When COVID struck, the response among the scientific community was to share DNA sequencing and to rethink traditional ways of working to solve the urgent challenge of creating a vaccine, writes Chipo Msengezi, Project Manager, Digital Development, CABI.

This leads to some big questions about what else can be achieved through more concerted approaches to data sharing. What impact could this have on farming across the global south and the drive towards more sustainable food production, for example? 

Researchers in agri-food systems often go to great lengths to collect data. They also know, better than anyone, the value of accessing data from third parties, and yet very little of the data collected become accessible to others. How can that be?

Value of Data Sharing

There can be good reasons why organizations, or individuals, feel uncomfortable about sharing data. But the more these issues are confronted in stakeholder engagement processes, the more likely it is that workable solutions can be found and appropriate safeguards can be put in place. 

At CABI, we are working with an increasing number of donors and partners on this issue. They see us as experts in the collection, storing and sharing of data. This is why in 2018 the Bill & Melinda Gates Foundation began working with CABI to investigate why data weren't always being used, even when everyone involved was committed to sharing data.

From a series of conversations with the foundation-funded soil projects in India and Ethiopia, some issues emerged that worked against data sharing. These were largely non-technical factors. The recurring challenges we found included a lack of resources, processes and procedures, such as a mandate for data sharing.

Also, there were practical issues. Typically, research platforms are tied to project funding and are not always available to use after the projects close. Too often data are lost, hard to find, or in formats that are not easy to use, even when the intention was the opposite.

Challenges in the Data Space 

In the wider data environment, these challenges have led to the development of a series of initiatives, such as the Open Data Movement. CABI has also done a lot of work on open data, and found this also creates challenges. For example, not all data that are collected can, or should, be shared. There are now global and national data protections to safeguard people’s personal information. This means most data need some curation and redaction.

CABI’s work interested the foundation. The foundation saw it could support grantees to implement the data access policies that the organization had recently adopted. CABI proposed using the FAIR data principles as a framework for improving access to data. 

FAIR data is defined as data that are Findable, Accessible, Interoperable and Reusable. As a framework, it’s gained wide acceptance, especially by  the EU Horizon programme, and with other research funders. 

With the foundation, we worked within funded projects to address the challenges the projects encountered. But, critically, CABI also started to work with foundation staff to better equip them to assess and support promising data projects to embed good practice from the outset. It was important to make the foundation aware of the potential pitfalls in a proposal as early as possible. But equally important, the program officers and the grantees needed to be able to access help and support from experts able to deliver the FAIR data principles.

FAIR Access

These FAIR data principles are a brilliant starting point. However, they were originally developed for an audience that is very comfortable with technical language, which can be stressful and off-putting for others who are also integral for designing strong projects.

CABI’s approach to data sharing is to work collaboratively with our partners to customize and define project-level FAIR principles. Our approach is to first understand the people in any data ecosystem, their relationships with each other, then the challenges, needs and opportunities. 

At CABI, we want those we work with to tell us what FAIR and data sharing usefully means to them. It is far more effective if the definition comes out of engagement with the community that supplies and use the data.

CABI supports practical implementation of the FAIR approach, designing better processes and training teams. This starts by understanding the problem. Then we map the landscape to work out where we can make the most difference in support of better data sharing. 

We are currently testing a process that aims to guide organizations towards developing their FAIR implementation strategies. Often, this is in parallel to helping donors and governments design better data governance, strategy and policy. 

Building Awareness

Getting to a position where an organization has its data available to share is a long-term, resource-heavy and time-consuming process. Furthermore, most of the benefits will accrue outside of the organization. The potential benefits flowing from these changes can be transformative. For example, data can help to support improved resilience among smallholder farmers, their communities and the natural environment, but for an organization to be ready to share their data effectively takes leadership from the top of the organization.