All Member Open Forum

 View Only

  • 1.  MTTR and MTBF Measures

    Posted 10-25-2019 08:46 AM
    ​What is best way to measure MTBF? Yearly, Quarterly, and etc?
    Also Automotive manufacturing, when measuring MTTR and MTBF on unit production not asset uptime. Where is the best locations to measure at, Departments or Zones within a department?

    ------------------------------
    Shaun Dick
    Equipment Reliability Engineer / MES SME
    Honda Manufacturing of Indiana
    Greensburg IN
    ------------------------------


  • 2.  RE: MTTR and MTBF Measures

    Posted 10-25-2019 09:57 AM
    Hi Shaun, I can't help but give you the most unsatisfying answer of all time: It depends.  It depends on what you're trying to achieve, it depends on the timescales of the problem, etc.

    A general rules of thumb about MTBF, as a metric, is: you should use if for long enough that you capture the time scales of the *relevant* failure modes you're trying to capture (eg: equipment that gets *touched* a few times a year requires a shorter time versus equipment that gets *touched* every several years), but short enough that you don't lose temporal trends - eg, my reliability significantly improved/worsened 2 years ago, but I'm calculating over 5 years so my answer is washed out.  There's always a trade off to anything in life, and this is the general trade off I've learned from estimating a classic MTBF metric.  But of course, it also depends on what you're trying to do with it too when you make decisions about what you want to trade off.

    And that applies to departments or zones - it depends on what you're trying to do.  And of course it also depends on the data as well.  For example, If two zones capture data totally differently but my application is at the department level of interest, you may still want to look at the individual zones because combining the data may be nonsensical.

    So, this answer is totally unsatisfying, but at least I hope I gave you some factors to consider in thinking about an answer.


    ------------------------------
    Sarah Lukens
    GE Digital
    Roanoke VA
    ------------------------------

    Original Message


  • 3.  RE: MTTR and MTBF Measures

    Posted 10-28-2019 10:47 AM
    This may not be the 'best' way, but when I start to assess MTBF and MTTR, I look at the last 12 months of WO data by equipment number to identify 'bad actor' equipment. (I focus on Emergency and Corrective work orders and exclude PMs). I get a baseline and trend from there. If your historical WO data is valid, you can go back earlier than 12 months, its a judgement call. I also cross reference the bad actors WO data Vs downtime bad actors.  ​

    ------------------------------
    John Cray
    Life Cycle Engineering
    N. Charleston SC
    ------------------------------

    Original Message


  • 4.  RE: MTTR and MTBF Measures

    Posted 10-29-2019 07:43 AM

    Shawn, your question (What is best way to measure MTBF? Yearly, Quarterly, and etc?) is actually two parts.

    MTBF is measured upon the failure history of the asset or body of assets in equivalent operating context. The calculated MTBF, which will reflect which of the six failure shapes applies to the asset. Your only choice is the window of history, you wish to base the MTBF calculation upon (see slide 48 of https://analytics4strategy.com/trn-modelinsigh ).

    The second part asks when you want to make the calculation for reporting or analysis purposes. That you are inquiring of report or analysis date, implies that you may find yourself calculating on multiple windows and comparing them.

    Richard G. Lamb, PE, CPA
    Educational website: https://analytics4strategy.com




    Original Message


  • 5.  RE: MTTR and MTBF Measures

    Posted 10-30-2019 11:40 AM
    ​Agree with Sarah on it depends on the type failure mode of the asset or component of the asset.  For example, an antifriction roller bearing will have a typical life in years not months.  If the roller bearing fails in weeks or months, you will not need any report or analysis to know about it.  Everyone will know there is a major problem.  MTBF on a roller bearing may need data from a larger sampling of similar bearing applications or a long time period 10 years or more.  However, a contact lip seal may have a full life at 4-6 months so a shorter time frame of 12 months may be sufficient.

    I personally mostly use MTBF for asset level analysis but it can be used I believe successfully too across an asset class or type.  For example, it I wanted to evaluate MTBF on all 100HP, 1800rpm motors across multiple areas then that could provide some useful information especially if we compared to 3600 rpm motors.

    If failure modes and equipment types are all combined, I'm not sure there is as much value comparing MTBF and MTTR.  Just my 2 cents.

    ------------------------------
    Randy Riddell, CMRP, PSAP, CLS
    Reliability Manager
    Essity
    Cherokee AL
    ------------------------------

    Original Message


Related Content