For the past two decades, I have worked on e-Agriculture, ICTs for Agriculture and Digital Agriculture as a practitioner, researcher, development officer, and now a policy adviser. My conclusion is that, for digital solutions and services to be sustained, scaled and then accelerate smallholder agriculture transformation, we need a better approach to data management at country level.

The problems/challenges we have been experiencing within the sector has little to do with the digital technologies, services and solutions. It is more about data – agricultural data.

According to Gartner, “At least 30 per cent of generative AI (GenAI) projects will be abandoned after proof of concept by the end of 2025, due to poor data quality, …..

Data is the rail upon which digitalisation thrives and has become the fuel for economic growth for many economies today, and more so tomorrow. The European data strategy for example aims to make the EU a leader in a data-driven society. By creating a single market for data will allow it to flow freely within the EU and across sectors for the benefit of businesses, researchers and public administrations. The United Kingdom on the other hand, argues that governments have an important role to play in laying the foundations for a flourishing data-driven economy by maintaining a secure and trusted data environment for data sovereignty. This will involve pursuing policies that improve the flow of data and ensuring that companies that want to innovate have appropriate access to high-quality and well-maintained data.     

The good news is that the issues we experience in our countries with agricultural data is not the lack but rather the overload. The issue is about the siloed, disaggregated and duplicated data systems that are often built without interoperability. This is the root cause of the challenges that we are experiencing. Multiple datasets exist on a given data point but because the data systems are not interoperable, access is a challenge. Hence, investors, financiers, researchers, service providers, policy makers, innovators, etc. cannot find the needed data not because they don’t exist but because they are not accessible. This leads to operational inefficiencies, failed business models, low adoption of services, failure to scale innovations, increased costs, revenue losses, limited financing and investment, inaccurate policy decisions, among others.

Your AI model is as good as the quality of data on which it runs, and the quality of data depends on the infrastructure that holds it.

And that is where infrastructure at country level comes in for the management of the data. A digital public infrastructure (DPI) for data? I strongly argue that what we need for the agricultural sector is DPI for data exchange at country level and not just another data exchange platform.

But firstly, what do I mean by “infrastructure” for agricultural data at country level? A solar energy “infrastructure” is more than just the solar panels; a telecom “infrastructure” is more than just the cell tower; and a road “infrastructure” is more than just the tarmac.

For road infrastructure, we will need to ask ourselves what does the speed limit signage represent on the road? What does a speed camera represent on the road? What does a tollbooth represent on the road? I argue that every national road infrastructure is unique to the country and has (I) technical (e.g., tarmac), (II) policy (e.g., speed limit), (III) governance (e.g., speed camera) and (IV) business (e.g., tollbooth) components.

Different vehicles of different sizes follow the guidelines of the road infrastructure to use the standardised tarmac. Governance mechanisms are in place to ensure road users adhere to the guidelines. And revenue streams exist to ensure income for maintenance and sustenance of the infrastructure.

So, every country needs a national infrastructure for their data, and the infrastructure should have the 4 components.

The national data infrastructure should take advantage of the existing data systems in the country. Therefore the concept of national agricultural data infrastructure (NAgDI) is about strengthening and harmonising existing individual data systems at country level through an interoperable national data infrastructure, enabling interconnected infrastructure at regional and global levels, thereby creating a superhighway for secure data exchange within and between countries, and across regions for macro-level decision-making.

Secondly, infrastructure for what? Infrastructure for agricultural data. So, what do I mean by agricultural data? For agricultural data, I mean two main groups of data. The first is content data – the data whether in digital, optical, or other form that conveys essence, substance, information, or intelligence in either its unprocessed or processed form (E.g., agronomic data, weather data, production data, etc.). And the second is user data – the data or different pieces of data when brought together, can lead to the identification of a particular person or entity across the agricultural value chain for service delivery (data on farmers, enterprises, field IDs, etc.). The data represent the vehicles that run through the infrastructure. For the user data, the identity acts as the number plate of vehicles. Once captured, other details of the owner can easily be discovered and therefore must be protected. But at the same time, the identity of the data might be a requirement for the provision of customised and precised services.

National agricultural data infrastructure with the four components will ensure equity with accessible, secured and quality data for all stakeholders and thereby reducing the lack of data in the age of data overload.

Learn more about Commonwealth NAgDI – https://thecommonwealth.org/news/blog-policy-practice-building-national-data-infrastructures-through-bottom-approach