I have a question pertaining to a nonconformance.

Situation:

Parts were manufactured as required by all applicable procedures, work instructions, drawings, etc. at the start of the work order. The operator realizes they are short approx. 800 pcs. to complete the work order. The operator notifies supervisor, who notifies the material handler, who inturn, has to have material converted to a specific width to run through machine.
Not a problem.

In the mean time the operator tears down machine and sets up to run another job while the material is being converted, again not a problem.

After finishing the other job the machine is torn down and reset-up to run the remaining 800 pcs. The operator runs the 800 pcs. and brings the final piece to the quality dept. to be inspected.

The quality dept. inspects the final piece and realizes that one of the dimensions is oversize, remember at the start of the order everything was correct.

After further investigation it was found that one of the dimensions was apparently transposed, resulting in 800 pcs being scrapped.

My question is would this be considered a process error or human error?

I ask because in my opinion the process wasn't in error the person failing to follow the process was!?

***Note***
The operator is aware of the requirement to have a first piece checked any time a machine is set-up or in this situation resetup.

Is there a checklist for the operator to follow/initial?

How experienced/trained is this particular operator?

Is there a long delay in getting the first parts checked?

Is there a checklist for the operator to follow/initial?
There is a control plan that outlines what is to be accomplished.

How experienced/trained is this particular operator?
The operator has been doing this job for 11 yrs.

Is there a long delay in getting the first parts checked?
No, there is not a significant delay in getting the first part checked.

IMHO - this was a process error.

Ask yourself - what could be improved in the process to prevent numbers from being transposed?
Could the set-up instructions be improved to prevent this?
Could the print be clarified to prevent this?
Could a differnt measurement technique be used to prevent this?
Could the set-up person benefit from some additional training to prevent this?

This, according to me sounds like a system failure -

by the way you have described, it looks to me as if this LONE operator is responsible for everything ?

Does your system not have Verification or other team members from Quality take part in "In process QC" checking at frequent intervals of the process ? Also there should a system of checking the first shot and if this is ok, the rest of manufacturing to continue.

Would it be not a good practise to ensure that machine parameters once set are checked / verified by supervisor for confirmation ?

With all these points, I would say that the process is inadequate to detect failures - remember that Quality has to be built in the process and not tested for Quality. :)

As with the others I would suggest this is entirely process error, particularly in the machine set-up there should be verification prior to running an entire lot. As suggested a first peice would work fine or some kind of poka yoke.

I have a question pertaining to a nonconformance.

Situation:

Parts were manufactured as required by all applicable procedures, work instructions, drawings, etc. at the start of the work order. The operator realizes they are short approx. 800 pcs. to complete the work order. The operator notifies supervisor, who notifies the material handler, who inturn, has to have material converted to a specific width to run through machine.
Not a problem.

In the mean time the operator tears down machine and sets up to run another job while the material is being converted, again not a problem.

After finishing the other job the machine is torn down and reset-up to run the remaining 800 pcs. The operator runs the 800 pcs. and brings the final piece to the quality dept. to be inspected.

The quality dept. inspects the final piece and realizes that one of the dimensions is oversize, remember at the start of the order everything was correct.

After further investigation it was found that one of the dimensions was apparently transposed, resulting in 800 pcs being scrapped.

My question is would this be considered a process error or human error?

I ask because in my opinion the process wasn't in error the person failing to follow the process was!?

***Note***
The operator is aware of the requirement to have a first piece checked any time a machine is set-up or in this situation resetup.

It's impossible to say for sure from this vantage point where the cause lies. Carol raises some good questions, but given an optimized process it still might be possible for mental errors to happen, and then be compounded by failure to observe process requirements. Unless there's some method of foolproof error prevention, the best we can do sometimes is minimize and not eliminate the possibility of nonconformance. We should think in terms of process optimization (making it as good as it can be in view of known restraints) instead of process perfection.

....The operator realizes they are short approx. 800 pcs. to complete the work order. The operator notifies supervisor, who notifies the material handler, who inturn, has to have material converted to a specific width to run through machine.
Not a problem.



