I am interested to know how many companies use both NCMRs and CARs and how they distinguish which needs to be used when. The system I'm working with uses NCMRs for any material processing error that is caught internally. CARs are used for any nonconforming material caught by our customer, procedural issues, customer complaints not product related and improvement opportunities.

I think I would like to eliminate the NCMR and use only a CAR for all items requiring corrective/preventive action.

Your input is greatly appreciated.

Mike

Here's MHO. Both or a modification of a CAR is needed. There are numerous issues that can fall under a "1 Off" situation where all that is needed is a quick CA. The event is isolated and contained and the cause is due to process limitations. It is just not feasible to put in permanent solutions. Track it on the NCMR, of course you can have the same info on a CAR/PAR but on some of the 1 Offs you just need a notification. The information compiled can be used to show feasibility of a permanent correction.
To me it is similar to the determination of foreign material and mixed parts.

I am interested to know how many companies use both NCMRs and CARs and how they distinguish which needs to be used when. The system I'm working with uses NCMRs for any material processing error that is caught internally. CARs are used for any nonconforming material caught by our customer, procedural issues, customer complaints not product related and improvement opportunities.

I think I would like to eliminate the NCMR and use only a CAR for all items requiring corrective/preventive action.

Your input is greatly appreciated.

Mike
In general, the two types of documents serve different purposes, if by "NCMR" you mean NonConforming Material Report. The NCMR is used to document the problem so that whomever is responsible for disposition of the material knows what's wrong with it, and of course to provide a record of the problem. The dispositioning authority then has options: Use as is, rework/repair, scrap, etc., and also might deem corrective action necessary, which is when the CAR happens.

JSW, that is what we are currently doing. The challenge we are facing is with our root cause so we can try to fix whatever allowed the problem to originally happen. Often, we are only taking containment action to correct the errors and ignoring the real issues.

The way our system was designed is to generate a CAR when an NCMR requires further action to correct a systemic flaw. In the spirit of lean, I would like to avoid duplication of our documentation.

JSW, that is what we are currently doing. The challenge we are facing is with our root cause so we can try to fix whatever allowed the problem to originally happen. Often, we are only taking containment action to correct the errors and ignoring the real issues.

The way our system was designed is to generate a CAR when an NCMR requires further action to correct a systemic flaw. In the spirit of lean, I would like to avoid duplication of our documentation.
I agree with JSW05, these two documents are for 2 separate things. The NCMR is to document and identify nonconforming material and the CAR is to initiate and documetn root cause corrective action. These actions are not the same. unless you are an expremely high quality shop with very few incidents of nonconforming material, then you wont' want to - or be physically able to - get to root cause of every nonconformance.

In the spirit of lean, perhaps you can identify and remove redundant information...I woudl think the CAR is a rather large form with many fields designed to take a lot of text input while teh NCMR is a simpel form to ID the NC and disposition (use as is, repair,rework or scrap)

JSW, that is what we are currently doing. The challenge we are facing is with our root cause so we can try to fix whatever allowed the problem to originally happen. Often, we are only taking containment action to correct the errors and ignoring the real issues.

The way our system was designed is to generate a CAR when an NCMR requires further action to correct a systemic flaw. In the spirit of lean, I would like to avoid duplication of our documentation.
The nature of the document isn't important--you can use some sort of hybrid and still capture the same information. It's the two processes that need to be kept separate, and there shouldn't be anything in your system that requires CA every time nonconforming material is discovered.

I think I would like to eliminate the NCMR and use only a CAR for all items requiring corrective/preventive action. Mike

That is the way I like to handle failures, be they customer complaints, internal failures, process failures, supplier or vender failures, or material failures.

Normally, all failures are handled the same way... All identify who wrote the NCR, the recipient, the problem, reason that it happened, immediate corrective action, preventive measures, and evaluation of the effectiveness of the corrective and preventive actions.

I also think one form simplifies and standardizes reporting.

