Large Research Infrastructure Building using FAIR Digital Objects

Abstract

Very high investments are planned for the coming decade at European level (EOSC), in Germany (NFDI), by the Nordic and other European countries in building large research infrastructures. The expectations are also very high to overcome the threat of a "dark digital age" (V. Cerf), to increase the efficiency of digital work (80% are lost with data wrangling), to offer best circumstances for advanced data intensive science and to enable participation in the exploitation of data for the benefits of societies. There is no doubt that solutions need to be based on global agreements that have the power to reduce fragmentation which is the result of an ongoing creolisation during the last decades. Technological innovation is extremely dynamic with new standards and thousands of tools being invented by smart people.

However, it is yet not commonly agreed how these agreements could look like. It is now widely agreed that the FAIR Principles should be taken serious as a guideline by all stakeholders. It is also agreed that we urgently need FAIR Maturity Indicators, as developed in RDA, to assess the FAIRness and not leave this to subjective evaluations. But such principles are not blueprints for building infrastructures. For a few decades some initiatives (CNRI, DONA, RD, GEDE) worked on the concept of Digital Objects (DO) and aspects of them such as protocol independent globally resolvable unique identifiers which are now in broad use in large data labs for about 20 years. DOs have the potential to reduce the enormous complexity in the digital domain by applying aspects such as abstraction, stable binding and encapsulation. Recently, the experts working on DOs on the one hand and those working on FAIR implementations on the other hand joined forces to agree on the concept of FAIR Digital Objects (FDO). FDOs are extensions of DO by introducing explicit semantics to meet the FAIR Maturity Indicators in a way that machines can process all information.

In newly designed research infrastructures covering many different sub-disciplines such as in biodiversity and sciences focusing on experiments with humans (from physiology to economics) this concept of FDO is now systematically applied. In addition, complex challenges such as creating joint metadata catalogues respecting the semantic spaces as defined in the sub-communities need to be solved. The talk will describe the challenges and discuss the solutions found.