IMHO, this is where the process may have some deficiencies. Not sure how thorough the verification is at this point, but this sounds like some opportunities for error. Is the notification written? When the conversion is made, is there a verification of proper machine setup? If the proper changeover process was followed, then the likelihood the equipment would have been set up correctly would be quite high.

As with the others I would suggest this is entirely process error, particularly in the machine set-up there should be verification prior to running an entire lot. As suggested a first peice would work fine or some kind of poka yoke.

The OP says that first-piece verification is already required. In most cases, humans operate processes, so the best-laid plans can still go astray. I have an issue with the idea of second-party verification to begin with; we talk a lot about everyone being responsible for product quality but no one seems to really believe it. Why is production the only place in the plant where inspectors are considered necessary? Why isn't there an inspector in Accounts Receivable?

Having said that, I was once in a plant where there was a QC first-piece verification requirement, and they had an 8.5x11 piece of green construction paper hanging in a vinyl protector in plain sight next to the machines--placed there by the inspector--after the first piece was approved. It was a simple way for anyone to tell whether the required procedure had been followed, and required an overt act of dishonesty on the part of the operator in order for the procedure to be subverted.

Such good responses so far.

I understand the operator "was aware" of the need for a first piece, but how is it that the process can run through without evidence of the first piece successful inspection? Therein perhaps lies the process weak point.

How is such a thing managed? If initials or results must appear on a job order at a certain step the lack of such information could help prevent the first piece being missed--no doubt he would have made sure to have that first piece done in every other single case...

My point is, a control plan like a router can also serve as a record as well as include "gates" for subsequent steps for this kind of mistake proofing. 11 years of experience doesn't help with such things (all that experience could, in fact work against a person through a sense of routine familiarity) - controls should be designed into the process.

IMHO, this is where the process may have some deficiencies. Not sure how thorough the verification is at this point, but this sounds like some opportunities for error. Is the notification written? When the conversion is made, is there a verification of proper machine setup? If the proper changeover process was followed, then the likelihood the equipment would have been set up correctly would be quite high.

Allow me to eloborate..

While all processes in the past may have been sufficient, it wasn't this time. Either they first article piece suffered a type 2 error (or whatever inherent inspection error is present), or there were not sufficient verification steps to assure minimal variance in the run after first article inspection. If I am reading this correctly, the only inspections performed are the first and last. I would for 800 pieces a few more pieces should be inspected (especially just given a changeover).

Hmm my apologies, I don't know how I missed that note at the bottom of the ops post. Anyhow, I do agree that a process should strive to achieve mistake proofing only to an extent that is reasonable. Still, my perspective stands, if my process requires first piece any time the machine is changed/set-up, I would want some way to ensure that requirement is followed every time. The truth is you're lucky if you can trust your employees to never stray from procedure, and you're naive if you believe they won't make mistakes. With that in mind, any process should strive to minimize the possibility of either of those happening.

The truth is you're lucky if you can trust your employees to never stray from procedure, and you're naive if you believe they won't make mistakes. With that in mind, any process should strive to minimize the possibility of either of those happening.

Well put, but I think that having trustworthy employees is an act of commission rather than just blind luck.

Lets add some things:

1. A first piece is required after set-up of machine prior to manufacture run to verify set-up.
2. In-process inspection is required and accomplished by the operator at a predetermined frequency. (i.e. 1 every some many pieces, 1 every hour/minutes, etc.)
3. Final piece is required at the end of the run, prior to packaging.

Also, the opinion and situation I have stated previouly was an opinion of someone in upper management! They are always quick to blame operator everytime a situation like this occurs and I am trying to get them to see beyond the operator and stop the finger pointing. :frust:

My suggestions were similar to alot of the ones mentioned here.
My first thought was why was the operator short material to begin with?
Just on that phrase I could think of many things that could of caused the order to be short material. Just for starters, The instrument used to count the amount of material required could be faulty/inaccurate.

Process error. The error was before any of the product was run. The planning was in adequate. There should have been adequate material, method, machine, measurement, manpower, and environment before any of the production started. This would have prevented the risk in machine change over in the middle of the production run.

This may be incorrect or too simplified. How would we really know without being there? But if you want a snap judgement, I would say it is a process error.

