A review of the session on “Digital agriculture and value chains” at EAAE170, Montpellier, 15-17 May 2019

From 15 to 17 May this year, 150 scientists were welcomed in Montpellier for EAAE170, the 170th conference of the European Association of Agricultural Economists, organised by the MOISA joint research unit on the subject “Governance of food chains and consumption dynamics: what are the impacts on food security and sustainability?”. The session on “Digital agriculture and food chains” gave rise to three #DigitAg-INRA-CIRAD-Montpellier SupAgro communications on the determinants of ICT adoption, innovation networks and market information systems. Key elements.

Understanding the determinants of IT adoption in agriculture using an integrated TAM-TOE model

The communication by Isabelle Piot-Lepetit addressed two research questions: what are the processes and factors in ICT adoption? And how do they influence a regular use? She presented an integrated model to study the adoption and use of digital services by farmers. Based on a review of theoretical models that facilitate the analysis of innovation adoption, two models were used, TAM (Technology Acceptance Model) by Davis (1989) and TOE (Technology-Organisation-Environment framework) by Tornatzky and Fleisher (1990), from which a new integrated model was proposed: “TAM-TOE, IT adoption and use model”.

With the TAM-TOE model, the factors explaining IT adoption in agriculture are:

  1. individual: perceived usefulness of the solution, its ease-of-use, training of users, their capacity to innovate, their knowledge/previous IT experience, observability of solutions/IT
  2. organisational: scope of the farm’s [A1] activities and its size
  3. technological: cost reduction, adaptability to user practices, compatibility of IT solutions and tools with existing processes and practices, confidence in IT solution or tools
  4. environmental: perceived trialability of the IT, solutions and tools, volunteering, availability of resources

The risk linked to IT solutions/ tools perceived by the user is a negative individual factor

The bibliometric analysis of studies on ICTs in agriculture was conducted on a body of articles from the Science Direct database for the period 2002-2016 and limited to a number of countries (Australia, Belgium, Canada, Denmark, Germany, Ireland, New Zealand, Norway, United Kingdom). It shows that 55% of the most prevalent factors are individual (perception of usefulness, training, ICT experience, perception of ease-of-use, age, observability), 20% are organisational (size and type of farm), 17% are technological (cost, adaptability, confidence, compatibility) and, finally, 8% are environmental (availability of resources, trialability).

It is now necessary to study how IT solutions/tools are used by farmers, and how they adopt them and sometimes adapt them over time.

Co-construction of innovation processes: what types of innovation networks exist in digital agriculture?

Technological innovation is a collective process involving a wide range of actors: new actors who promote the use of ICTs, traditional actors in agricultural innovation systems, such as basic and applied research, and agricultural actors who are the intermediate or end users of technologies. In order to explore and analyse innovation networks in digital agriculture, the literature on innovation systems is mobilised, in particular with the sectoral approach. A Sectoral Innovation System (SIS) refers to all actors, networks of actors and institutions that promote the production of new knowledge and the development of innovation in a given sector (Breschi and Malerba, 1997; Malerba, 2002).

Three main elements of the SSI have been taken into account: (1) the sector-specific knowledge mobilised; (2) technological innovations, which represent the object of interactions between actors and are a major constraint on the diversity of actor behaviour, on organisation, and on the potential forms of innovation networks; and, finally (3) innovation networks, which represent temporary sets of partnerships, made up of private or public laboratories, farms[A2] , clients, suppliers, and financial institutions, with active participation in the development of new products.

 

What types of innovation network organisation exist to promote knowledge development and technological innovation in digital agriculture?

17 interviews were conducted with the different actors in the system: companies/start-ups, research units, agricultural education, technical agricultural institutes, chambers of commerce and industry, chambers of agriculture, and agricultural advisory services. The data collected concerns three pillars: technological innovations (uses, specific knowledge used and innovative companies in the sector), knowledge (type, actors involved in its production, its usefulness in innovation processes), and innovation networks (types of relations between actors, their characteristics and functioning, their roles in the innovation process).

 

The construction of the networks of actors in favour of innovation is determined by a number of factors that make interactions possible (diagram) Two organisational forms of innovation networks are identified: simple and complex.

