Evaluating Digital Resources for Farmers

This resource evaluates a collection of existing digital tools for use by farmers from the perspective of adoption by the cooperatives of SEWA, an association of over 2 million self-employed women workers centred in Ahmedabad, India. It is hoped that this evaluation and analysis will be helpful to similar groups of cooperatively organised informal workers, many of whom may be engaged in subsistence farming.

This analysis looks at technologies which are to some extent specifically intended for use by farmers, although many general purpose data technologies may also be highly appropriate, especially in the early stages of developing a data infrastructure. These range from ordinary office productivity software, whether on the desktop or in the cloud, to more specialised data tools such as AirTable, Tableau and others. Some such tools are suggested for use by cooperatives on the International Cooperative Alliance’s website. At the time of writing, some cooperatives within SEWA are prototyping the use of Looker Studio as a framework for surfacing basic quantitative information (turnover, headcount and the like) for communications within and amongst the federation.

Our evaluation must be done with particular sensitivity, since as reported in a recent IT for Change publication, How Digitally Restructured Value Chains Are Reshaping Labor Futures for Women in the Global South (December 2021), “Female farmers on digital platforms are not able to attain the same productivity gains as their male counterparts due to problems related to tenure security, informal institutional constraints and intra-household dynamics”. Further, the report notes that, although “A survey of 400 Ugandan entrepreneurs engaged in different stages of agricultural value chains found that access to and ownership of mobile phones, which appear to be positively correlated with enterprise development”, that “there is a prominent gendered digital divide in ownership”, partly through lack of awareness, but also through “unique challenges related to formalization. Women avoid registering their businesses and formalizing, as keeping their business ‘invisible’ can be an important way to protect themselves against their husbands or male relatives taking charge of their earnings and resources.”

In contrast, a 2021 CACM paper, Digital Agriculture for Small-Scale Producers: Challenges and Opportunities, presents a perspective based on the “catching up” narrative which is often attached to communities labelled as part of “developing countries”. Under this narrative, the goal is to bring workers in these areas into parity and consistency with economic and technological practices that are widespread in wealthy nations in the West. This narrative has long been criticised from several traditions, such as that of Ecofeminism, from which Maria Mies in a 1993 essay “The Myth of Catching-Up Development” points out that this is an economically infeasible deception targeted at the “third world”.

The Data Communities for Inclusion project takes the view that these technologies will have to be deployed in a way which is sensitive to the unique perspectives of the cultures and communities involved. Technologists should be more prepared to listen to members of these communities as they explain what data means to them, and what uses they would like to put it to, rather than assume that their trajectory must resemble those of people who are in positions of economic and cultural dominance. Use of existing technologies that were not grown within or intended for the cultural context of worker-ownership and grassroots self-organization is necessary and unavoidable, but involves a process of working with communities to find ways of orchestrating them and reconceiving them in ways useful to this context.

Towards Reparative and Pluralist Data Infrastructure

Part of the work of this evaluation is to uncover the implicit framing behind a technology and what questions it is trying to address. For example, many existing digital technologies for farmers have a central aim of applying a security model to data, and preventing its use for inappropriate purposes. This could be seen as a “preventative” approach to data breaches – for example as seen in JoinData and FarmStack. However, this may not be the most important need as experienced by the farming community – JoinData’s questionnaire showed that only 6% of respondents were interested in this question. 

The “preventative” approach could be contrasted with a “reparative” approach under which the technology makes it easy to track the provenance of data as it travels around the system, and puts tools in the hands of communities to let them discover and remedy inappropriate uses. This reparative approach seems more realistic under the view that in a world populated with powerful information actors such as cloud computing corporations, national data agencies and massive-scale artificial intelligence, breaches are very likely to occur during the ordinary processes of working with the data. That said, any system should make straightforward efforts to ensure that data does not routinely fall into the wrong hands.

The pluralistic data infrastructure developed by the WeCount project demonstrates how provenance can be tracked as data travels throughout a system. The infrastructure allows for the surfacing of data provenance through visualisations that can feed into a reparative model where it is more clear which community’s viewpoint is responsible for a visible piece of data.

According to Data Feminism, “the most complete knowledge comes from synthesizing multiple perspectives, with priority given to local, Indigenous, and experiential ways of knowing” (Chapter 5). This definition of pluralism can support an inquiry into our notion of what data is and how communities can decide what to collect, from whom, and for what purpose.

