iSixSigma

FMEA

Viewing 26 posts - 1 through 26 (of 26 total)
  • Author
    Posts
  • #45240

    pedro
    Participant

    Hi…
    I used to work in automotive company, and as you know they handle FMEA’s and as guide the FMEA manual created by the AIAG is used..
    I am now working in a Fiber Optic company.. I was reviewing the FMEA process to try to implement it here in this field.
    But the Severity Evaluation Criteria table is set up for vehicle manufactures..
    I was wondering if there is a simmilar table for general products or somebody have had the same experience and wanted to shear how the table was used.
     
    thanks
    Pedro

    0
    #147108

    Dale
    Participant

    The ranges in your definitions need to be applicable to your environment.  The underlying definitions normally don’t change.  And consistent across your company.  a $1 million severity might be low in a $100 million process.

    0
    #147147

    Monk
    Participant

    Pedro,
    The FMEA will work the same way. the only difference will be the severity of 8 and above. In vehicle manufacturing you have failure modes that can lead to safety issues, where as in te fibre optic, there may be very few or no instances of safety occurance due to process failures.
    So you may have very few faliure modes with a severity of 8 or above.
    I have used FMEA in transactional processes as well and it work well there as well. All the best !
    Monk

    0
    #147167

    NEBB
    Participant

    Pedro,
     
    I have a similar working experience i.e. Certified as a BB in automotive environment & now working in a similar field to yourself.
    I have used a similar set up to the AIAG ratings for severity, but changed the occurance & detection to suit the needs of my company eg 1:50000 etc for occurance would mean 100 year of production for us at present!, so we have reset the ratings to something like 1 – every year, 2- every quarter 3 – every month 4-every week 5-every day (the 10 point rating has been reduced to 5 by us as its best fit. As a previous poster suggested, if you are going to formulate your own scoring ‘ratings’ / ‘criteria’ ensure that this is used “across the board” at your company to reduce / eliminate subjectivety.
     
    hope this helps
    regards
    NEBB

    0
    #147747

    Philip
    Participant

    The key thing to remember is that the Severity, Occurance and Detection levels are not numbers. They are ordered categories which have nemerals as labels. FMEA is a very powerful process for SUBJECTIVELY assessing risk. However, multiplying S, O, & D together to get an RPN is mathematical nonsense. “Numbers” can only be multiplied together if they are from a ratio scale of measurement (i.e. a severity of 10 is 10 times as bad as a severity of 1, and twice as bad as a severity of 5, etc). You should either use a three digit code for RPN (e.g. 2-5-9) and subjectively rank those, or make your S,O,D ratio-scale by using the actual value of the risk ($), a true probability (0 to 1) of Occurance and a true probability of (non) detection. This will give you an “Expected” loss or “Regret” that you can prioritise. Multiplying together nemeral labels for ordered categories is a recipe for confusion, which is why there are so many overrides (such as “a severity of 10 must always be addressed” and “a severity of 1 should not be assessed” etc)

    0
    #147752

    Monk
    Participant

    Philip,
    What you have written or suggested is a total nonsense. PL. don’t misguide people with your experiments.
    FMEA is a tried and tested tool and it helps to bring sense to things that we ‘subjectively’ access. Thats the reason, RPN is the multiplication of severity, occurance and detection. Moreever for prioritising, we have to use the severity factor and not rpn.
    Monk

    0
    #147753

    Philip
    Participant

    This is not nonsense, it is mathematical fact.
    If you perform mathematical operations such as multiplication or division on ordinal scale data you will end up with answers that are nonsense.
    Despite the widespread use of FMEA, Risk Priority Numbers do not prioritise risk, and are not even numbers. It is perfectly possible to get high RPN for low risk items and low RPN for high risk items.
    FMEA is an excellent, and powerful, SUBJECTIVE tool, which is destroyed utterly by attempts to make it objective. Much success derived from FMEA has largely been obtained from the fact that practicioners use common sense and often ignore the RPN calculations.
    You yourself state that prioritisation is better done in terms of Severity, which is tacit acknowlegement of the limitations of RPN.
    I am disappointed that correspondents to this forum should try to stifle important debate in order to promote mathematical ignorance.

    0
    #147754

    sophos9
    Member

    Philip, what you have written is not correctThe RPN is an overall reflection of the priority of risk that needs mitigation, this allows for tools such as CDAM to be used across a common measurement system and easy propagation into a ParetoIt is also common to check on the severity flag, a severity rating of 8-10 should warrant mitigationsophos9

    0
    #147755

    sophos9
    Member

    I’m not sure that anyone is trying to stifle debate. People are assisting in the clarity of answer to the original question.All the practitioners I have worked with (IBM, Rath & Strong and multiple global organisations) use the FMEA in the way demonstrated.The FMEA is not a ‘be all and end all’ its just another tool used to deliver a robust solution

    0
    #147756

    TJ
    Member

    Monk,
    You’re quick to complain when you think others bash you, but you’ve got no problem bashing others. Is this why you hide behind a multiple aliases?
    If you’ve got a problem with Philips argument – provide a counter argument – don’t just bash him like you bashed the other poster this morning.
    TJ

    0
    #147757

    Monk
    Participant

    TJ
    I only use one alias and it is consistent. So your accusation is totally wrong. I use only one alias.
    Coming to bashing. I have just put my opinion. I have not made any personal remarks. Pl. read my message once again and you will release.
    In any case, if the originator feels that I have offended him, I owe an apology and I will ask for it. It was not my intention to insult anyone in this forum.
    Monk

    0
    #147758

    Monk
    Participant

    TJ,
    I agree that I was a bashful in my other poster, but I have been forth-right in putting it. This is done in good intention.
    Monk

    0
    #147760

    TJ
    Member

    Monk,
    On the surface you seem a reasonable guy, but you seem to enjoy being blunt.

