SelfX and NLP

Semantic interoperability through natural language processing (AI) as the basis for self-X capabilities of asset administration shells in semantically heterogeneous asset networks
Current practice in industrial and building technology is characterized by a heterogeneous semantic description of assets (technical systems, system components, automation applications). Due to the wide variety of manufacturer-specific asset descriptions and inconsistent national and international standards, the realization of an interaction application of assets will continue to be decided in the future by weighing up the benefits against the costs. This is where the research project comes in. It extends current research concepts for the interaction of Industry 4.0 (I4.0) components (asset and asset administration shell) to include new exploration and interaction mechanisms in semantically heterogeneous networks. Assets with different semantic characteristics are identified and automatically mapped to a common syntax (information model) and semantics (meaning). This makes them semantically interoperable, a prerequisite for the self-configuration of their interactions. These new exploration and mapping mechanisms are based on artificial intelligence (AI) methods, specifically natural language processing (NLP), for recognizing the meaning and structure of asset functionalities. Prototype cloud-based energy and asset monitoring applications are being developed that demonstrate Self-X capabilities in heterogeneous and homogeneous semantic asset networks for technical building equipment. Software developed in the project will be made available as open source; product development or commercial use is not planned. All research results will be published and incorporated into standardization and guideline work through close networking with I4.0 working groups.
Project partner: -
Funding provider: KSB-Stiftung
Duration: 01.01.2021 - 31.12.2023