I have seen that some companies want to quantify the cost of failures, material, labor, etc associated with corrective actions. In that case, it might be sensible to have a separate form. But one has to be careful not to have too many failure reporting venues or the users will get lost.

unless you are an extremely high quality shop with very few incidents of nonconforming materialLOL!

Allow me to add more information so I can paint a better picture of what’s occurring within the building. Over the past 9 months we’ve moved to a building 3 times larger than our first 30 years of business. Our sales have increase by 70% and our employee roster went from 160 to 225. That’s the good.

The bad, our errors have increased from 17 in FY04 (with one customer complaint) to 22 in FY05 (with 2 customer complaints). The first 4 months of FY06 have yielded 14 processing errors (with 6 customer complaints).

Our way of business in the past was quite simple, nothing changed. The same employees worked for years. Average tenure here was better than 10 years. The jobs remained the same. The customer base remained the system. The system we had in place worked and worked well for what we did. The culture we are in now requires the system to change in several areas, some drastically.

I see two routes to accomplish this. This first is to start fresh with a new management system. That isn’t the answer, in my opinion. I feel we need to take the current system and help it grow to match the growth we did as a company. Everytime nonconforming product is produced (not normal scrap) that tells me something within the system went wrong. If it was just human error, fine. I can live with, accept and acknowldge I am part of that problem. Everything else is the result of the system not functioning as well as possible. That creates an opportunity or need for improvement. Hopefully, as we fix the holes our processes will become what they need to be so we can be successful.

If we only take the NCMR’s (used to document the processing error and disposition) and not funnel them into our CAR/PAR program, wouldn’t I be missing opportunities to fix the holes when they are identified?

well, not sure what you make or what your volumes are but if there were only 22 incidents this year then that isconcievably a managable volume of CARs - assuming that you can support that many at your staffing level...My current and past experiences are with hundreds and thousands of nonconforming material incidents in a year. clearly unsupportable to generate a CAR for every nonconformance...

The nature of the document isn't important--you can use some sort of hybrid and still capture the same information. It's the two processes that need to be kept separate, and there shouldn't be anything in your system that requires CA every time nonconforming material is discovered.
I agree with JSW05 that the processes are separate, and use two forms for that reason. Corrective actions are initiated based on risk, cost, or customer request. I would not like to have an audit finding due to an incomplete combined form - not when the nonconformance has been resolved in a way that is acceptable to the customer.

That said, if every one of your NCMR's results in a corrective action, there is no real reason to have two forms - especially if you're just duplicating information like, "problem description", "Disposition/Action Taken", and "Cause".

There's a good example of a combined report/Corrective Action form in the book, "Correct! Prevent! Improve!" by Jeanne Ketola and Kathy Roberts (ASQ Press).

In my view, the concept of a report of a "suspected" nonconformance is different from a final determination of a nonconformance, at which point, "someone" makes a decision to

use as is
"correct" the nonconformance (supply missing documents, repair or rework the product or service)
scrap and replace
scrap and reimburse (unable or uneconomical to replace)
Depending on the individual circumstances, "someone" makes a decision whether to implement a root cause investigation and subsequent "corrective action."

In my own practice, "suspects" were referred to a Material Review Board. The MRB had power and authority to make unilateral decision on suspects detected in-house. If a suspect arose at a customer, MRB worked jointly with customer to determine true status and devise a remedy acceptable to all parties. Often, a customer would be called in to affirm a "use as is" determination, regardless of where the suspicion arose.

Accordingly, regardless of the exact form title, the function of the initial report was "request for determination of suspected nonconforming material" at which point material would be segregated (quarantined) until final determination of status and a plan of action (up to and including root cause and corrective action activity.)

I get suspicious when I see folks worrying more about a specific title for a document than about the underlying process which will be triggered by the document. My suspicion is they have compartmentalized themselves so thoroughly they can't see the "big picture."