As we survey digital solutions, the following questions are in our minds:

• Who becomes empowered in a transition to a digital solution?
• Are the aims addressed by technologists those which are most urgent to the population?
• What becomes of the citizens’ data and workflows if a commercial solution fails?
• What ideologies are embedded in the technologies, their orchestration, and the organisational structure of those delivering the technology, that will end up being inscribed on the communities standing to benefit?

Resources

Farmers’ Portal, Government of India

Last evaluated February 2023

The portal’s purpose is to give farmers access to “all relevant information on specific subjects” at all geographical scales, delivered in the form of text, SMS, email and audio/video in a language that they understand. It is also intended as a clearing house for queries and feedback through a “Feedback module”.

This site loads very slowly, and has a “Last Updated” status Jan 2021 with latest News from 2018. It has links to a collection of apps – e.g. “AgriMarket mobile app can be used to get the market price of crops in the markets within 50km of the device’s location”. The app links are broken and the site’s JavaScript seems to have been broken for years.

JoinData 

Last evaluated February 2023

JoinData is an independent platform focused primarily on dairy farmers in the Netherlands. The primary function is of a data platform, enabling sharing of farm data with government, accountants, suppliers and customers. The cost to farmer members is fixed at 50 Euros per year. There is a well-elaborated technological platform, with a strong emphasis on permissions and sharing as capabilities of the infrastructure. It is advertised that apps are supported but these are not publically available. The corporate structure is interesting and relevant, since JoinData is organised as a cooperative, with general assembly twice a year of the member organisations. It operates a “Data Sounding Board Group with six experts in the field of software and data exchange”.

JoinData’s ontology and language are fairly tightly aimed at the particular target market which is Dutch-speaking dairy farmers. Their development documentation is well-organised and readable. There is a well-developed, fairly open integration model with publicised data model and API – e.g. from https://boerzoektapp.nl/documentatie/api-documentation/. There are helpful descriptions of intended use of the architecture, for example:  “If you have data flows that do not make use of the Data Hub, for example if you retrieve farmer’s data from or share farmer’s data with third parties directly, you probably need to verify if the farmer has given consent. You can store those mandates in our Purpose Registry and use the API’s to check the existence of consent.”

Most helpfully they have performed a survey of the priorities of farmers using their system under “What do farmers want” at https://boerzoektapp.nl/wat-willen-boeren/. The results are rather interesting – contrary to the focus of most platforms on data sharing and permissions, this ranks lowly:

• “Spending too much time on administration” – 44%
• “Ever more changing rules” – 25%
• “My data is not making money” – 25%
• “Who can access my data” – 6%

Community Videos from Digital Green

February 2023: these videos are continuing to be regularly produced with a good cadence over the last couple of years – typically one finds 5-10 within a given week.

These videos, produced in collaboration with grassroots partners and rural farmers, are freely available on YouTube, with more than 6000 locally relevant videos available in more than 50 languages (primarily Hindi, Telugu, Kannada, Odia). The purpose is to identify information gaps (for example, paying excess prices for produce) and to disseminate impactful practices in agriculture and nutrition behaviours. They are very well-produced and on a wide range of topics (health, nutrition, farming practice – majority on agriculture). A good illustrative sample is “Indigenous Technical Knowledge for Plant Nutrition in Chilli” in Telugu at https://www.youtube.com/watch?v=igVxFNm8cC0, co-produced with Government of Andhra Pradesh Department of Horticulture & ITC.

FarmStack from Digital Green

Last evaluated February 2023

FarmStack the protocol, “FarmStack Open”

The “FarmStack” name is used variously by DigitalGreen to describe some collections of its services, including the community videos, but most recently seems to be attached to a particular open source product, described at https://docs.farmstack.co/ as “a reference implementation of an open and interoperable data sharing protocol in agriculture sector.”

In its architectural detail there does not seem to be anything specific localising FarmStack particularly to the farming sector – it does not impose a strong schema on the data which is exchanged or the roles of those exchanging it. This is at the same time a strength and a weakness – it is possible to adapt FarmStack for many different contexts, but dedicated development resources would need to be available to do so.

