This presentation by Giovanni Borzi was made at the NAFEMS eSeminar "Simulation & Digital Twins - Behind the Buzzwords" on the 2nd of May 2018.

A Digital Twin may be defined as an operational virtual model of a physical component or machine. As such, the idea of Digital Twin has been around for several years, especially in the Aerospace and Automotive sectors. What truly differentiates a Digital Twin from a simulation model is the area of application, that is shifted towards operations and servitization, and the integration with IIoT technologies or automation design tools. Referring to a simulation based model of a production machine, for example, the requirements may include the ability to detect issues with the real machine counterpart throughout its operating life, or the ability to enable strategies for compensating decreased performances without slowing a production line. Other than predictive maintenance, digital twin enabled servitization may entail optimization or zero defect manufacturing strategies based on process or machine models.

To exemplify these scenarios, two different approaches to Digital Twin development will be presented. A model based approach for the development of Digital Twin of production machines will be described, whereas a model of the machine is created and afterwards integrated with automation design software using the FMI technology in order to provide virtual commissioning capabilities. A second example will be presented whereas a process simulation model is used to create a surrogate process model (meta-model), that is afterwards used in combination with data recovered by sensors and HMIs in order to provide production quality prediction and control capabilities. In this latter example implications regarding the management of data will also be presented.