cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
JMP is taking Discovery online, April 16 and 18. Register today and join us for interactive sessions featuring popular presentation topics, networking, and discussions with the experts.
Simulation Experiment

 

In this JMP in Action segment, JMP Systems Engineer Robert Anderson takes us through two alternative approaches to optimizing a process or product: maximized designability and simulation experiment. First, Robert demonstrates how to build a model and to view and manipulate it within the Predication Profiler. Next, Robert show you how JMP can run a series of simulations on critical factors to allow you to identify the settings that minimize the defect rate. Finally, should you want to try simulation experiments on your own, Robert shares where to go to get more information.

Comments

@robert_anderson  thank you for this lovely presentation, i especially liked the added visualization (Scatterplot 3D and Graph Builder with Contours) tools that you used to show the simulated values across the 3-D factor space. actually this is a more in-depth version of another analysis that one of your colleagues Byron Wingerd gave awhile back on the Mastering JMP series that was really great as well.  https://www.jmp.com/en_my/events/ondemand/mastering-jmp/tracking-and-trending-manufacturing-metrics....

 

Patrick, thanks for your nice comments. I agree that Byron does a good job of explaining it in his Mastering JMP recording. I believe that Brad Jones and John Sall were involved in developed this really useful capability in JMP many versions ago. I've done a Discovery talk on the same subject if you want to see slightly longer discussion of the particular example I used: Robust Optimisation of Processes and Products by Using Monte Carlo Simulation Experiments

Wow! @robert_anderson absolutely thank you so much for the reference, I look forward to diving into this further.  This is a very useful tool in mfg. process optimization in any industry: I've been using some of the basic simulation capabilities more recently in my work in medical device manufacturing and I thoroughly enjoy the "click and simulate" capabilities that JMP offers (very user-friendly compared to "spreadsheet driven" Monte Carlo simulation add-in software for MS Excel for example). - @PatrickGiuliano 

Thanks @PatrickGiuliano , I agree, there are lots of really useful things in JMP. You might also want to check out this white paper by Bernd Heinen https://www.jmp.com/en_in/whitepapers/jmp/pharmaceutical-quality-by-design-methods.html

Excellent walkthrough of the simulator with some great tips - thank you!

Thanks Julia, I'm glad you found it helpful.

Gilly

@robert_anderson great demo, very useful. Many Thanks

Thanks Gilly, that's good to hear. 

Excellent and clear presentation Robert. Thanks very much for demonstrating this - another feature to keep in the back of my mind now!

 

Best,

Greg

Thanks Greg, much appreciated

ar2

Great Demo Robert. I will certainly build the simulation experiments into my  training modules!

Thanks Andrew, glad you will be able to make good use of this.
Usman

Great demo!

Thanks Usman

Marco1

Hi Robert Anderson,
I am pleased to greet you, excellent explanation! Regarding the "Noise Factor" option located in the graphic menu, when using it...would you get the same result?
Regards,

Marco

Marco1

Hi Robert Anderson,

A query,....... when using the "Simulation experiment" option and specifying the distributions Triangular for dollar price, Uniform for change dollar, Uniform for risk, indicate in "Specification limits" ....total change : LSL = 400 USL= 3600, total risk: LSL = 0 USL= 90, total change: LSL=3.36 USL= 3.9, indicate 128 experiments, 1 (full range) and run the simulation experiment.... JMP uses the ranges of the "exchange of dollars" table.......lines below I detail the procedure with images.....what am I doing wrong?

Regards,

Marco

Marco1_0-1673210593964.png

Marco1_1-1673210671888.png

 

Marco1_4-1673211025241.png

Marco1_7-1673212600156.png