Optimizing Batch Sizes 1

[McKagan13]

I am working on multiple projects within my cell to reduce WIP inventroy, lead time, increase productivity etc. Im sure all you manufacturing engineers are constantly doing the same thing. Let me preface by saying my degree is Mechanical engineering and I have limited manufacturing experience.

In any event, I would like to examine the possibility of changing our batch sizes to optimize efficiency/performance. We have a great product mix (means change over) as well as inconsistent batch sizes between parts. In any event, I am looking for a model or some advice on batch size optimization. Can anyone point me in the right direction?

 

[MikeHalloran]

My machinist friend Elmer said he loved rattling around alone in a sizable machine shop in a shrinking company, because he left multiple machines set up for different kinds of jobs. His setup time was essentially zero, so he had an economical batch size of one.

[ That of course neglects the cost of providing more machinery than he could use at once, because it was already there. It also neglects the pennies on the dollar that an aggressive accountant would recoup by selling off the surplus machines, or the astronomical cost of buying them back from the tool dealer should the company recover. ]

Similarly it's possible to set up CNC machines with flexible tooling or idle stations on large beds to also get zero setup time and an economical batch size of one.

Beyond that, you can invest in fast-change tooling, e.g. robotic die changers for large stamping presses, to reduce the setup time to just a few minutes ... if you have the budget for the automation.

You can probably drive yourself crazy reading books about this stuff, but major changes require major money and major backing from management, both factors that are often lacking.

In the meantime, it might be productive to break down the processes as they are now, beat on them in a spreadsheet, and look for ways to make incremental improvements. That exercise may eventually reveal the actual bottlenecks where it would actually be productive to spend money.




Mike Halloran
Pembroke Pines, FL, USA

 

[tygerdawg]

I have similar ME / MFG background except a lot of IE courses and training. I have been called a Heretic by my IE brethren, but I used IE methods in a non-sophisticated way to great effect. While formal IE methods are powerful & useful, my company and job situation would not allow me to get into heavy analysis. There were so many opportunities for improvement in this organization, that it was all I could do just to hit that 80% improvement target and move on to the next raging fire. It worked out well and I looked like a Hero to the bosses. The real power of sophisticated IE methods kicked in for the last 20% of the improvements.

I used Discrete Simulation software to do quickie analyses to model manufacturing operations and determine bottlenecks, batch sizes, evaluate alternative production line layouts, etc. Years ago I was so successful with that method that I published a paper on it and it was well received. You can download it here

http://www.bluetechnik.com/ImprovedOperationsThruSimulation.pdf

. If you want to discuss, then contact me through the website.

The last I looked, there was a inexpensive and powerful little simulator package

http://www.simul8.com

that may serve your purposes well.

TygerDawg

 

[mrainey]

This is a very inexpensive program I wrote that will help a non-machinist to quickly lay out and understand the times and costs involved with a CNC operation, at various lot sizes.

http://mrainey.freeservers.com/ProJobPlanSummary.html

Software For Metalworking

http://mrainey.freeservers.com

 

[PSE]

Take a look at "The Goal" by Eliyahu Goldratt

http://www.goldratt.com/

You can also find it in regular bookstores. There may be a few things in there you will find of interest in terms of dealing with process improvement, bottlenecks, and subsequent effects.

Regards,