In my view, the concept of a report of a "suspected" nonconformance is different from a final determination of a nonconformance, at which point, "someone" makes a decision to

use as is
"correct" the nonconformance (supply missing documents, repair or rework the product or service)
scrap and replace
scrap and reimburse (unable or uneconomical to replace)
Depending on the individual circumstances, "someone" makes a decision whether to implement a root cause investigation and subsequent "corrective action."

In my own practice, "suspects" were referred to a Material Review Board. The MRB had power and authority to make unilateral decision on suspects detected in-house. If a suspect arose at a customer, MRB worked jointly with customer to determine true status and devise a remedy acceptable to all parties. Often, a customer would be called in to affirm a "use as is" determination, regardless of where the suspicion arose.

Accordingly, regardless of the exact form title, the function of the initial report was "request for determination of suspected nonconforming material" at which point material would be segregated (quarantined) until final determination of status and a plan of action (up to and including root cause and corrective action activity.)

I get suspicious when I see folks worrying more about a specific title for a document than about the underlying process which will be triggered by the document. My suspicion is they have compartmentalized themselves so thoroughly they can't see the "big picture."
Yes. All "nonconforming" material should be considered suspect until disposition takes place, because the purpose of the disposition is that determination, and what to do about it. I also concur in the idea of a cross-functional MRB being responsible for dispositions and assignment and tracking of followup activities such as CA.

If I am reading this correctly, Wes and JSW feel that a review body should/could determine if a CAR is necessary based on the information provided to them in the initial report. That adds a step to our process but isnt necesarily bad.

I wonder if part of my confusion is when NCMRs are generated. Within my industry, bad product is expected. Our labor force is primarily persons with disabilities. While we error proof as much as possible when establishing work instructions, we also build in insepction into the equation to confirm quality. For example, I may have an employee peforming a light assembly task. The work is inspected at a predefined level (often more than 50%). The product found to be nonconforming is returned to the individual to correct. This could be 1% or 50%. Either way, I do not document that as a nonconformance.

What we document are items not caught within the normal operations of the job. If, during the same light assembly example above, the error wasnt caught until the product moved onto the next operation, then I would document the nonconformance. It is documented then because it is outside the normal and expected practice of the job/employee performing the job. At this time we most likely have a system breakdown. Faulty product, lack of inspections, lack of resources, lack of instructions, lack of training.....

It's because of this that I am thinking we would be better off with root cause for all processing errors.

How off base am I?

I concur with both Wes and JSW05 but would like to add some past experiences. In the past (defense electronics) a certified electronic test operator, upon finding a test failure of say a PCB, would immediatedly perform diagnostics to determine root cause of failure. The test operator would then generate the NCR and perform a preliminary disposition which essentially was a rework to drawing. For example, suppose the diagnostics revealed that a resistor was wrong value the test operator's preliminary disposition would be remove, replace and retest. If the rework was successful then the NCR was dispositioned by the QA supervisor to accept the PCB and forward the removed part to MRB.

MRB would investigate and track for trends. MRB would also close out the NCR by issuing sheet two which scrapped the removed resistor.

Now let me add that we also performed in process inspections where defects (if found) were documented on a "Yellow Tag". These were returned to the operator for rework to drawing. Upon successful rework the "yellow tag was stamped off as accepted and sent to the QAE assigned to the project. It was the QAE function to track all yellow tags for trends.
All of the above, plus what Wes and JSW05 have posted, were performed under MIL-Q-9858A.

How off base am I?
You're right on. You can use one form and proceed to corrective action whenever nonconforming product is discovered. There's no requirement that you have a review board.

In general, the question comes down to one of "empowerment" which may be tempered by the complexity of a product and subtle nuances which may make a product usable or unusable.

In a machine shop environment for instance, an operator is fully justified in scrapping a "short" (a bar end which is shorter than the product specification.)

The nonconformance and its cause are readily apparent and may even be a planned part of the process because it provided a gripping surface for the machine to work on the next to last piece of the bar.

