Discrimination of hidden information in measurement results

In a growing economy, standardization promotes the exchange of products, people and processes but it is during industrial crises that standardization fosters the economic recovery. The Geometrical Product Specification (GPS) Project aims to develop a new, universal, univoque and mathematical based language to express the geometrical variability of manufactured products. The knowledge of GPS standards not only provides a competitive edge to manufacturing enterprises, which design, produce and control products with higher quality, but also increases the competitiveness of a whole industrial district or country promoting the cooperation between enterprises. GPS explores components of uncertainties, from the specification to the verification phase. Even if the GPS project provides a description of variability along the product life cycle, it is significant to discriminate hidden information on measurement results, in order to gain the most quantity of information into the manufacture variability. Methods to discriminate information are analyzed for macro geometry with a partitioning probabilistic method and for profiles with independent component analysis techniques.