Eclipse Semantic Modeling Framework (ESMF)

Digital Twins – digital representations of physical as well as abstract assets, e.g., a drilling machine or a production process – are the foundation of digitalization efforts in production and logistics as they aim to achieve consistent data homogeneity and interoperability. We conceive Digital Twins as a collection of Aspects, where the Digital Twin establishes identity by representing a specified asset and the Aspects provide a domain-specific view on this asset. For the Asset Administration Shell, the implementation of a digital twin for Industry 4.0, an Aspect corresponds to a submodel.

Technically an Aspect is a software service that offers functionality and data related to the represented asset. Each Aspect references a concrete Aspect Model, which formally describes the data that is provided by the Aspect. For the Asset Administration Shell this linkage between aspect and aspect model is done via the semanticId of a submodel. A meta model is required allowing to express both schema information to perform data validation and domain semantics to make implicit information explicit. The Semantic Aspect Meta Model (SAMM) is a semantic meta model for defining the semantics of a submodel via an aspect model.

An Aspect Model contains both information about the runtime data structure (e.g., that there is a property in the data called "temperature", and that it has a numeric value) and information that is not part of the runtime data (e.g., the physical unit and the value range). It does not, however, contain actual runtime data (e.g., a numeric value representing the current temperature), as this will be delivered by the submodel of the digital twin.

The Eclipse Semantic Modeling Framework (ESMF) project develops accompanying tooling to create, manage, and leverage Aspect Models. This helps to support manufacturing organizations in setting up an environment to make data-driven decisions that can help manage risk, optimize production, and adopt and explore new revenue streams or business models.

The bridge to the Asset Administration Shell is provided by already existing importers in the Eclipse AASX Package Explorer as well as AASX command line generators in the SDK of this project.

Known standards such as IEC 61360 and international data dictionaries like ECLASS [1] and IEC CDD [2] but also existing ontologies can be used as basis for building aspect models. 

Aspect Models express a schema with a defined Resource Description Framework (RDF, [3]) vocabulary and are validated by a comprehensive set of rules in the Shapes Constraint Language (SHACL, [4]). Domain semantics are captured by a combination of structural elements, relations, namespaces and reified named concepts. In comparison to OWL [5], SAMM favors aggregation over inheritance in modeling, thus enabling domain experts without background in ontology engineering to create and maintain Aspect Models.

There will be close relationship to other activities that also aim for open source approaches, for example Industrial Digital Twin Association (IDTA), Gaia-X, International Data Spaces Assocation  (IDSA), Eclipse BaSyx, Catena-X (Eclipse TRACTUS-X), Manufacturing-X and the Digital Twin Consortium. 

References

1. ECLASS https://www.eclass.eu/, last accessed 2021-07-28
2. IEC 61360-4, Common Data Dictionary (CDD - V2.0014.0017).
3. Cyganiak, R., Wood, D., Lanthaler, M.. RDF 1.1 Concepts and Abstract Syntax.
4. Knublauch, H., Kontokostas, D., Shapes Constraint Language (SHACL).
5. Bock, C et al.: OWL 2 Web Ontology Language Structural Specification
6. Catena-X. https://catena-x.net/de/ last accessed 2021/07/28