Simulation

[jamtilimpact]

I am the only continuous improvement/lean/mfg engineer at my facility and pretty new to the job (9 months). I'm wondering if anyone has a preference in simulation software. I used Arena quite a bit in college and have been tinkering with a VERY limited demo version of FlexSim lately. I have been whispering in my bosses ear that a simulation program is cheaper and easier than disrupting production and rearranging cells to try an idea that may or may not work. While this may have value for some, the disruptions and the (often) scrap paint a different picture. Like anywhere else, resistance to change here is fierce. I think if everyone could see a pretty 3-D animation of the future state they would be more readily convinced that there is a better way after all.

 

[tripleZ]

I've used both ProModel and Arena. Personally, I'm more comfortable with ProModel and have an easier time nesting programs with that platform.

A word of warning on sim programs though. The time study data needed is pretty complex, and is fairly time consuming if you're going to do it right. The time you spend collecting all of this data might be worth it. However, you might see more immediate improvements by focusing on bottleneck improvements and problem operations. Case in point: I've got a molding operation that has run well for the past year. But I'm tapped on how much faster I can make the press go. I can play all sorts of computerized balancing games to optimize & coordinate the line, or put the effort into a riskier double cavity design. The latter worked, and is netting the company around $175K per year in cost savings, not to mention capacity.

If they're not receptive to change, work within the system before trying to reinvent it. When your boss sees your ideas working, he'll likely be more receptive to suggestions. Just my 2 cents.

 

[jck26]

Use what you know since there are no other users. Make damn sure that the data you load the system with can very closely match the actual operating conditions that you have on the floor. If you enter in an actual set of present conditions it must correlate to what you have or those who can't understand the concept of simulation will never be on board with what you are going. Never use the phrase "game theory" in you discussion of simulation because the negative crowd wants nothing to do with any methods that use the word "game" in the description.

Before you show the results of any work to a group that is new to simulation prepare a simple demo of what the software does. Don't get deep into the technical points of the software or the concepts of simulation for fear that you will lose most people with the exception of other IE's and technical types.

To get back to you original idea, this does make an excellent method to demonstrate how changes impact the shop floor when considering changes but it can get out of hand if too many people want to add their inputs to the model. Keep this method as a tool of engineering until people are more conversant of the features of the system and you have learned the ins/outs of the software tool yourself.

jck26

 

[tygerdawg]

I did a lot of manufacturing improvement projects with simulation way back when (DOS-based SIMAN, then later Witness). So that makes me Old-School. Besides, I'm just a dumb Mechanical Engineer, so I don't know the nuances or rigor of my Industrial Engineering colleagues. Ignorance is bliss, I suppose.

I would propose to you that you DON'T need to go to the N-th degree on a manufacturing simulation project because the work environment is so dynamic. I agree with the other guys: bottleneck elimination would be the easiest route to improvement. In these situations, incremental improvements are much more palatable to managers than "Grand Scheme" projects. I wrote a paper on this methodology years ago and my premise was very well received at several trade shows and professional society venues by "dumb MEs" and rigorous IEs alike. Well, except the Academics...they thought I was the Anti-Christ or something. I can send you a copy if you email me at my blind email address tygerdawg@excite.com .

Pretty 3D animations are one thing, but as a Production / Engineering Manager, I would be more impressed with:
(1) realistic data with payback analysis
(2) a plan to do your proposed improvements that cause minimum diruption
(3) whether the cost of implementing your changes are included in your payback analysis.

I was going to get back into this game as a consultant a couple years ago, and I was going to choose Simul8 since it did everything I needed, plus was cost effective. I did not have the funds to spend $15K on a software package.

TygerDawg