Simple organisational forms are generally composed of two types of actors and three main organisational structures:

  • Business to business: this is a partnership between a client company and a supplier company, linked to technological constraints and to the size of companies that do not necessarily integrate all skills. For example, BeApi outsources the “information system” part to SMAG and part of the “satellite image processing” of its technology to the ISAGRI group, which specialises in remote sensing and satellite image analysis.
  • Company – research: this type of structure links a company to a research actor. Indeed, many companies mobilise a stakeholder network in which the direct contact is a research actor (Inra or Irstea, for example), with which they develop communication and cooperation relationships around the construction and evaluation of innovative technologies. This type of partnership covers the outsourcing of R&D within companies for technology benchmarking purposes, for example, or to mobilise skills and research work in order to develop a new service.
  • Company – agricultural intermediary (cooperatives…): in this partnership, an agricultural intermediary enables the company to assess the expectations of farmers (the end users of the technology) and to test innovative solutions in real conditions with farmers, which all helps to more effectively meet their needs.

Complex organisational structures are organisational forms that bring together more than two actors, and are mainly structured around cooperation, communication and exchange relationships. This “necessary” interaction results from the desire of different actors to grasp all aspects related to innovation in this sector, not only upstream of innovation processes with R&D and prototyping activities, but also downstream, in the context of the launch of new products and their adoption by end users. The specificity of these actor ecosystems lies in the specific knowledge and skills of their members, but also in the purpose of their interactions. The range of interaction subjects can, for example, include reducing farm dependence on crop protection products, using remote sensing to improve agricultural yields, the well-being of farmers, etc.

We see several different organisational forms, which are financed by various means and bring together actors in research, training and development. This is the case of the Joint Technological Units (UMT), in which public research teams and professional technical organisations (technical agricultural or agro-industrial institutes) cooperate on national research and development subjects. One example is the UMT EcoTech Viti, which fosters linkages between Irstea, IFV and Montpellier SupAgro around collaborative projects to reduce the dependency of winegrowing enterprises on crop protection products through the evaluation of innovative technologies. The Joint Technological Networks (RMT), on the other hand, decompartmentalise research/training/development actors and foster closer linkages between public research and business. One example is the RMT Modelia, which comprises Inra, Irstea, Cirad, Arvalis, ITK and Agrosolutions around the Modelling and analysis of data for agriculture. Another form is the Digifermes network, which brings together several ACTA technical institutes (ARVALIS, IDELE, ITB, Terres Inovia) and innovative companies for the co-construction and development of digital tools for agriculture.

 

The study highlights the plurality of organisational forms that coexist along the innovation chain and process to form an innovation “community”, with relatively limited involvement of agricultural organisations in the innovation network, which are clients or end users of the technology. The interactions between the different actors within these networks contribute to building a new innovation system, which leads to a change in the technological regime for agriculture. It contributes to discussions on inbound open innovation in companies, and its implications in terms of the management of external knowledge and collaborations with different stakeholders. Finally, the study focuses on innovation in SMEs and the construction and functioning mechanisms of innovation networks. Extending research to large companies in the sector (Airbus and Geosys, for example) could be a complementary approach to understanding the innovation chains and networks in this emerging sector.

Market information systems efficiency: questioning access and use before seeking impact

Mobile phones have spread rapidly in the developing countries, making it possible to connect rural populations to markets and sparking new interest in market information systems since the early 2000s (Galtier et al. 2014). As a whole, the scientific community is showing a growing interest in ICT and development issues. Economists are focusing their research on assessing the impact of MIS (market information systems), but little is known about the adoption process, the constraints or the impact pathway.

These systems still reach only a small number of farmers. What are the constraints of access for users? What information is actually required and what is it used for? The goal here is to test a rapid assessment method to adapt MIS to users’ constraints and needs.

