Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations cowski on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

SPC "war stories" 3

Status
Not open for further replies.

PSE

Industrial
Apr 11, 2002
1,017
I have begun developing a proposal for management to introduce a SPC pilot program on one of our product lines. I would like to be able to use some genuine case histories both good (to show how it can improve processes) and bad (to identify pitfalls). The Harvard Business Review has only a few case histories available. Would anyone care to divulge some of their experiences?

Regards,

PSE
 
Replies continue below

Recommended for you

For a pitfall....

I've heard several stories of operators refusing to record out-of-tolerance data. They were afraid that if they recorded defective parts, they would be fired. (I.e. "You built this part wrong, so leave.") Of course, most problems with machines are not caused by ill-spirited employees, and SPC won't work without recording defects....
 
I agree. Find some way to cover the operator’s butts before you start. I was called in as a brazing consultant for a company totally devoted to SPC. They had a tool breakage problem. My first question was “Where are the numbers for which tools are breaking where?”

Think of a carbide tipped end mill. Breakage was either the carbide or the carbide and part of the steel behind it. The brazers identified the grinders. We found the left hand tools broke more than the right hand tools because the grinders were set up for right hand tools. When they ground a right hand tool the coolant was collected. When they ground a left hand tool the coolant hit the operator in the chest. This solved 60% of the breakage problem.

Pretty much the causes of failure will be attributed to someone else if possible. Machinists hit fixturing and broke tools but you very rarely saw that as the reason given. The brazers would take a tool with a damaged carbide tip and steel holder. They would braze on a new carbide tip and then fill the gap behind it with braze alloy. Braze alloy does not have the same properties as steel so the tips broke readily.

The company would not let the brazers see the original drawing for the tools so the brazers would know what was in tolerance and what wasn’t.

1. Try to come up with a way to cover the operators’ rears. I like starting out by saying that it looks like everyone is doing a really good job but we have a communication problem.
2. Many embarrassing things will come up at all levels.
3. Expect that some people will quit rather than share information.
 
True Story:

A Quality Manager is presented with out-of-tolerance parts (diameter) to evaluate/disposition.

Immediately in front of him are:

1) Samples of the out-of-tolernace parts.

2) A micrometer.

3) The lot paperwork, which includes SPC charts.


Quick, what does he pick up?

Does he pick up a part and measure it with the micrometers?

No.

He picks up SPC charts and announces that the parts couldn't possibly be out-of-tolerance because that's what the SPC charts tell him. He refuses to even look at the parts.

When the "charts become more important than the parts" you've lost sight of the goal.
 
wow common sense at last from Kenny, and from someone in aerospace , a very rare trend (note my irony).
While i agree that spc and other monitoring tools are useful, thats all they are a tool.

It seems to me that more and more is being place onto having meetings about a meeting your going to have

Alot of time SPc is used for the sole purpose of impressing the customer on thier audits. Who can honestly say that if the process if going okay that they spend time looking over the SPc charts, or dutifully fill them in every day, unless that is your sole responsibility.

SPC only works if the person driving it has enough personality and responsibility to alter processes and drive forward thier findings and outcomes. How many times have you went to the shop floor and asked "why are you doing it like that? and the response was " we have always done it like that."

Anyway we dont need SPC we dont make any defective parts (Irony again)

 
Some Very good postings so far (at least from the downside point of view). Keep 'em coming. By the way, has anyone out there had a POSITIVE experience?

This is what I am hoping to be able to implement:

Automated Data Collection: This should preclude operator intervention on the data set. Currently the product line is about 75% there in saving test data to a common database.

Data Mining Tool: Non SPC data retrieval for use in assembly configurations (product traceability etc.)

SPC feedback to the operator: Provide feedback to them (preferrably in as close to real time as possible) on the stability of the process being performed.

Operator training: They need to know how to react to what they see.

I wholly agree that SPC is simply a tool and not an ends in and of itself. The data gathered and charted is pretty much useless unless it is used to react to what is going on. My biggest concern at this point is that it will simply be a monitoring tool with data being taken for data's sake.

Regards,

PSE
 

Your good operators are already using some form of SPC. Good operators in good operations know what kind of results they should be getting and monitor their performance.

Many, many years ago I had a job where I went into operations and turned them around. After three times my reputation was to the point where people would quit when they heard I was coming. However maybe 40% of the good people were really happy someone was coming to make everyone do their job properly. Good SPC with the openness and feedback that requires will make alto of good employees very happy. It does work.

We have a brazing operation with six belt furnaces. We established a process so there is a very clear, simple visual indicator of wetting. The operators constantly monitor the results and adjust for drafts, power fluctuation, discrepancy in part size or density, etc.

Start by assuming the operators are really good and start by checking how well the equipment is working. When I go in I assume the operators are all good but we probably ought to go over the procedures anyway. This gives people a chance to quit doing things wrong and / or to start doing things right without every having to admit they were making a mistake in the first place.

I like to start really early by making a change that improves quality but also improves the operators quality of life. Lighting usually works really well for this. Maybe next best are simple repairs to keep the workplace cleaner. If you can smell, taste, or feel oil (welding fumes, grinding dust, etc.) in the air start by checking the ventilation. Try the book Industrial Ventilation from the American Council of Government Industrial Hygienists. Warning – an open door can really screw up the ventilation but some people really like an open door in nice weather.

I once had a Boeing supervisor comment on my “Columbo’ like approach. Go in, hang around and ask a lot of questions until they figure out the answers on their own. The only problem with this is that you don’t look smart because they figured out the answer on their own and you just happened to be there.

Good news – Deming can work better than you ever dreamed. Kaizen can be both fun and very productive.

I like to take what we call a modified Toguchi approach. You go in, assess the situation, consider all the factors and then see how much good and how many kinds of good you can accomplish with the elast amount of effort anfd expsanse. The better you do this the more elegant the solution.

Ask (yes, really, politely, honestly ask) the people at all levels what they think. We have an engineer who doesn’t like doing this because eh is an engineer and he knows best. After enough years they have pointed out ways to make enough things better that he now sees value in the system.

If you have honestly worked on the plant floor it will help. Do not go down and run a 5 minute test then extrapolate that it to an hour, day, etc.

Read Goldratt then reread Goldratt.

Put yourself in an operators place. In my earlier example, no one in management realized the operators were getting hit in the chest and they had no idea how unpleasant that was.

Tom



 
the problem is aminly due to the mentality of the organization.
If the organization provides necessary safeguard to the workforce in recognising their efforts rather than results.
I think book by alberto Galgano on Company-wide Quality management, throws more light on this topic.
thank you
gmkme
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor