A Python based Simulation of Industrial Scenarios - An investigation of throughput time

Abstract

The goal of the thesis is to find a simulation package suitable for industrial application, create a model, run simulations with it and investigate the goodness of the choice. This thesis uses the object-oriented programming language Python, with additional package salabim to conduct a discrete simulation.

The package was chosen with the idea that Industry4.0 will be the future of production. It was briefly explained as well as the different simulation packages found suitable. The right package was found using approaches described by various literature sources. To find the right model, the right type of simulation had to be found. To find the right type of simulation the topical case had to be found. The case was given to the author from SINTEF Raufoss Manufacturing, and as many other manufacturing cases, discrete event simulation was suitable. An explanation of discrete event simulation, statistical distributions, scheduling and parameters was done to investigate if the chosen package could manage to simulate it. A part on how to conduct simulations was found necessary for the reuse of this thesis. Results from running different scenarios shows that the choice of package displays the information in an organized and desired manner. The throughput time and standard deviation varied much like expected. Thus concluding it as a successful approach, worthy of reuse.

The examination of different scenarios led to the conclusion that the package chosen was suitable for industrial scenarios. It is considered a well-documented, easy to learn approach of both object-oriented programming and simulation. It is concluded as a middle way between visual simulation and fully statistical simulation.