Work had been done on FarmStack in an open source GitHub project farmstack-open but there has been no substantial work there since the departure of Manyank Gautam in April 2022, the author of the great majority of core work as Technology Lead at Digital Green.

The technological basis of FarmStack is the Fraunhofer Institute’s’ International Data Spaces protocol, which was renamed from “Industrial Data Spaces” in 2008. The architecture emphasises “sovereignty over data”, and in its description of an IDS Clearing house, states “Data in the Clearing House is stored encrypted and practically immutable. There are multiple ways in which the Clearing House enforces Data Immutability”. This has close connections with, but not identical to the blockchain model which is popular in many modern platforms, including the Regen Network.

FarmStack the SMS and IVR platform

In the context of delivering SMS and IVR (interactive voice response) messaging, this incarnation of FarmStack is described as using “the integration of farm and farmer data to develop customized and demand driven information and services offered to farmers via multiple integrated channels of communication”.

The primary document Farmstack Evaluation Results records an evaluation conducted between November 2019 and August 2020 by IDInsight and was delivered February 2021 to the Gates Foundation. However, the product evaluated appears to be unrelated to the data sharing protocol described above. The evaluation compares treatment of groups of farmers who receive just video vs those who receive video + SMS + IVR (interactive voice response), and shows most impact resulting from videos with small additional effect from text and voice messages.

As well as “Knowledge Impact”, other measures were also evaluated such as production and yield which showed much smaller though measurable benefits. The cost of the platform was estimated at $53,056 for reaching 5770 farmers resulting in a per-farmer marginal cost of $3-$4.

Pluralistic Data Architecture, Project WeCount

Last evaluated February 2023

This architecture “supports communities in taking collective ownership of data that relates to them and curating its relationships with data from other sources.” It is somewhat similar in positioning to FarmStack’s protocol/platform. Similarities are:

• Emphasises working with simply structured tabular data in CSV/Sheets
• Does not mandate a schema 

Differences:

• Does not provide a protocol, expects to be orchestrated in a cloud or locally
• Does provide tracking of provenance i.e. where each cell of data has come from

Rather than a fully developed solution this is purely a technology demonstration – however all the elements are open source and easy to start working with. In addition there is a continuously running cloud deployment which demonstrates the infrastructure running on a free platform which is also easy to copy and adapt. Similarly to working with FarmStack Open, this would require dedicated development resources.

FarmOS 

Last evaluated February 2023

Describes itself as “a web-based application for farm management, planning, and record keeping. It is developed by a community of farmers, developers, researchers, and organizations with the aim of providing a standard platform for agricultural data collection and management.”. It has a field kit app, which works offline, and is “modular, extensible and secure”. 

The project has been running for more than three years, has good test cases, and is primarily maintained by one developer, Michael Stenta, although there is a wide network of minor contributors. It has received sponsorship from a wide variety of institutions, mostly based in the US. 

Architecturally FarmOS is an extension of the popular open source PHP content management system (CMS) Drupal. It would be relatively easy for a community to host their own installation or commission one in the cloud, although this would require access to dedicated development and operations staff.

FarmForce

This Norwegian company, with global reach, offers digital solutions to help farmers with sustainable sourcing, improve their quality of life and protect the environment. They were the subject of a November 2022 report of from the Data Economy Lab, Exploring the value of adding a data layer to cooperatives of a visit from the Aapti Institute to SEWA’s Megha cooperative. The paper describes FarmForce in the context of data cooperatives, as “a digital procurement solution allows cooperatives to track interactions with farmers such as input supply. Examples of data points typically captured in the app include farmer ID, crop information, farm geolocation data, farm size, crop cycles and a photo of the farmer”.

However, it is not clear the extent to which FarmForce constitutes or supports a data cooperative. Its governance structure appears to be that of a conventional company, offering services to, amongst others, cooperatives, described as part of their general offering to cooperatives, aggregators and SMEs.

FarmForce’s solution “digitize food’s first mile and enable comprehensive farm management among smallholder farmers. Improved monitoring of farm development, increased production, and compliance with certification standards leads to more opportunities to access local markets. Our technology allows smallholder farmers and cooperatives to create a digital economic identity and financial history to secure financing”. FarmForce also advertises that a mobile app will be available to support access to Carbon credits, with a pilot study tracking information from 2000 surveys of avocado farmers together with Sunvado of Ethiopia.