    Some people would find this offensive, especially when they have a wife and children to feed.
    You  told Philip he wrote nonesense.
    He’s trying to open people’s mind and teach them how to think. If you haven’t learnt anything that is your loss. As someone who wants to learn and would like more insight from someone like Philip – I find your attitude offensive?
    You seem to be able to hand it out well enough – but can you take it without cascading into a multitude of other aliases.
    TJ
     

    0
    #147761

    Monk
    Participant

    TJ
    You seem to repeat the same mistakes.
    You can give your opion to the related message.
    Let Pilip put his case forward and if you are a disguise of Philip – the lat line of the previous message applies to you very well.
    As for me, I am monk and will remain monk. I don’t use any other alias.
    Monk

    0
    #147762

    TJ
    Member

    I am not Philip .. and you can’t take someone else being blunt! You can only hand it out!

    0
    #147764

    TJ
    Member

    There you go again. I disagree with you and you resort to bashing me.
    What don’t you understand about a visitor to this forum wanting to hear more from people like Philip?
    Just because something is in a book, doesn’t mean it’s correct. Just because everyone accepts it’s correct, doesn’t mean it’s correct. What’s wrong with a different point of view? Just becuase you say it is, doesn’t mean it so.
    If Philip is wrong, but up a counter argument using the same alias. Don’t suddenly invent a new character just in case you loose the argument.
    You can hand it out, but you can’t take it!!!!!
     

    0
    #147765

    CDH
    Participant

    I, like many of the other poster have used FMEA in a variety of businesses and industries. I totally agree with re-working ratings to ensure that they are relevant to your company, remember we are aiming to prioritise risks in a specific process within a specific organisatio, so teh FMEA should reflect this. With regards to the argumant concerning multiplyimng the scores to determine an RPN. Personally, even if it is not mathematically correct, I do not see taht as an issue. The whole point is to identify the major risks from a list that generally runs into tens and potentially hundreds.
    My view,for waht it’s worth,  is that the FMEA tool does this more than adequately.

    0
    #147767

    equinox
    Participant

    Sophos9,
    Philip makes a good point …. It is possible for and intermittant fault with a ‘car radio system’ to have a higher RPN than a fault with a brake line – This will occur on the basis that the end customer is more aware of it. Therefore, Detection and Occurance scores will be higher.
    RPN Numbers in themselves, have no value or meaning. RPN numbers should only be used to rank potential process weaknesses for consideration  .. Sometimes the use of the word ‘Priority’ is mis-used ..
    As in some paretos, The most obvious element is not the most important one …
    ‘Severity’ should always be the governing factor ..
    Kind Regards,
    Equinox.

    0
    #147768

    equinox
    Participant

    TJ + Monk,
    This forum is considered to a valuable base for people who wish to develop and improve their environment .. Some people are struggling with Change and risistance from old cultural bastians of industry ..
    It would be nice if we all felt supported by like minded people without finding an ongoing cat fight when we are seeking advise ..
    Please consider other peoples time and if you wish to carry on please swop E mail addresses and leave the rest of us to get on with real benchmarking and personal development ..
     Monk, before you consider an acidic reply .. Dont bother,  I never argue with stroppy people, Invariably, they drag me down to their level … Then beat me on experience ..
    Equinox.

    0
    #147769

    Torrance
    Participant

    Agree with both sides of the argument.
    Agree with Phillip – it’s important that issues that would be for example “STOP SHIP” should not be lost into a calculation. They must remain as fundamental issues – no debates, must be fixed. Identify these, and take them out to be dealt with seperately.
    Agree with Monk – those issues that are not critical (safety etc) can be “prioritised” by the RPN. Otherwise we can be focusing on the minor points rather than the major ones.
    What’s wrong with this balance?

    0
    #147770

    Rob M
    Member

    I agree …  we need both sides of the argument. Telling someone they’re talking nonesense is a conversation killer. It’s better to seek clarification.

    0
    #147772

    equinox
    Participant

    Davy T,
    Good point, well presented.
    Will have to improve on my ‘Lean’ techniques …
    Regards,
    Equinox. 

    0
    #147776

    sophos9
    Member

    EquinoxI do not agree that the statement you made that RPN’s have no meaning or value for the following experienced reason. The RPN is the driver for further focus/mitigation, without it, project/programme resource time can be ill spent working on issues that have no or little value. As most lean/sigma tools, its a tool that assists organisation and mitigation NOT a tool that solves allI more than agree with the issues surrounding ‘severity’ which is why I flagged it, again, ANYTHING that has made it on the FMEA must need a further sense check/mitigation, the only difference is the granularity of investigation. This is what the function of the REVISED RPN is forMeasurement systems must also correspond to the process/project, there is no one size fits allWhilst I partially agree with your statement regarding Pareto’s, in my experience and in Sigma, if you have quantified the mitigated FMEA data and used the REVISED RPN, the Pareto should be correctBest, Sophos9

    0
    #147813

    RC
    Participant

    Hello Monk,
    I agree with Philips approach, I think it’s very innovative. People shouldn’t listen to you.
    I think you’re a shame to the six sigma profession, how about that?
    RC

    0
    #147839

    Monk
    Participant

    RC
    Do you remember – how you struggled to do your DOE ? Would you like to share he lessons learnt with all here.
    So you know now, how you have put yourself in shame.
    Monk

    0
    #148007

    SB
    Member

    Grow up, people, and stop wasting the time of those who are trying to gain insight from others knowledge, experience and points of view. Take your personal fights somewhere else.

    0
Viewing 26 posts - 1 through 26 (of 26 total)

The forum ‘General’ is closed to new topics and replies.