Building with bits and bytes in search of the core offering of a digital research infrastructure
The term ‘research infrastructure’ readily conjures up the image of large-scale research facilities that can be used by different studies. In other words: physical infrastructure for classical research. But we are now living in a digital era. Increasingly more research is being carried out with digital data. The emergence of big data, open data, and ever further-reaching progress in artificial intelligence facilitate the development of knowledge and insight without random samples and experiments. In a society that is becoming increasingly digitised, the growth in the number of digital data sets is explosive. Research is thus becoming more and more data-driven. However, the data sets originate from different sources and have diverse structures, variables and measurements. Innovation at both a technical and organisational level is needed in order to use these data sets effectively and efficiently. Challenges in this regard include: How do you find the right data? How do you ensure that the data are interpreted correctly? How do you guarantee the integrity of the data? And what arrangements need to be made regarding data access and use?
The answer to these questions lies in the development of a digital ‘road network’, or a digital research infrastructure, an important theme in EU research programmes. Besides researchers, such digital research infrastructures can also assist other stakeholders with their data-driven research, for example in the development of citizen science in which consumers and the public are becoming increasingly more engaged.
Core offering of a digital research infrastructure
But how do you develop a digital research infrastructure? Where do you begin? Who must you involve? And what added value must this research infrastructure provide? How do you make such a valuable research infrastructure sustainable and robust? In order to answer these questions, you first have to determine the core offering of the research infrastructure.
An example of a digital research infrastructure is the ‘Research Infrastructure for Food, Health, and Nutrition’ programme with its various associated research projects. The EU H2020-project (www.richfields.eu) is one of these. I have been working as a researcher on RICHFIELDS since 2015. It is an ambitious project because we want to lay the foundation for a joint European infrastructure for research into nutrition, health and behaviour.
The research infrastructure we are working on is primarily for researchers, making it far easier for us to share, link and use scientific data and insights. But we cannot find solutions for social challenges that are related to nutrition and health without the involvement of the other links in the chain. These include consumers and businesses that supply us with data on purchasing behaviour, preparation, consumption and lifestyle. They also include policymakers who base policy decisions relating to nutrition, health and behaviour on validated research data.
It is tempting to devise all types of functionalities for such an infrastructure in order to be able to serve as many stakeholders as possible. But a research project has a limited lifetime and resources. This means that not everything is possible within the available time and resources. A choice from the core value is then unavoidable. In other words: what is your core offering? What do you want to have achieved in four years, the lifetime of the project? And what can be developed further afterwards? Will it become a data hub? Or is it mainly a portal with links to different data sources? Are you building new standards or mainly harmonising existing ones?
Defining building blocks together
The research infrastructure should ideally provide added value to all links in the data chain. However, searching for the core offering of RICHFIELDS has been a long process, precisely because we want to provide added value to all links in the chain.
The research infrastructure design has come about through close consultation with consumer, business and government representatives. An important outcome of this has been the definition of building blocks. Building blocks of the digital research infrastructure include data models in order to make variables from different databases comparable. For example, if we want to research how breakfast contributes towards the health of European consumers, we must first properly define what we mean by ‘breakfast’. After all, breakfast habits vary greatly in Europe and data from different studies cannot simply be bundled together.
Building blocks such as data models are the technical elements of the digital research infrastructure. A digital infrastructure also needs organisational building blocks. This will be the topic of a subsequent blog by my colleague, Marc-Jeroen Bogaardt.
Search for the core offering of RICHFIELDS
Do you already have ideas for the application of the infrastructure? Leave your ideas or questions below. You can, of course, always contact us directly!