In the case of a single sample of a 10,000 piece lot being "suspect," much more goes into the determination of whether to scrap or save the entire lot and whether there is economic justification if the pieces can be sorted, root cause performed, etc. As in all things having to do with production, much rests on the downstream consequences of each action and who the managers want to give responsibility and authority to.

We each need a sense of "proportion" in assessing our response to a specific case. Certainly we cannot blindly apply a response fit for a one-off operation where the operator is extremely expert to a situation where there may be a question about the sufficiency of the "suspect" determination. Certainly, we have all heard of instances where customers rejected product, only to find they had an instrument out of calibration or the wrong version of the specification.

Certainly, my own MRB often empowered individuals to make decisions and determinations without bringing the decision before the full board. Sometimes, even those individuals chose to still bring it before the board because the decision was "sorting pepper from flyspecks."

If I am reading this correctly, Wes and JSW feel that a review body should/could determine if a CAR is necessary based on the information provided to them in the initial report. That adds a step to our process but isnt necesarily bad.

I wonder if part of my confusion is when NCMRs are generated. Within my industry, bad product is expected. Our labor force is primarily persons with disabilities. While we error proof as much as possible when establishing work instructions, we also build in insepction into the equation to confirm quality. For example, I may have an employee peforming a light assembly task. The work is inspected at a predefined level (often more than 50%). The product found to be nonconforming is returned to the individual to correct. This could be 1% or 50%. Either way, I do not document that as a nonconformance.

What we document are items not caught within the normal operations of the job. If, during the same light assembly example above, the error wasnt caught until the product moved onto the next operation, then I would document the nonconformance. It is documented then because it is outside the normal and expected practice of the job/employee performing the job. At this time we most likely have a system breakdown. Faulty product, lack of inspections, lack of resources, lack of instructions, lack of training.....

It's because of this that I am thinking we would be better off with root cause for all processing errors.

How off base am I?
Not off base at all, given the special circumstances. In general, it's assumed with questions like your original post that some sort of ISO system is involved, and there aren't any extenuating circumstances such as disabled employees for whom expectations are perhaps necessarily different. You should work the system in a way that makes the best sense for your circumstances.

For example, I may have an employee peforming a light assembly task. The work is inspected at a predefined level (often more than 50%). The product found to be nonconforming is returned to the individual to correct. This could be 1% or 50%. Either way, I do not document that as a nonconformance.Seems like hidden cost.

JSW, your correct. We are ISO 9001:2000, ISO 14001:2004 and CARF (rehabilitation) registered as a non-profit. The workers we employ are all held to the same standards regardless of their ability. Those standards are established based on the product, not the person. We have three missions working at the same time, neither of which can function independantly. First, to provide quality services that will assist persons with disabilities in achieving their employement goals; Second, to provide quality quality products that meet our customers' needs; Plus an EMS mission.

If you anyone posting in this thread goes home thinking they didnt accomplish much today, you did. Thanks for all the input. I greatly appreciate it. I'm pretty confident I have solid idea for our MR meeting tomorrow. If not, I'll be back with more.

Mike

I am interested to know how many companies use both NCMRs and CARs and how they distinguish which needs to be used when. The system I'm working with uses NCMRs for any material processing error that is caught internally. CARs are used for any nonconforming material caught by our customer, procedural issues, customer complaints not product related and improvement opportunities.

I think I would like to eliminate the NCMR and use only a CAR for all items requiring corrective/preventive action.

Your input is greatly appreciated.

Mike


I am a 3rd party auditor. You can use your CAR process for NCMRs if you like. They serve a similar objective. However, most companies I work with (about 200) use both. The reason is that an NCMR tends to have many fileds related to product issues (quantity, where found, when found, no. of defects, disposition, etc.) which would tend to bog a CAR form down. BUt either approach could be compliant if done properly, and if the proper procedures describe the approach correctly.