I am trying to educate our management on the meaning of "true" preventive action. I've been searching for some examples to help show them the way. Does anyone have any examples of "real" preventive action that they can share with me?

You are driving home and run out of gas, stranded, walk to gas station, buy gas can and gas. Corrective Action.

You are driving home and notice that gas gage is between E and 1/4. You stop to get gas before you run out. Preventive Action. (you reacted to a trend of the gage moving from Full to Empty)

Or how about a case where a three-hole molded part is cracking because one feature of the mold is too sharp? For corrective action you create a larger radius on that feature.

However, a very similar four-hole molded part hasn't shown any cracking yet, so you radius the equivalent feature on that as preventive action.

Alf

Some examples of preventive action: training, planning, FMEAs, knowing system and product capabilities and not promising more than either are capable of delivering, maintenance, monitoring processes for indications of failure, poka yoke, & 5s. Anything done to avoid or mitigate a potential problem.

Alf-
I beg to disagree, but what you have with the mold is a corrective action, because it (process) has already happened with another job or part. True it is probably more in the area of corrective action impact, but corrective action none the less. I see how you got there though, we used to think that way too, but with our last audit our eyes were opened and we were forced to look very seriously at our preventive action program.
Thanks for your input

Russ-
Thanks for the posting, but now you've got me wondering, especially since you infer that a registrar wouldn't consider my example preventive action.

Your point about applying it to processes makes sense, but I'm not convinced that applying PA to individual parts is inappropriate. Are you saying that, for example, establishing a guideline to ensure that future molds have proper radii (affecting the process of design) would qualify as PA, where simply fixing something that already exists, even though you're preventing the potential problem from showing up, would not? I can kind of see that. In fact, as I write this, I'm agreeing with your original statements more.

One other comment, with all the debate over CA vs PA, the registrars I've dealt with seem to be happy as long as you're doing something/anything.

Alf

Corrective Action = Reactive
Preventive Action = Proactive
This helps me to understand the difference.

I have part numbers 1,2,and 3 that each have a four cavity die to punch out tree sizes of rubber gaskets.

All dies were put into production on the same day.

On the third day of production we notice that die number 1 is starting to produce oversized gaskets but dies 2&3 are running within expected limits.

Die #1: We do corrective action and find that the punches we used need to be made of harder steel. We run the new "harder" punches and everything is dandy.

We look at each other and say, we don't want this to happen with dies 2&3 even they are running fine. Let's use the new punches that we got for die #1 in the other dies. Preventive Action.

It doesn't matter what the driver is, if it has not already happened it is a preventive action.

What hepls me is this:

Corrective action: Already happened.
Preventive action: Hasn't happened but leans toward product.
Continuous improvement: System/process related.

I think the finest line to walk is that between preventive action and continuous improvement, but that has been covered many times in this forum.

ASD...

I differ on the "car/gas" example given earlier. Filling gas in a car that has run out of gas would be "correction". Monitoring the gas level through the level gauge in the car to ensure that one doesn't run out of gas again would be "corrective action" (prevent recurrence of the same non-comformity". Installing and monitoring other "meters" to brake-oil, battery, tyre pressure etc to prevent potential problem of "outage" of the corresponding service would be "preventive action"

I think you are splitting hairs semantically on the car/gas example.

Some of what we use for Preventive action includes:
Process Improvement Teams - The product is not out of spec, but reveiws are performed to see how we can improve the process or product.
Analysis of Data - This is also 'preventive action'. The study of data in itself as well as review of the trends to make any changes BEFORE there is a problem.
Management Reviews - Also a 'preventive action tool' This is one of the best forums to communicate potential Management Systems changes (usually discussion here lead to start of an Improvement Team).

The list can go on and on. Semanitcally speaking you're splitting hairs ad nausium whenever this topic is discussed. There has been an evolution in what preventive means. With the same registrar, we've had several auditors and some do chew over the fine points while others lump them together.

Basically - If no 'problem' or 'deficiency' exists but changes are made (remember that even a review of the data is preventive) is it a preventive action. If you 'Correct' a problem on one part, and review other parts before their is a problem any steps you take for these parts is 'preventive'.

Hope this helps, not confuses more :rolleyes:
Best Regards, Eileen

Probably posted this in too many threads in the past but.... What made the best sense for the situation to my management folks.


Corrective action - A problem has found you - correct it.

Preventative action - you find something that "will be a problem" if a change isn't made before it happens.

IMHO a pre-production process FMEA can be considered a preventative action as well as pre-release design reviewes etc.

Regards
Jim

Jim, I've seen this elsewhere so please help me....What does IHMO stand for?

Thanks, Eileen

In My Humble Opinion

ASD...

Eileen,

Keep up the good posts! Consider going to the benchmarking section and give us some information about your company.

All new members, I need a larger sample group to put together the benchmarking matrix.

No more shameless exploitation of this forum from me. http://16949.com/ubb/biggrin.gif

ASD...

Gee Thanks AL!

Also see:
http://Elsmar.com/ubb/Forum32/HTML/000002.html
and
http://Elsmar.com/ubb/Forum5/HTML/000237.html for some other thoughts.

registrars are so glad to see any effective action ,,,most don't diddle about CA vs PA unless the actions you document are weak, and don't go back to the root cause......I agree Correct/fix it, and prevent its recurrance, ID the similar impact on other stuff then PREVENT the occurance....head it off at the pass so to speak

How about some PA examples for an internal test facility...

Would the testing of materials be considered PA?

How about keeping extra supplies on hand to prevent downtime of a piece of equipment?

