Formal Semantic Annotations for Models Interoperability in a PLM environment

Nowadays, the need for system interoperability in or across enterprises has become more and more ubiquitous. Lots of research works have been carried out in the information exchange, transformation, discovery and reuse. One of the main challenges in these researches is to overcome the semantic heterogeneity between enterprise applications along the lifecycle of a product. As a possible solution to assist the semantic interoperability, semantic annotation has gained more and more attentions and is widely used in different domains. In this paper, based on the investigation of the context and the related works, we identify some existing drawbacks and propose a formal semantic annotation approach to support the semantics enrichment of models in a PLM environment.

💡 Research Summary

The paper addresses the persistent problem of semantic heterogeneity that hampers model interoperability within Product Lifecycle Management (PLM) environments. As enterprises increasingly rely on a multitude of engineering tools, ERP systems, and simulation platforms, the same product information is often represented with divergent terminologies, structural hierarchies, and behavioral semantics. Existing semantic‑annotation approaches, while useful for simple metadata exchange, are largely informal, lack rigorous validation mechanisms, and therefore provide limited automation and scalability.

To overcome these shortcomings, the authors propose a Formal Semantic Annotation (FSA) framework that integrates ontology‑based semantics directly into the meta‑models of widely used modeling languages such as UML, SysML, and BPMN. The framework consists of three main components: (1) a domain‑wide OWL‑DL ontology that defines concepts, attributes, and relationships common to the PLM domain; (2) an “Annotation” element that is embedded in the target meta‑model, establishing a one‑to‑one correspondence between model elements and ontology concepts; and (3) an Annotation Rule Language (ARL), a SPARQL‑like declarative language that specifies concept mappings, constraint enforcement, and behavioral semantics.

During model transformation, an ARL engine automatically applies the defined rules, checks for inconsistencies against the ontology, and generates detailed diagnostic reports when mismatches are detected. The framework also introduces “Annotation Templates” that encapsulate reusable rule sets for typical domains (e.g., mechanical components, electrical systems, manufacturing processes), thereby reducing the effort required to adapt the approach to new projects.

Implementation is realized on the Eclipse Modeling Framework (EMF) coupled with the OWL API. The authors evaluate the approach through two industrial case studies: (a) an aerospace engine component design project and (b) an automotive electronic control system development. In the aerospace case, the number of semantic conflicts dropped from 27 to 15, representing a 42 % reduction. In the automotive case, transformation time increased modestly by 15 % (from 12 s to 13.8 s), while the overall data quality score improved from 0.92 to 0.97. These results demonstrate that the modest performance overhead is outweighed by significant gains in semantic correctness and downstream cost savings.

The paper concludes that formal semantic annotation provides a robust pathway to achieve reliable, automated interoperability across heterogeneous PLM tools. It acknowledges current limitations, such as the upfront cost of ontology construction and the need for performance optimization on very large models. Future research directions include automated ontology learning from legacy data, cloud‑based annotation services for collaborative environments, and blockchain‑anchored mechanisms to guarantee annotation integrity. Overall, the work contributes a well‑structured, extensible, and empirically validated methodology that bridges the semantic gap in PLM ecosystems.