The Vision of the FAIR Digital Object Machine and Ubiquitous FDO Services

In addition to the previous intensive discussion on “Data Deluge” with respect enormous increase of available research data, 2022 Internet-of-Things conference confirmed that in near future there will be billions if not trillions smart IoT devices a very wide range applications and locations, many them computational capacities. This large number distributed create continuous streams data require global framework facilitate their integration into Internet enable controlled access services, name but few aspects. would tracking these measure resource usage for instance globally address UN Sustainable Development Goals. Additionally, policy makers are committed define regulations break monopolies sharing. The result an increasingly huge domain accessible digital which one hand allows addressing new challenges especially cross-sector ones. A key prerequisite this is find right across boundaries supporting specific task. Digitisation already being called fourth industrial revolution emerging information 21 st century's resource. Currently vision mostly unrealised due inability existing resources findable, accessible, interoperable, reusable despite progress providing thematic catalogs. As result, capacity latent underutilized. There no level infrastructure currently exists process by all made consistently accessible. patchworks localized limited trusted created communities have been funded or directed collaborate. To turn commodity, description, to, validation, processing needs become part we call Global Integrated Data Space (GIDS). main pillars approach services identified accessed, predictive descriptions control make findable. researchers relying partly informal knowledge such as knowing labs persons maximize chance trustworthy method limiting use suitable data. scenario, other mechanisms possible. public space Google-like searches using terms accepted solution documents human consumption. however, does work GIDS numbers contributors from institutions, millions worldwide, where types automatic procedures dominate. Indeed, successful apply complex models describing surrogates can automatically leverage labs. makes often operationally applied manual stitching operations too costly both time competitive option. researcher looking brain imaging study has options: Rely network colleagues. Execute known registries appropriate departments researchers. He/she engages agent execute profile matching sub-spaces. We assume creators capability interested detailed metadata different researchers, who able specify precise profiles they for. Two characteristics carry out at ultra-high speeds lead various subspaces according some facets self-organizing mechanisms. Of course, poses high requirements quality used potential consumers share about semantic operate, mappings brokers self-provided. Metadata must highly schemas developed communities. usual metadata, users need credentials form roles purposes indicate opportunities. Changing current practices yield richer presribed FAIR principles simple, since seem far away formalizing broadly way, pressure rich standardized increase. It should noted course sensors, defining proper action only requested once times. Why FDOs special scenario? bundling required secure i.e., via gloablly unique resolvable persisten identifiers (PID) FDO PID security basis establish trust. provided capable computing efficiently coded representing properties relevant matching. speedup enormously. two major questions characterizing “FDO Machine” envisioning: Which kinds representations could much more efficient? How FDO-based sub-spaces help layer offer specialized example UN’s Goals? Brokers might want agents along important domains, trustworthiness, roles, etc. Such ephemeral virtual structures top integrated space.