The European Metrology Network for Mathematics and Statistics was created in 2018 by European metrology institutes committed to the development and application of mathematical and statistical methods for metrology.
Mathmet is a “backbone network”, and its expertise is applied to a number of fields, based on the requirements of the other metrology networks and the needs suggested by the relevant stakeholders.
Mathmet aims at providing a point of reference for metrology institutes, standardization and accreditation bodies, academia, and industry, on mathematics,statistics and computational science in metrology. Mathmet provides guidelines, develops computational codes, facilitates the cooperation of its members within scientific projects, and supports the dissemination of the knowledge through scientific publications, conferences, and didactical resources.
Specifically, this network plays a main role for the integration between measurement science and the mathematical and statistical methods, to address current key challenges, like energy sustainability, health, and climate and environmental monitoring. In these fields, the network exploits its experience on mathematical modelling and uncertainty evaluation to investigate innovative scenarios, where multivariate and complex processes take place, and where advanced algorithms become an integral part of the measurement instruments, giving rise to peculiar uncertainty contributions.
Currently, special attention is paid to new and cross-cutting topics, as digitalization, big data, and artificial intelligence, with the aim to quantify the level of trustworthiness of the corresponding tools. At the same time, the network develops and characterizes advanced modeling tools (e.g. numerical multiphysics or multiscale models) suitable to perform virtual experiments useful to aid the interpretation of empirical observations, or to replace measurements that may be too complex, costly, or invasive.
One of the specific objectives of the network is the preparation of digital twins, i.e. virtual models for the prediction of the behavior of real-world systems for which it is possible to acquire experimental information.
INRiM, which currently has the chairmanship of the network, contributed to the foundation and the development of Mathmet, and participated actively to the preparation of the EMN strategic research agenda and to the development of quality assurance tools for software, datasets, and guidelines.
INRiM organizes the periodic workshop Mathematical and Statistical Methods for Metrology (MSMM), which is one of the main Mathmet-related scientific events, open towards the whole scientific community. Moreover, INRiM participates at the MU Training Activity, aimed at supporting didactical activities in the field of uncertainty evaluation.
Within Mathmet, INRiM’s mathematical and statistical skills are applied to a number of sectors, including the harmonization of statistical approaches and the standardization of the terminology for interlaboratory comparisons and conformity assessments in analytical chemistry, the development of Bayesian approaches for calibration issues, conformity assessments and sampling procedures, the uncertainty evaluation for distributed sensor networks, the investigation of therapeutic and diagnostic techniques, as hyperthermia and magnetic resonance imaging, the development of quantitative medical imaging methodologies, the study of nanomaterials and nanodevices for precision medicine, the evaluation of human exposure to electromagnetic fields, and the validation of artificial intelligence systems in diagnostic imaging.
