Use cases: FMB Digital Twins

Automation is innovation at the service of industrial plants: Process intelligence and Digital Twin

Organizations involved

Three manufacturing industries: Ocrim S.p.A., Desmet Ballestra S.p.A., a dairy industry and a solution provider, spin-off of the University of Parma: FMB – Engineering Innovation for Enterprise S.r.l.

Objectives obtained

The real model of the production processes is monitored in real time by equipping machinery and plants with an ad hoc designed sensor system. The continuous monitoring of the process phases allows the creation of an appropriately validated dataset, which represents the material for comparison with the virtual model (digital Twin), capable of reproducing the behavior of the real plant through advanced tools, such as process simulation , CFD simulation and complex calculation models. Thanks to process modeling, the digital twin will have a reliable and constantly updated response surface available, capable of representing a real guide for regulating the process. The data collected in the field is analyzed by a supporting Big Data Analytics solution. of decisions, capable of detecting deviations between ideal and real operating conditions, deciding on the implementation of any corrective actions.

The challenge

The three manufacturing companies in question, despite belonging to different industrial sectors, had the same problem: increasing the efficiency of the production and maintenance process. This challenge was faced by entering into a relationship with an authoritative and recognized solution provider capable of interacting with competent research centers and discovering the enabling technologies necessary to obtain the expected results.

The solution

The three companies Ocrim S.p.A., Desmet Ballestra S.p.A. and a dairy industry came into contact with FMB – Engineering Innovation for Enterprise S.r.l., a company that deals with research and development from a plant and process point of view, founded in February 2014 as a spin-off of the University of Parma and based at the Campus of the same University which, valorising its industrial research and development experiences, after a careful analysis has proposed in all cases to apply digital twinning techniques.

Digital twin is to be considered a virtual copy, a model, of a real physical asset in operation capable of reflecting the current condition of the asset, including in the analysis series of historical data characteristic of its operation.

The Real model of the three industrial processes is now monitored in real time, through an ad hoc supervision system, capable of detecting and historicizing the signals coming from specific field sensors. In all cases the methodology is the same, while the application in the field is necessarily different. In any case, the dataset, appropriately validated, is material for comparison with the Virtual Model (digital Twin), which operates on the basis of a numerical model created through the application of dedicated software or the programming of generic tools. The data collected in the field is then analyzed by a Big Data Analytics solution to support decisions, capable of detecting deviations between the ideal operating conditions and the real ones, deciding on any corrective actions to be implemented.

For a better understanding of the techniques and application details, we refer to a presentation that analyzes and describes the various cases listed above, expressing useful considerations on the techniques used.

Are you interested in adopting this methodology?