Data-driven invariant modelling patterns for digital twin design

The Digital Twin (DT) is one of the most promising technologies in digital transformation market. A twin a virtual copy physical system that emulates its behaviour to predict failures and opportunities for change, prescribe actions real-time, optimise and/or mitigate unexpected events. Modelling rather complex task requires availability large amount information set accurate models adequately represent reality model. At present, modelling depends on specific use case. Hence, need design solution suitable context twin. paper proposes new approach DT by endeavouring concept "modelling patterns" their invariance property. patterns are here thought as data-driven, they can be derived autonomously from data using devised reach an feature, allow these used (and re-used) situations problems with any given degree similarity. potentialities proved grace real industrial application, where dedicated has been built proposed here.