At the IoT Week 2022, AIOTI hosted a session on the IoT and Edge computing standardisation in the context of data spaces (Wednesday, 22 June 2022).
Speakers of the session:
Svetoslav Mihaylov presented the focus of the panel, i.e., standards related to IoT and Edge computing to enable the integration of data spaces, including the move to the Far Edge. He presented the strategic picture for Europe on a common open digital platform and ecosystem for the industrial internet of things, presented examples of projects (OpenDEI, Synchronicity), as well as the launching of projects on a MetaOS for next-generation IoT and Edge Computing (ICOS, FLUIDOS, NEMO, NEBULOUS, AEROS, NEPHELE) and their support by (OpenContinuum, Unlock-CEI=.
Gyu Myoung Lee presented challenges towards data-centric ecosystems: an ecosystem of ecosystems, from data provisioning to data usage, data lifecycle, common language for data interoperability, common data models, data curation, trust in data sharing, governance and ethics, decentralization, data fabric. He concluded with the vision that a data-centric approach is essentially a trustworthy decentralized data ecosystem with AI capabilities.
Silvia Castellvi presented IDSA activities on the IDS-RAM 4.0 (reference architecture model), its structure into layers and perspectives, its availability in GitHub, and associated portfolio of open source building blocks, as well as a 3rd party certification scheme. She also pointed out IDS standardisation roadmap including promoting DIN SPEC 27070 to ISO/IEC, participating to data spaces standards. She also informed of the Open DEI building block survey that is ongoing.
Juanjo Hierro presented the plan for addressing technical alignment within DSBA (Data Space Business Alliance), which includes BDVA, Fiware, GAIA-X, IDSA. A starting point is the 12 building blocks described in the OpenDEI design principles for data spaces, the existence of convergence catalysers (W3C verifiable credentials, TM Forum based marketplace, DOME proposal for the federation of marketplaces). The approach is to deliver versions of a minimum viable framework (MVF) with a timeline aiming at 2023.
Milenko Tosic presented the Interconnect semantic interoperability framework, addressing the interoperability and distributed requirements of data spaces through distributed facilitating digital platforms, generic adapters, ontology agnostic approach, security, privacy and trustworthiness and multiple deployment options. These enablers provide the basis for the creation of federated knowledge pools as the foundation for building data spaces. The InterConnect semantic interoperability framework also includes a registry of interoperable services, a configurable security framework and supporting services for system administration and framework integrators.
Fabio Coelho presented the value of a semantic interoperability layer to build a data space based on federated platforms. He reminded the action plan on the Digitalisation of Energy (DoEAP) that is calling for a global data space on energy. He contrasted three architectures: one where there is no data space and where a service has to cope with multiple custom data integration to interface services from different systems, a data space architecture based on a common domain data representation, and the data space architecture based on a domain agnostic data sharing technology. Interconnect has implemented the latter by developing a semantic interoperability building block that can be added to future mainstream data space frameworks (e.g., GAIA-X). It is, leveraging an ontology agnostic toolset and has been used to implement energy, smart homes and smart building cross-domain applications based on SAREF.
The session was concluded by Antonio Kung. He announced the availability of 3 further position papers (AIOTI: guidance for the integration of IoT and Edge computing in data spaces, BDVA: Data sharing spaces and interoperability, OPENDEI: Reference architecture and interoperability in digital platforms) that will be the starting point for the submission of standards on data spaces. He then confirmed with the speakers on the takeaways: