Topic 1: Product Specification and Verification

(Team led by C.M. Souzani)

This theme concerns, on the one hand, the formalization of 'multi-physics' requirements: Operators, Operations, Entities, Uncertainties

Approaches and tools used:

➢ Laplace-Beltrami Operator
(multi-scale descriptors, sampling in shape space, meta-operator for GPS)

➢ Geometric Algebras
(situational elements: ToLiP, generalization of SATTs: Multi-ToLiP)

➢ Polyhedral Approaches
('exact' numerical computation, open-source software)

On the other hand, it involves progressing in the semi-automatic generation of tolerancing specifications (life cycle approach)

➢ Knowledge formalization (ontologies, category theory)

➢ Integrated Tolerancing (Generative Design under uncertainty, Functional Optimization)

➢ Machine Learning (Graph Neural, ChatGPS)

Finally, we will focus on the digital processing of measurement data for the reconstruction and identification of shapes/entities/products.

➢ Filtering by composition and decomposition
(contour/shape deformation, image/volume correlation (VIC/VVC), operator spectra, Physics-Informed Machine Learning)

➢ Single/multi-sensor multi-scale fusion
(reference algorithms, software standards, explanatory AI)

➢ Data/model coupling for reconstruction
(updating part and assembly JN, semantic reconstruction, formalization of the reconstruction operation – ISO GPS standard)

Highlights:

Automatic Multi-sensor Inspection Planning