Any other ideas?



Analysis of data...would it apply to the material or the equipment or both? How would you determine?

Al,

I wouldn't mind providing some info for the afforementioned benchmarking section.

Maybe it's me being a bit 'tik, but I couldn't locate it.

Could you point me in it's direction.

Thanks & Regards,

Chris May

Ken,

I wouldn't generally consider testing to be a preventive action because it's purpose is to either verify that conformance exists or not - after the fact so to speak - and doesn't prevent the occurrance of nonconformance itself. (although you could extrapolate that doing a test on incoming materials will prevent the use of bad stuff that would affect the conformity of your own product...)

In your second example, the keeping of supplies handy won't prevent the equipment failing....but it would reduce the impact of the failure so I would consider that to be a preventive action.

There are all sorts of little things that can fall under preventive action: tweaking a filing system or developing a checklist or placing a spill tray within easy reach or sending email reminders about computer safety or even ordering plastic sleeves for papers that are used in damp areas. All those things that you do to head off a potential problem simply because you thought about what could go wrong.

Just thought I'd throw the testing out for an opinion.

You mentioned the little things. Things you'd just normally do in the course of a day. Things you never documented.

I'm working on trying to document a preventive action procedure for ISO17025. Sort of like you know what you do but you can't put it into words where it would make any sense. :o

Thanks for your reply.

Ken

Dont think the testing undertaken by a test lab could be PA.

When talking of a test lab, and considering ISO/IEC17025 one must consider the product of the test lab to be accurate test data/reports. As such any PA must be targeted at achieving accurate test data.

The testing itself may be a PA for the general company, but not the lab.

PA in this instance should be things like ensuring the test method is valid, environmental conditions are properly controlled, suitable equipment is employed, uncertainty of measurement is known, etc, etc. In fact all the things that ISO/IEC17025 tells you to do !

Alf and Russ, Ihate to throw in another term, but here goes. Corrective Action Impact - Learning from our mistakes. The mold would fall under this.

If a PA is to prevent a problem that hasn't happened yet (or any of the before mentioned definitions) how do you document the fix you put in place actually worked?

Maybe the problem you defined would never have happened anyways?

Just curious as to how you would handle this. :confused:

Ken,
Why and when are you testing materials? If it is for verifying conformance then, as Lucinda stated, it is not a preventive action. If you are testing materials in the design and development phase of a project to determine the window of acceptance for the product it would be a preventive action. Its not the type of activity or action that determines if it is preventive, it is the purpose, and perhap timing, of the action. Many actions that are taken in the design phase are preventive while the same action in the production phase would be corrective.
Russ

Russ, we do annual material certifications on all products we produce and we also do quite a bit of development work for new materials.
The reason I asked the questions is because I am trying to write a Quality Manual for ISO17025. This is what the guidance manual I'm using states for 4.11 - Preventive Action:

Preventive action is a pro-active process to identify improvement opportunities, rather than react to the identification of existing problems or complaints. The policies and procedures should reduce the likelihood that a non-comformance will occur. Apart from the review of the operational procedures, the preventive action might involve analysis of data, including trend and risk analyses and proficiency testing results. Once a need to improve has been identified, an action plan should be developed, which includes tasks, assigned responsibility and timing.

Once the plan is implemented, it should be monitored to verify that the needed improvements have been realized and effective. There should be some evidence that a follow-up of the effectiveness of the preventive action has been performed. Several methods exist for doing a follow-up. One method is a formal process by which each preventive action is reviewed and the effectiveness is evaluated. The effectiveness of the actions should be evaluated within a reasonable amount of time to assure that the activity has not slipped into nonconforming activities. If it is discovered that the corrective actions are not effective, the laboratory should have a process for ensuring that further investigation is performed to determine why the actions were not effective.


Whew! OK - have you read this over...do you understand why I'm confused. How do you write a policy and procedure to cover this?

:frust: :frust: :frust:

Ken,

This sounds like the Design of Experiments approach to me...but I could be wrong.

FMEA maybe??

Also, if you find a problem and fix this, then that is corrective action, but when you have fixed it so it doen't recurr, it can also be preventive action.

In an SPC environment, one of the data required is the number of Defect Opportunities.
Analyse what could possibly go wrong and prevent it. (Risk Analysis ??).

Regards,

Chris

Per me -

Corrective action is - to prevent recurrence
Preventive action is - to prevent occurence

Some sources of preventive actions are :

1. Employee suggestions
2. Internal Audits
3. FMEAs

I agree with the earlier post - throw out the term "Corrective Action Impact" and you get into a bigger discussion. QS (the big three) in their wisdom brought this corrective action impact into the picture and now the picture is even more cloudy. What we use to call preventive now could fall under corrective action impact. It is harder and harder to find a true preventive action without tying it back to some type of corrective action. It has to be addressed in a practicle method that allows for the understanding of what your organization accepts as CA, PA, and CAI. Then through in mistake proofing and continuous improvement, which really all of the projects should be CI or you are wasting your time. For all of the audits I have been on, this is one of the biggest discussion during the audit.

We have developed an "Actions Matrix" in where we list at a minimum one of each of the types of actions - corrective, preventive, corrective action impact, continuous improvement, and mistake proofing. We try to draw a line between them so as to clearly and confidently call them what we feel is the best catagory. Our facilities have gained an understanding as to our definitions of these actions and it has been effective during our quality team meetings and audits. We require at least one type of action for each catagory to be open and being worked on. This helps us with continuous improvement and problem solving.

Just my thoughts.