Integration of Artificial Intelligence in the life cycle of industrial Digital Twins

Abstract

Digital twins (DT) constitute a major concept of future industrial systems. They are expected to enable efficient virtualization of manufacturing systems and enhance various decision-making processes. In parallel, many initiatives exhibited how artificial intelligence (AI) could increase the performance of the DT on specific applications. By reviewing the literature combining AI and DT, a lack of contributions on the whole life cycle of the DT was exhibited. Therefore, the main contribution of this paper is to define a global integration framework of AI into DT, focused on the exploitation phase of the DT. A case study, using a relatively simple physical twin, illustrates the potential of such integration for the response of the DT to unpredictable modifications of the physical twin.