration with AVR and ILVO through DjustConnect. Potato farmers can now automatically exchange field boundaries, detailed yield maps and technical crop indicators between AVR harvesters and the MyEasyFarm platform, and data can only be shared with the explicit consent of the farmer. Let us read it through our lenses. Interoperability: this case touches all four layers we described, technical, with the interfaces; semantic, with the common data descriptions; organisational, with the rulebook; and legal, with consent and data sovereignty. Ecosystem: a manufacturer, a public research institute operating a neutral platform, software companies, and farmers' consent as the gate. Business model: farmers can use DjustConnect for free, the manufacturer saves repeated integration work, and the software company gains access to better data. ILVO also reports that farmers received on average nine data-sharing requests on their dashboard, and that new options are being developed, such as group data management and data vaults where sensitive data stays on the farm. The lesson is that interoperability has a business value for every actor, but it needs neutral infrastructure, shared rules and the trust of farmers..

[Audio] Our fourth case returns to DIVINE, the Horizon Europe project we introduced in Module 3 to test the cost-benefit of agricultural data sharing. It ran from October 2022 to January 2026, with four real-world pilots. In Slovenia, the pilot led by the Innovation Technology Cluster and the Chamber of Agriculture and Forestry in Murska Sobota developed benchmarking tools for dairy and pork farming. It connected fourteen public and private data sources, and thirty dairy farmers and eight pork farmers gained three years of insights into the economic and sustainability performance of their farms, with automated data entry and improved advisory support. In Greece, the pilot led by NEUROPUBLIC combined 88 digital farm calendars and ten IoT agro-environmental stations, connected farm management data via interfaces and semantic translators, and automated the calculation of CAP indicators, which links to what we discussed about the CAP in Module 5. Twenty-one farmers received customised data-driven advice. In Spain, the pilot integrated data from over a hundred weather stations, sensors, farm management software, field notebooks and imagery for olive and almond production. In Ireland, a 56-hectare cereal pilot tested three decision support tools. Now apply the evidence lens. The project reports indicative results for the Greek pilot: for grapes, 102 percent more production over three years, 55 percent across two pilot rounds, 16 percent less irrigation and 46 percent less pesticide use; for olives, 120 percent more production across two pilot rounds, 24 percent less irrigation, 15 percent less fertilisation and 37 percent less pesticide use. The Spanish pilot reports overall reductions of environmental impact indicators between 21 and 30 percent. These are large numbers, much larger than the 8 to 20 percent average input savings and 2 to 6 percent yield gains found in the independent review we saw in Case 1. The news item does not say what the baseline was, whether there was a control group, how many farms contributed to each figure, or whether weather differences between years were taken into account. That does not mean the results are wrong. It means that, before quoting them, you should read the full pilot case study, which the project publishes, and look at the design. It is an excellent exercise for students. What stands out for our purposes is the process: DIVINE reports that the Greek pilot faced data quality issues, lack of interoperability, farm management software lock-in and farmer reluctance to share data, which are precisely the barriers we discussed in Module 3. And its pilots show that value arises when data from different sources is combined, for example into benchmarks or CAP indicators, which no single farm could produce alone. For discussion: what questions would you ask the project team before recommending these tools to farmers in your region?.

[Audio] Our last case is about the human side: how digital tools reach advisors and farmers, and how skills are built. I will use three examples, and two of them involve partners of TALLHEDA. The first is agros.ai, a digital farming platform built by Smart Agro Hub with support from the Agricultural University of Athens. It is designed for farmers, agronomists and businesses on one shared platform. For farmers, it offers free use for up to ten parcels and then one euro per additional parcel per year, with import of parcels from the Greek land parcel identification system, a fourteen-day weather forecast, NDVI satellite imagery, crop stage prediction, notes with photos and video, and access to a network of professional agronomists. Its communication feature between farmers and agronomists is described by the developers as the first of its kind with features tailored to agricultural workflows. Read it through our lenses. Business model: a freemium logic, as we discussed in Module 4, in which the entry barrier is very low for small farmers, and the value of the network of advisors and companies grows with use. Data sources: it uses open data, such as satellite imagery and parcel boundaries from public registries, which connects to Modules 2 and 5. Ecosystem: the platform connects the farmer with the advisor, who is the trusted intermediary in AKIS, which research on the digitalisation of agricultural knowledge and advice networks identifies as a key area of change. Please check the current prices and features on the platform's site, as they may change. The second example is TRUST-FOOD, a Digital Europe project that ran from 2023 to 2025 on advanced digital skills in blockchain for trusted food supply chains. It offers twenty short courses in seven languages, including Greek, aimed at reskilling and upskilling employees, particularly in small and medium enterprises, and jobseekers. It covers practical applications of blockchain, such as traceability and smart contracts, and is coordinated by Rezos Brands with Smart Agro Hub leading communication. Remember the TRUSTyFOOD finding that few blockchain projects reach full operation: training is one way to address the capacity gap behind it. The third example is TALLHEDA itself. The project's Digital Tools booklet, published in 2025, informs producers, agronomists, students and researchers about modern digital tools. Its brainstorming visits discussed artificial intelligence and data analytics, blockchain and distributed ledger technology, and digital twins. And staff mobilities bring university researchers to field sites: Smart Agro Hub reported a three-day mobility by researchers of the Agricultural University of Athens on 27 to 29 April 2026, covering remote irrigation control, fertigation units, field measurements via a mobile application, and a live drone flight with multispectral and thermal cameras. The common lesson is that technology adoption depends on trusted intermediaries and skills. A low-cost platform, a network of advisors and practical training form together an ecosystem that individual farmers can plug into..

