# Desktop Client: How do I simulate a Lean VSM to show variability in process times or queuing?

Simulation, such as Discrete Event Simulation (DES), is not possible in a Lean VSM type diagram. Simulation is available in the iGrafx Process or Process for Six Sigma products, using a Process or BPMN type diagram.

A Lean VSM diagram is focused on averages or the current state observed from the factory floor (Gemba), and even hard-codes inventory levels to a specific amount. Using a Lean VSM for analysis helps focus on exposing the fact that there’s NVA or muda-- such as inventory-- in the first place, not how the waste such as inventory varies, nor in fluctuations in processing time.

However, that said, there are two other methods available to you with iGrafx that may be useful:

- Export the (average) times you’ve captured from the VSM into Excel. From the Lean menu, choose Export VSM Table to Excel, and then do your analysis in Excel. For example, you could enter a standard deviation for processing times in the Excel sheet, and use Excel functions (and/or the 'solver' optional add-in) to get a range of cycle times.
- Use the iGrafx simulator to do a variability type of analysis, handling 'real-world' variations in the process behavior. Use a Process type diagram, draw the material flow process (or copy-paste shapes from the VSM into a Process type diagram), enter the processing times as variable data, run simulation, and analyze the iGrafx report or even log the detailed results to Excel (or a statistical application if using Process for Six Sigma) if desired. For example, iGrafx can handle mean and standard deviation in Task page Duration times by using the NormDist() function in a duration expression. There are many different statistical functions that may be used, including Weibull, LogNormal, Exponential, etc. If you have iGrafx Process for Six Sigma and a statistical package (i.e. Minitab or JMP), you can even fit empiracally-measured data into one of 5 different statistical curves for use by your simulation model.