We had tried our best to reduce the setting time (changeover time) from X hrs to Y hrs and had been successful for 6 months but it has started to deteriorate from the 7th month onwards.The setters who do the setting are skilled in their job and constant training is being given . One observation was that setting time have been less whenever the supervisor is beside the operator for the complete setting . But we are not able to ask the supervisor to followup the setting because he is held up in other machines ?
What can the forum users suggest to improve this performance? What do you find is lacking?

Andrews,

It is necessary to fully understand the inputs and outputs of a process and determine if the process is already in statistical control. For instance, what are Xbar and UCL/LCL for the changeover process for a given product/tool? Are only Common Causes present, or do you have some special causes at play?

It is difficult half a world away to know for sure, but I’ll hazard a guess that the naturally established changeover time was probably in statistical control suggesting that only Common Causes were in play. When you introduced a Special Cause (increased supervision), the outcome met your arbitrarily established lower changeover time. Once the Special Cause left, then the process returned to its former state and your gains were lost. The Supervisor is a Special Cause in this instance.

To make lasting change, you must establish a statistically controlled process (no Special Causes) and understand the Common Causes of variation. Next, establish a goal within the process limits (higher or lower depending on what it is you are trying to do). Then, you set to work to improve on the Common Causes. When these improve, the process becomes more robust and the process mean improves.

Regards,

Kevin

How was standard time Y decided? Did you try to establish the 'work elements' and the time required for these? Did you use techniques such as 'MOST'?

Hi Andrews,

I may be totally wrong here, but I'll assume that you mean setting up, like in tool change and things like that? If not, please disregard this. :bonk:

Anyway, If you do: Have you considered rebuilding the equipment for faster and easier setup instead of fine tuning the existing methods? A lot can be achieved that way.

/Claes

Took the exact words out of my mouth Kevin. I had typed a very similar response but had to quit before I posted it.

I wouldnt try however to reduce the natural variance of the existing process. I think you need to look for a step change to the process, i.e. re-design of the tools themselves probably, for example reducing the number of fixing bolts.

Andrew,

When the supervisor is present is he actually helping in some small way with the setup ??...maybe he is.

Break the setup down into very small elements including all movement, travelling etc;

You may find that something has moved or been relocated or just broken.

Have you got access to the original time study??

If yes compare notes..if not then do it again using an average rated worker (not your best one).

Regards,

Chris

M Greenaway said:

Took the exact words out of my mouth Kevin. I had typed a very similar response but had to quit before I posted it.

I wouldnt try however to reduce the natural variance of the existing process. I think you need to look for a step change to the process, i.e. re-design of the tools themselves probably, for example reducing the number of fixing bolts.

Exactly... That's what I was on about too.

/Claes

Setting time reduction:
Hi Kevin,
Thank you very much for your reply.Even though you are half a world away ,you have hit the nail in the head.The problem we face is that the operators tend to slow down the setting when the supervisor's head is turned.We are unable to change the operators attitude.
What do the forum users feel is the solution to the problem? How can one improve the attitude of operators?Are there any training programmes that may help improve their attitude?
Sometimes we find that the operators commit the same mistakes despite training them on numerous occasions.
For your information I would like to state that reducing the setting time to Y hrs can be easily achieved because we have been in the manufacture of these parts for atleast 20 yrs and the tooling has bee stabilised.

Hi Atul,
Thanks for your input but I was not able to understand what the "MOST" technique means? Can you please elaborate?

Hi Claes,
Rebuilding the equipment is a very good suggestion but it seems to cost a lot which we are not able to achieve.Moreover the reduction we are planning seems to require dedication from the operators.

May,
Work study is something we have not done till now. After your suggestion, we shall try to conduct work study as soon as possible.This might help us gain some knowledge on where we are losing time.

Andrews,

I was actualy asking how the time Y was determined?

MOST ® (Maynard Operation Sequence Technique) is a revolutionary work measurement tool designed to simplify and speed up the process of setting engineered time standards. It measures work by dividing it into basic motions such as reach, grasp, move, position and release, documents the motions, sequence of motions and time. The time for each motion is picked from tables of predetermined times.

In India, Mahindra & Mahindra have successfully implemented MOST.

However, I doubt if this would help in a case where 'operators tend to slow down the setting when the supervisor's head is turned'. :frust:

- Atul.

Do you already practice the technique of SMED? ( Single minute exchange of die).

Using this technique you isolate 'On line activities' from the 'off line' ones, leaving the machine down for a very little time.

Once the procedure is standardized, there is very little scope for ambiguity or variation of set up time, whether in presence of the supervisor, or behind his back.

The best case of SMED I have come across is an improvement of setup time of a press machine from 1 shift (8 hours) to 10 minutes!