[Audio] Let us compare the five cases and extract lessons that you can use in teaching, advising and project design. First, value depends on context. In the precision crop management case, projects reported large savings, while independent reviews found more modest averages and showed that complexity does not guarantee better economics. When you read results, ask for the baseline, the sample, the period and the source. Second, the orchestrator matters. In the Walmart case, a dominant retailer could create critical mass by requiring participation. In the DjustConnect case, a neutral public research institute built trust among cooperatives, companies and farmers. In Europe's fragmented chains, neutral and well-governed orchestration is often the realistic path. Third, interoperability needs more than technology. The potato harvester case required platform bridges, a shared rulebook and farmer consent. All four layers, technical, semantic, organisational and legal, were needed. Fourth, data sharing creates value by combining sources, as in the DIVINE benchmarking and CAP monitoring pilots. But data quality, lock-in and farmers' reluctance were the recurring bottlenecks. Fifth, advisors, cooperatives, platforms with low entry prices, and training are part of the ecosystem. Without them, even good technology stays on the shelf. That is the point of projects like TALLHEDA and TRUST-FOOD. Sixth, the funding and business model must outlast the project, a lesson from Module 4 that is confirmed by many of the cases, since several of them were project-financed. Now the exercise. Choose a digital agriculture case from your own region, a real one if possible, or one from the list we saw. Analyse it with the six lenses: problem and actors; chain stage and technology; data flows and governance; ecosystem roles; business model and who pays; and evidence and limits. Write one paragraph for each lens, and finish with two lessons and one question that you would ask the people behind the case. This is a good format for a student assignment or a workshop with advisors. If you are watching live, share one lesson in the chat. If you are watching the recording, pause here for about ten minutes. In the next module, we turn to people and institutions: the skills that this transformation requires, and the role of higher education, which brings us back to TALLHEDA's core mission..

References. Agdatahub / ILVO / DataSpace Europe (2024). Agdatahub networks with ILVO (Flanders) and DataSpace Europe (Finland). https://agdatahub.eu/en/agdatahub-reseau-ilvo-dataspace-europe/ agros (Smart Agro Hub). Digital farming, made simple. https://agros.ai/en DIVINE project. News and pilot overviews (Pilots 1-4); Pilots. https://divine-project.eu/news ; https://divine-project.eu/pilots DjustConnect / ILVO (2024). ILVO and DjustConnect pull off first transnational Data Space. https://www.djustconnect.be/en/ilvo-and-djustconnect-pull-first-transnational-data-space ILVO (2024). ILVO kicks off the first transnational agrifood Data Space. https://ilvo.vlaanderen.be/en/news/first-transnational-agrifood-data-space ILVO (2025). DjustConnect celebrates its fifth birthday with strong growth. https://ilvo.vlaanderen.be/en/news/djustconnect-celebrates-fifth-birthday-with-solid-growth 1001 Lakes. Realizing the value of networked data, Part 1: Potato farming (Potato-X). https://1001lakes.com/realizing-the-value-of-networked-data-part-1-potato-farming/ Fielke, S., Taylor, B. & Jakku, E. (2020). Digitalisation of agricultural knowledge and advice networks: a state-of-the-art review. Agricultural Systems, 180, 102763. https://doi.org/10.1016/j.agsy.2019.102763 Fraunhofer IML. TRUSTyFOOD: stakeholder-driven pathways for blockchain implementation in the agri-food sector. https://www.iml.fraunhofer.de/en/projects/2022/trustyfood-blockchain-technologie-agricultural-food-sector.html IoF2020. Internet of Food and Farm 2020 (CORDIS project 731884). https://cordis.europa.eu/project/id/731884 Kovács, E. & Szőllősi, L. (2026). Economic aspects of precision crop production: a systematic literature review. Agriculture, 16(7), 820. https://doi.org/10.3390/agriculture16070820 MyEasyFarm (2025). Interoperability: MyEasyFarm and AVR via DjustConnect. https://www.myeasyfarm.com/en/category/press-releases/ Papadopoulos, G., Arduini, S., Uyar, H., Psiroukis, V., Kasimati, A. & Fountas, S. (2024). Economic and environmental benefits of digital agricultural technologies in crop production: a review. Smart Agricultural Technology, 8, 100441. https://www.sciencedirect.com/science/article/pii/S2772375524000467 TALLHEDA project (2025). Booklet: Digital Tools. Zenodo. https://zenodo.org/records/17243349 ; 1st and 2nd Brainstorming Visits to the Agricultural University of Athens. https://zenodo.org/records/17242335 ; https://zenodo.org/records/17242918 TRUST-FOOD project (2023-2025). Advanced Digital Skills on Blockchain for Trusted Food Supply Chains (GA 101100804). https://trustfoodproject.eu/ ; https://zenodo.org/records/15348956 World Economic Forum (2019). Walmart is betting on the blockchain to improve food safety. https://www.weforum.org/stories/2019/01/walmart-is-betting-on-the-blockchain-to-improve-food-safety/ Kamath, R. (2018). Food traceability on blockchain: Walmart’s pork and mango pilots with IBM. The Journal of the British Blockchain Association, 1(1), 3712..

[Audio] Thank you!. thank you!. TALLHEDA has received funding from the European Union's Horizon Europe research and innovation programme under Grant Agreement No. 101136578. Funded by the European Union. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Research Executive Agency (REA). Neither the European Union nor the granting authority can be held responsible for them..