Topics on Digital twins: Hybrid modeling, befitting cross validation and dynamical systems

The talk is designed to provide an introduction to digital twins, mainly in the context of system and industrial applications. Three other topics will be covered which provide directions for further research. These include: i) hybrid modeling, ii) befitting cross validation and iii) dynamical systems. Hybrid modeling combines engineering based mechanical models with phenomenological models: Typically, mechanistic models are developed from first principles based on some information about a system of interest. This is in contrast to phenomenological models, which are data-driven in nature, viewing the system as a black box. Hybrid modeling is the integration of the two modeling approaches. A second topic relates to context dependent cross validation of predictive analytic models. Specifically, a methodology labeled befitting cross validation (BCV) will be presented. BCV is particularly suited to validating digital twin models. A third topic to be covered is modeling dynamical systems with deterministic ordinary/partial differential equations. Mathematical and statistical models permit to measure and predict the state of a process in time and space, and enable the development of monitoring, diagnostic and prognostic capabilities for process control purposes, as well as for efficiency and predictive maintenance implementation. These are the main analytic capabilities of digital twins. Contatto: piercesare.secchi@polimi.it

Professor Ron Kenett is Chairman of the KPA Group, Israel, Chairman of the Data Science Society at AEAI, Senior Research Fellow at the Samuel Neaman Institute, Technion, Haifa, Israel and Research Professor at the University of Turin, Italy. Past president of the Israel Statistical Association and the European Network for Business and Industrial Statistics. Authored and co-authored over 250 papers and 16 books on applied statistics topics. Awarded the 2013 Greenfield Medal by the Royal Statistical Society and, in 2018, the Box Medal by the European Network for Business and Industrial Statistics. PhD in Mathematics from the Weizmann Institute under supervision of Sam Karlin and BSc. in Mathematics, with fist class honors,from Imperial College. Ron is emphasizing the role of information quality as a main objective of statistical analysis