The study was conducted in Burkina Faso and Tanzania. It focuses on two systems, SIMAgri by the NGO APROSSA – Afrique Verte in Burkina Faso, and MAMIS by the MVIWATA farmers’ organisation in Tanzania, which recently developed mobile applications to collect and disseminate market data. Through its site and mobile networks, SImAgri provides information on market prices (a wide range of agricultural products), individual offers to buy and sell, as well as alerts on events. The prices of a wide range of agricultural products and individual offers to buy and sell are also available through the MAMIS mobile application at MVIWATA. All of the messages sent and received by the two servers were also extracted and analysed (88 287 messages in Burkina Faso for the period 2013-2015, and 51 893 messages in Tanzania, from 2010 to 2016). To complement the traceability of message flows, a short telephone survey was conducted among 112 users in Burkina Faso and 165 in Tanzania.

The analysis of message flows received by servers shows that:

  • The use of services is strongly linked to the marketing season for the main crops. No net increase in the number of messages is seen over the years.
  • The main staples are the most requested products. Maize represents 47% (Burkina Faso) and 32% (Tanzania) of messages. It has a leading position in markets, exceeding traditional staples (BkF: sorghum, millet; Tz: cassava). Some more marginal but specifically market-oriented crops are also of interest to users: sesame and soybean in Burkina Faso, and onions and tomatoes in Tanzania. It would therefore be more efficient to focus on a few strategic products, ensuring the quality and regularity of data disseminated, than to cover a very wide range.
  • User requests mostly concern market prices (BkF: 82%, Tz: 85%). The submission and consultation of individual offers and requests are marginal.
  • There are few regular users. Most users test the services and then give up: 51% of users in Burkina Faso and 89% of users in Tanzania did not consult the service more than six times over the period. There are various types of constraints: poorly formulated requests, information requested unavailable, poor network coverage, language barrier.

Surveys among users show that:

  • Regular users are mostly farmers, but 25% are retailers and processors in Burkina Faso. This is linked to the fact that APROSSA is actively engaged with this type of actor. In Tanzania, 18% of regular users are local organisation leaders and instructors, as MVIWATA is made up of a network of producer organisations. User training is emerging as a key driver of adoption: in Burkina Faso, where structured training has been provided, 85% of users discovered SIMAgri through this training.
  • Information is used to expand commercial opportunities (new buyers, sellers, products, places), to improve negotiating capacity and to save time.

Globalement, l’adoption de ces Systèmes d’Information de Marché  est très lent, même s’ils recueillent des appréciations positives.
Mieux répondre aux besoins et capacités des petits exploitants agricoles impliquerait :

  • d‘adapter la technologie aux utilisateurs peu instruits par des approches interdisciplinaires TIC / développement rural…
  • de prendre en compte les processus d’apprentissage (compréhension => confiance => capacité d’utilisation => prise de décision) au travers d’un environnement institutionnel qui favorise le transfert de connaissances : formations, groupes d’utilisateurs, « intermédiaires » de connaissances, plateformes d’innovation…
  • de mettre l’accent sur la qualité des données et des éléments d’analyse, notamment les tendances, pour des prises de décision à moyen terme
  • d’adapter progressivement les dispositifs par des processus itératifs de suivi et d’ajustements :
    • intégrer dès la conception du service mobile les possibilités d’extraction des flux de messages (principale source de suivi)
    • avoir recours à des enquêtes légères, à des approches participatives, pour capter l’usage et les perceptions des utilisateurs
    • tenir compte de la diversité des effets, sans se limiter aux effets directs sur les prix de vente

Overall, the adoption of these market information systems is very slow, even if they have positive appraisals.
Better meeting the needs and capacities of small farmers implies:

  • adapting the technology to less educated users through interdisciplinary ICT/rural development approaches, etc.
  • taking account of learning processes (understanding => confidence => capacity for use => decision-making) through an institutional environment that fosters knowledge transfer: training, user groups, knowledge “brokers”, innovation platforms, etc.
  • placing the emphasis on the quality of data and elements of analysis, especially trends, for medium-term decision-making
  • progressively adapting systems through iterative monitoring and adjustment processes:
    • integrating from the mobile service design stage possibilities for extracting message flows (main source of monitoring)
    • using light surveys and participatory approaches to identify uses and user perceptions
    • taking account of the diversity of effects, not just direct effects on selling prices

 

  • Contacts:  Digital Agriculture Session > Leila Temri – leila.temri [AT] supagro.fr – Cirad MOISA – EAAE170 > Paule Moustier – paule.moustier [AT] cirad.fr
  • Conference Website