Virginie Fortineau
Virginie Fortineau has a total of 4 co-authors. Here are the top co-authors by number of shared publications:
- Samir Lamouri
LAMIH-UMR CNRS 8201, Arts et Métiers Sciences et Technologies, Institute of Technology, 75013 Paris, France
2 shared publications - Asma Talhi
Khtema sas, Lamih, CNRS, Arts et métiers, ParisTech
1 shared publication - Jean-Charles Huet
EFREI Research Lab, Université Paris-Panthéon-Assas, Villejuif, France
1 shared publication - Thomas Paviot1 shared publication
Latest Publications
- Open Access
- 5 June 2026
- Published by Elsevier in SSRN Electronic Journal
Symbolic AI for Semantic Interoperability Verification Across Lifecycle Bill-of-Materials Views
This paper introduces a symbolic AI workflow that combines knowledge representation (geometric theories, bridge morphisms) with automated reasoning (SMT solving) to provide machine-checkable verification and obstruction diagnosis for semantic correspondences between heterogeneous data models.The workflow targets semantic interoperability across engineering, manufacturing, and service Bill-of-Materials views in product lifecycle management, where data exchanged between design, manufacturing, and maintenance platforms may be structurally correct yet semantically wrong. Existing schema-matching, ontology-alignment, and model-transformation approaches help construct correspondences, but they rarely provide formal evidence that a transferred fragment is semantically valid or pinpoint why a mapping fails. The proposed method encodes schemas as geometric theories, correspondences as bridge morphisms, and uses SMT solving to decide whether a lifecycle-data fragment preserves meaning under stated applicability conditions or, when it does not, to localize the obstruction.Three layers are evaluated: a pedagogical Part--Material example establishing the formal mechanism; a curated aerospace-style case study covering eBOM/mBOM/sBOM fragments with 16 entity sorts, 77 function symbols, and 33 theory-internal axioms; and a public NIST benchmark used as a reproducible bridge-coverage study on open lifecycle artifacts. The solver-backed campaign comprises 30 verification checks. On the curated industrial case, median runtimes range from 2.8 ms for obstruction checks to 204.9 ms for full consistency, with all checks below 1 s. On the public fragment, the method supports exact manufacturing-to-inspection correspondence for 56 of 60 runs (93.3%) and localizes four unmatched runs to identifier-level obstructions. These results show that symbolic AI can deliver verifiable semantic-assurance evidence on bounded lifecycle-integration corridors.
- 5 June 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessSymbolic AI for Semantic Interoperability Verification Across Lifecycle Bill-of-Materials Views
This paper introduces a symbolic AI workflow that combines knowledge representation (geometric theories, bridge morphisms) with automated reasoning (SMT solving) to provide machine-checkable verification and obstruction diagnosis for semantic correspondences between heterogeneous data models.The workflow targets semantic interoperability across engineering, manufacturing, and service Bill-of-Materials views in product lifecycle management, where data exchanged between design, manufacturing, and maintenance platforms may be structurally correct yet semantically wrong. Existing schema-matching, ontology-alignment, and model-transformation approaches help construct correspondences, but they rarely provide formal evidence that a transferred fragment is semantically valid or pinpoint why a mapping fails. The proposed method encodes schemas as geometric theories, correspondences as bridge morphisms, and uses SMT solving to decide whether a lifecycle-data fragment preserves meaning under stated applicability conditions or, when it does not, to localize the obstruction.Three layers are evaluated: a pedagogical Part--Material example establishing the formal mechanism; a curated aerospace-style case study covering eBOM/mBOM/sBOM fragments with 16 entity sorts, 77 function symbols, and 33 theory-internal axioms; and a public NIST benchmark used as a reproducible bridge-coverage study on open lifecycle artifacts. The solver-backed campaign comprises 30 verification checks. On the curated industrial case, median runtimes range from 2.8 ms for obstruction checks to 204.9 ms for full consistency, with all checks below 1 s. On the public fragment, the method supports exact manufacturing-to-inspection correspondence for 56 of 60 runs (93.3%) and localizes four unmatched runs to identifier-level obstructions. These results show that symbolic AI can deliver verifiable semantic-assurance evidence on bounded lifecycle-integration corridors.
- 30 November 2017
- Published by Springer Nature in Journal of Intelligent Manufacturing
- Vol. 30 (5) , 2171-2192
Ontology for cloud manufacturing based Product Lifecycle Management
- 30 November 2017
- Published by Springer Nature in Journal of Intelligent Manufacturing
- Vol. 30 (5) , 2171-2192
Ontology for cloud manufacturing based Product Lifecycle Management