Even though the name suggests 'exchange of die' it is used for all kinds of machine setups.

Great thread everyone!!

Marc started a forum some time ago called Total Productive Maintenance. Some of you following this thread may want to glean through the topics to see if there is something there of interest.

When I first read of TPM in Masaaki Imai’s book Kiazen, I thought to myself, “How slick!” In my experience, huge cost to the supplier is passed along to the customer because the supplier hasn’t figured out how to reduce the cycle time to effectively manage a kanban/JIT production system. As a result, the West struggles to break from the grips of mass production. In order to get piece costs to a desirable target, the producer is often left with the decision to over produce and shelve the product for later use. This is neither optimal for the Customer, who will bare some of this cost regardless, or the Organization. Huge gains can be made by endorsing TPM techniques including the outfitting/retrofitting of your existing equipment to facilitate quick changeovers and ease of preventive/predictive maintenance. But I digress.

From Andrews explanation, I get the impression that a good study has been performed and many of the variables are known. The next task is to work on these variables and to find ways to reduce the negative effects and improve the inputs. Much of the focus has been put on the employee, which may or may not be appropriate. Too often, manager’s expect improvement by expecting folks to work harder. This is especially true if a group that under close supervision established a high mark above the process average and management expects this to be the new average. After all, they did it once, they can do it all the time. They must be goofing off otherwise, right? Dr. Deming warned us about setting goals outside of the process limits (of a stable process). Many times the target can be reached, but at great cost to other components in the system. This is suboptimization. Other times, the goal cannot be reached, thus demoralizing the workforce. I would look into these two possibilities:

Is management expecting to much by arbitrarily establishing targets outside of the limits?
Are the workers demoralized because they wish to hit the mark established by management, but cannot?

My second guess is that sustained improvement has not been achieved because a realistic target has not been set and because management has not mitigated factors controllable by management. If several workers continually make the same mistake, management is at the root cause of the failure (poor training; wrong/inadequate resources, tools, materials, etc.). I would look at the problem, 5Why them, and trace the root cause to a management oversight. If that produced nothing, then we might look into subversive tactics by the employees.

Regards,

Kevin

P.S. Folks, if you guys only knew how many times I have typed a response to post, copied it from Word, returned to the Cove, hit Refresh, to find that one or more of you had said what I had... Great minds think alike. ;)

Any new thoughts on this?

Remember the oldie, but goodie, http://www.ChangeOver.com/ - It looks like it was 'sorta' abandoned July 2003, but it still has some stuff.

I looked at and thought about this thread. I did some contract work for Stolle (a division of American Trim of of Lima, OH) not so long ago. I thought of them and thought about the US and where the country is going. Press operators at Stolle will be replaced with migrants. American Trim will not pay a living wage for much longer.

I've worked with line folks and what I mostly saw was people whose benefits are reduced each year and other related 'management' factors.

One can increase training and other factors which should help. But I keep seeing the same thing: People don't care because they no longer believe their company cares about them. When this happens, training and such mean nothing.

At my organization we have Key Indicator called Utilization. This is our scheduled up-time and we have a designated goal that we are required to meet. This goal and its achievement is tied into our incentive program so our change-overs tend to get faster with each year as new skills are learned, new equipment is brought in, and the goals are tightened just that little bit extra.

Granted, Utilization takes into account not only change-overs, but unscheduled downtime. As we have learned that unscheduled downtime has frequently been attributed to something missed during a maintenance or production check, it is part of the equation.

Not saying it's a flawless system...the blackout impacted this number by no fault of our own, but our employees accept and understand this. (They're an awesome group of people!)

But how do we improve this number? How do we get better Utilization percentages?


Tool carts - each work cells has its own tool cart and there is a designated team leader responsible for ensuring it meets specifications (i.e., all parts are there that need to be there (from tools to bolts) and that it is in its proper location)
Shop organization - the "distribution centre" for our change-overs is well-organized and the layout is logical. Frequenly used parts/tools are closest to production. Shelves are labelled, multiple spares available for frequently used parts, etc.
Technical agreement - a "partnership" with a competitor overseas to exchange knowledge
Designated responsibilities - everyone has certain responsibilities to fulfill during a change-over, from the person who sweeps the floor in the area to the supervisor...this keeps people (a) busy and (b) from sticking their noses in jobs they shouldn't be
Incentive program - my least favourite option, but money can be an awesome force in getting people to do things
Training - explaining issues like new tools, new processes, new methodologies
Team building exercises - teach the group the benefits to working as a cohesive unit instead of individuals always watching their backs


Time studies and other SPC tools will help you, as well, when you wish to determine where you are and where you would like to be. But the above tools may be the keys to getting from Point A to Point B. :)