The background for this thesis stems from a wish to move the production a step further towards Industry 4.0 at Siemens Energy Trollhättan. They wish to start making this journey by researching the possibilities of simulating their production flows and extending that simulation into a Digital Twin, with alarms helping to optimise their production. In this thesis, the focus will lie on the Digital Twin towards implementation of Real Time Data within that, while the focus on alarms will be handled in the partnering work performed by Niklas Hansson [1].

The simulation model produced within this degree work came to encompass one of the initially six suggested production flows at the Siemens factory. The data has been prepared so the expansion of the model can proceed without any major difficulties. According to theory, the model has been proven to have a high level of accuracy, the throughput only missing the calculated one with five percent. This means that the simulated model behaves in a way closely related to the actual factory floor.

One of the most interesting results found by experimenting with the simulation is that if too many products are initialised in the system the output will actually decrease. Beyond experiments pertaining to the inflow of products, other parameters were tested as well. The chosen parameters to test were changes in sick leave, size of the workforce, scheduling, and buffer sizes. The scheduling, and as mentioned before, the inflow, were the parameters that most greatly affected the yearly production, parameters such as sick leave almost having no visible effect at all. This is thought to be because of the small scale of the current simulation, if more production flows would have been added, a greater effect would have been seen. Although no clear solution for the task of implementing Real time with a simulation was reached within the timeframe of this degree work, a solid understanding of what needs to be further researched and implemented has to be summed up in the thesis.