International audienceThe authorities of the standards organization International Organization of Standardization (ISO) advocate mastering any uncertainties in all parts of the industrialization process. In the three-dimensional (3D) measurement process, uncertainty is usually obtained at the end of a battery of tests. It is defined as a whole because it includes several types of errors, known systematic components, unknown systematic components and random components. Automated calculations of uncertainty can be made based on statistics. This method is based on statistical concepts, which are in accordance with " The Guide to the expression of the uncertainty in measurement " (GUM). It also enables us to generate uncertainties on the verification of ISO specifications (or specs in the ISO directives). In the course of this article, a usage will be presented that takes the knowledge of uncertainties into account: this usage will help the operator to take a decision on the conformance of a mechanical part in reference to its conformance to geometric tolerance

International audienceIn the scope of quallty control, accurate evaluatlon of measurement uncertalntles is a real challenge to Improve the use d Coordlnate Measurlng Machlnes (CMM). In our work, a new method, based on a statistical approach of the problem, was therefore developed, to deduce instantaneous measurement uncertainties directly from the set of acquired coordinates. The covariance matrix of the intrlnsic parameters whlch characterize each analyzed surface Is also evaluated, thus allowlng an accurate propagatlon of the measurement uncertainties to the ISO specifications to be controlled. The experiments carried out in our study illustrate thls new statistical approach and demonstrate its relevance

No abstract