ESTEQ Discrete Event Simulation use-case at SIEMENS.2014-04-01
SIEMENS has recently completed a use-case of the application of Discrete Event Simulation (DES) by ESTEQ using their software; Tecnomatix Plant Simulation.
ESTEQ uses Plant Simulation to provide planning services and reduce costs.
DES is used to model and analyse the process flow and logic of any process-driven environment. Using Plant Simulation it is possible to understand how a whole system lives and breathes as a whole.
The use-case gives an example of how much value DES can potentially have as well as showing the quick turnaround time of building models and getting results in an ideal environment. It concludes with a brief explanation of some of the advanced tools that can be deployed.
The case-study involves an automotive customer with a CAPEX to acquire additional resources within a skid system costing approximately R200,000.00 per skid. The question was how many skids should be in the system and within three days the model was set up due to all the correct data being available. Figure 1illustrates the experiments wherein one additional skid was added into the system with each run. Experiment 1 (Exp 1) was the base model from which it can be seen that the addition of a skid increases the output (y-axis) of the system until Experiment 5 in which the output ceases to increase. It was determine that only three additional skids were needed with the original estimate of six implying a cost saving, or rather, cost avoidance of R600,000.00 in three days.
Plant simulation can solve multiple problems using built in tools such as genetic algorithms which can easily sequence very complex alternative combinations. Other tools such as the Sankey Diagram and Bottleneck-Analyzer makes it possible to look at the over-all status of the system in a glance. Plant Simulation is very proficient at handling variation in large amounts of statistical data which is traditionally difficult due to the the amount of computations required increasing exponentially with every variable.