I don't know. Maybe not. I do not see the connection between time reduction and number of changeovers.

And if the analysis of internal and external activities during consecutive changeovers find that changeovers differ only a small number of activities and by lengthen (slight) time of one changeover we can eliminate another changeover? Do you think it is possible?

We start research in company in October. You can join us if you want :)

Hello Justyna,

SMED helps you in many cases: you can reduce EPEI (Every Part Every Interval) in order to improve flexibility or improve “Takt” by better lot sizes of product families. It helps to standardize workflow, new dedication of product and machine and very often enables to reduce organizational waste, like priorisation, as well.

Your benefit depends on your business case: reduction of change overs sounds that you can optimize lot sizes and use "flight plan" scheduling of work orders, a kind of "OXOX" or Heijunka board. This increases your productivity already but don’t stop bringing down lot size and reduce EPEI, this will give you even more flexibility and utilization.

SMED is a waste reduction activity and usually opens a big range of opportunities to improve value added portion per process by waste reduction.

These are only a few quick thoughts;

hope they help

Kind regards Gregor

Dear Gregor,

Thank you very much for your comment.

Production schedule we accept as constant and will not interfere in it. I have in my team leveling production expert and production schedule developed by her we will take as input to this project.

We would like to analyze the sequence of activities performed during changeovers, their repeatability, and analyze possibility on limiting the number of changeovers.

Once again, thank you very much for your inspiring comment!

I agree with Beck Gregor Schulte, the aim of SMED is to reduce the time of model change through the elimination of waste and reduce number of tasks internal, if the goal is to reduce the number of model changes then increases the interval, but that involves increasing inventory and cost, in my experience I have not observed any relationship

Thank you Martin for your suggestion. We will check impact of reducing number of changeovers on inventory, cost but also on other production indicators. Maybe reduce the number of changeovers will have a negative impact... I assume that in certain conditions of production it can bring a positive effect. But of course I might be wrong.

I worked for Toyota before. We were taught by the Japanese managers that primary intention of SMED is to reduce WIP inventory of each model if you are running multiple or  mixed-model production in support of Heijunka. Less WIP means shorter Lead time while maintaining your Throughput. The line becomes very flexible to demand. In effect, it is to have more frequent change-overs. So this is in contrast to your objective of lessening frequency of changeovers. However, if you are running  few models in your schedule, I think you have a point of having less changeovers.  

I agree with Erick, today because of the diversity of products is not possible to have dedicated lines, this means that the lines are more flexible and therefore more changes over to meet the demand of customers in the shortest time possible.

So at the end it depend how many different parts you are running in line.

Thank you Eric and Martin for the answers.

I also agree with you that the objective of SMED is to increase the flexibility of production and that the number and time of changeovers depend on production plan.

Do you know maybe (referring to your experience) types of processes in which the changeover (some) consist only of external actions or only of internal actions?

In my experience all SMED processes include internal and external tasks

Thank you Martin.

I agree that thinking in Lean the objective of SMED is to increase the number of changes in order to reduce batch sizes and, in consecuence, WIP and lead time.

In any case a rule is required in order to define the number of changes because every change is reducing the capacity of the machine (reducing its OEE)

In my practical experience, I propose SMED as an investment in flexibility. The cost of this investment is the % of OEE you "waste" in changeover. So this % has to be defined.

Frequently 10% of OEE is used as this value (I am still looking for the origin of this "magic number" I apreciate if somebody has information about).

In any case this number must be defined by management.

After that you can defined the number of changes per period of time and, everytime changeover is reduced, number of changes can be increase always maintaining the % of OEE "invest" in flexibility.

Dear Francisco

Thank you for very interesting suggest.