Monday, March 15, 2010

Where's the SMED?

The SMED system which stands for Single Minute Exchange of Dies, developed by Shigeo Shingo and published in his 1985 book, guides us to achieve a machine setups in under 10 minutes. The quick changeover thinking has been around for a couple of decades now but many companies still have not achieved this level of changeover. To this day, we have setups taking 60 minutes and longer. Why?

Are these long setups not viewed as a problem?

Do we just accept the status quo of long setups?

Do we even track and monitor setups?

Is it easier the just buy faster (more expensive) machines than to roll up our sleeves and figure out how to reduce our changeovers?

Is it not a priority? Are we too busy with out limited resources (yet we let go resources in the last layoff)? Short term thinking wins again?

Where’s the SMED? Can any company report that all of their setups are 10 minutes or less?


Mark R. Hamel said...

Many times I run into folks who explain that they have applied SMED and have indeed achieved the "single minute" internal cycle time.

The problem is that they can rarely demonstrate their achievement. Direct observation will often surface internal set-ups that exceed the single minute mark. With that, they can rarely point to any set-up standard work, especially one that properly reflects steps, sequence, cycle times and anything that reflects which steps are to be done externally or internally. In other words, no standard work sheets or standard work combination direct evidence of how they supposedly and routinely execute SMED.

Yet another critical item that is usually absent is any sort of plan vs. actual reporting for internal set-ups. With plan vs. actual charts comes accountability, performance history and the identification of any barriers/issues/opportunities.

So, while many folks have yet to apply the rigor of SMED, still many have failed to sustain their SMED.

John henry said...

More important, what is the cost of SMED? Or rather, the cost of not doing SMED.

I have been asking this question of all my clients as well as audiences when I speak at conferences. I find that the hourly cost of downtime is known in about 10-20% of companies.

I tell my clients that it is at least $10,000 per hour but I have never been in a company where it was actually that low.

If it is $10,000/hr, even a savings of 30 minutes a week will save $250,000/yr.

It must be calculated and the calculation must come from accounting or finance. Numbers from anywhere else will not be believed by management. They will not be "official"

I have an Excel template that calculates the dollar costs of changeover. If anyone wants a copy, send me an e-mail at

John Henry