I would caution from putting too much trust in inspections. Inspections are often human based. Inspection does not affect the conformance of product. Inspection is an effort to minimize escapes on non-confomances. If the process can produce a non-confoming output (red bead), then the employee will always be at risk of being blamed for the creation of a non-conforming output when really the process created the non-conforming output. The employee did not identify the non-comforming output of the process. They did not create the non-conforming output. They just did not stop it.

Inadequate planning and process design of the input process caused the error downstream. The employee can only be as good as the processes provided,. In this case, the process provided included an unnecessary risk due to an unnecessary machine change over.

Check out the Red Bead Experiment for more thoughts.

Regards,

Dirk

Also, the opinion and situation I have stated previouly was an opinion of someone in upper management! They are always quick to blame operator everytime a situation like this occurs and I am trying to get them to see beyond the operator and stop the finger pointing. :frust:


The most "successful" and persistent managers are those who best mimic the management strategies of the top managers. If that strategy consists primarily in blame-seeking, then lower managers who don't follow that path generally won't be around for long. If operators are afraid of being hit over the head, they'll devise some sort of protective headgear whether or not the written procedures allow for it. Expect the sort of thing you describe to continue to happen until there's a fundamental change in management style. Whether or not it's possible to make that happen--or facilitate it somehow--depends on how far gone the top manager is. Sometimes they won't admit that they're drowning until they're going down for the last time.

Process error. The error was before any of the product was run. The planning was in adequate. There should have been adequate material, method, machine, measurement, manpower, and environment before any of the production started. This would have prevented the risk in machine change over in the middle of the production run.

Inanimate objects are incapable of error. Process design might be at fault, but that's not process error--it's another form of human error. Which is to say that it might well be that the operators of the process were not at fault, but the designers of it might have been.

Then the answer is human error, somewhere in the system.

This also would allows us to never even ask the question, is it a process error?

Of course it is some form of human error. It is a human created business system.

If there is no malitious intent - voluntary and concious mis-process, then I would look to the design and implementaiton of the processes. I would look at the MMMMME that I listed in my previous posting. I did not list the MMMMME to mean that the material or machine casued the problem.

I would look at how the processes were designed and implemented by humans. And I would not look at specific individuals. My opinion.

Regards,

Dirk

Then the answer is human error, somewhere in the system.

This also would allows us to never even ask the question, is it a process error?

Of course it is some form of human error. It is a human created business system.

If there is no malitious intent - voluntary and concious mis-process, then I would look to the design and implementaiton of the processes. I would look at the MMMMME that I listed in my previous posting. I did not list the MMMMME to mean that the material or machine casued the problem.

I would look at how the processes were designed and implemented by humans. And I would not look at specific individuals. My opinion.

Regards,

Dirk

The reason for trying to determine whether a given problem is due to the process or the operator is to avoid blaming operators for things they have no control over. This is a fundamental misunderstanding in prevention activities. Once you've determined that the operator isn't at fault, you don't stop there. You have to fix the process, but you also need to try to understand why the process was poorly designed in the first place. Nonconforming conditions that can't be traced to human error (either errors of commission or omission) are relatively rare.

The reason for trying to determine whether a given problem is due to the process or the operator is to avoid blaming operators for things they have no control over. This is a fundamental misunderstanding in prevention activities. Once you've determined that the operator isn't at fault, you don't stop there. You have to fix the process, but you also need to try to understand why the process was poorly designed in the first place. Nonconforming conditions that can't be traced to human error (either errors of commission or omission) are relatively rare. Very true. I wonder if the process was made as per a control plan based on a FMEA? I won't hold my breath.

My point is, a process should be designed with the intention of preventing errors as a key focus. Instead, I find processes are usually designed to make stuff.

Sometimes there is a lot of weight placed on X years of experience. Fault with that goes both ways:

1. If highly skilled people feel they're being dumbed down by a record that serves as a sort of checksheet.

2. If people in positions of authority will say "So-and-so has X years of experience and shouldn't have this hand holding" or some similar copout.

Both are attitudes. Both need to understand that the steps in place are just a process. Jim is right; management may never understand because whipping some operator is so much easier than taking a system view of things.

Good luck with this.