We are thinking of using defect elimination process to prevent random failures. This is because normal time-based prev maint will only address age-related failures and root cause analysis is reactive in nature ie after the failure happens.

Could anyone kindly share their experience in defect elimination process ie success or failure stories? Tq

Dear Josh,

Is there something wrong with eliminating defects and the underlying causes (physical, human and latent) after the event by doing RCA?
In my view, the learning process often changes mind-sets that cause such failures and can well prevent future failures.

I find that people get hung up on words like pro-active and reactive; reactive work can be very effective when proper RCAs are done. From my perspective, a good RCA program is a good way to eliminate defects in future.

V.Narayan.

Vee,

I have to agree. WHile it is true that RCFA is a reactive method. (After the event) And if things were managed well the need for such a program would be reduced to a bare minimum. (Eg proactive risk management, driving maitnenance and asset management intervention strategies)

But.... this takes time. Sometimes a long time. There is a need to be pragmatic and to eliminate defects where possible.

There are some good case studies on the following sites:

http://www.reliabilityassurance.com
http://www.reliability.com
http://www.apollorca.com

Vee, your comment is absolutely correct. After doing many RCAs, we have prevented recurrences. However, new bad actors come up. (A great shadow show!) So RCA program never ends and should be continuous. Therefore, we think we need to complement the RCA program (not abandoned) and are looking for a more encompassing program like defect elimination program (most likely using FMEA)which we think will look into a wide range of failures (known & expected). Any experiences on this approach?

Daryl, perhaps you would like to explain what you have commented above on proactive risk mgmt, driving maint & asset mgmt intervention strategies? Sorry, sorry what is this thing? TQ

Josh,

When you do your RCAs, at what stage do you call it complete? Do you go beyond the physical root causes and identify human and latent causes? If you do, and apply the learning across the board, you should expect a rapid fall in the number of significant failures. More importantly, you should expect a change in the way people think and behave. If this is not happening, perhaps your RCA process needs review.

Proactive measures tend to be expensive, in terms of time and money. It may be better to defer them till you get some of the basics right (cleanliness, lubrication, joint tightness, alignment, balancing and PM compliance). Do you feel you are ready?

RCM is a great way to 'catch' defects before they occur. The first four steps of RCM(Functional FMEA) do this well. You can also use FTA or FMEA to do defect identification. The difficulty is in knowing where to start, on which item to work. So you can end up with a mountain of analysis without a clear path.

Stopping known defects at source through RCAs is usually more efficient and cost-effective.

V.Narayan (Vee)

We always try to reach the latent cause. What "learning process" do you refer to here? What do you mean by "across the board"?

Josh,

A well done RCA produces recommendations covering physical, human and latent causes of failure.

When you implement those relating to physical causes, the immediate problem get solved (e.g., the bearing failed due to lack of correct lubrication).

When you eliminate human causes, failures of other equipment also get eliminated, e.g., if the incorrect lubrication was due to poor oil quality, then this affects other machines not connected directly to the current RCA.

When you eliminate latent causes, you can make improvements in the whole facility, e.g., if the latent cause was loss of skills over time. re-training will improve people performance not only on these machines and just in lubrication, but also on other issues over the whole facility (across the board, both geographically and over a range of skills). Further it will help change attitudes and behaviours that prevent future failures, especially if the RCA was done by people directly involved in the incident.

Does this make sense to you?

V.Narayan.

Thanks mate, making a lot of sense. Would you like to have a look into an RCA done by me recently at your free time to identify any areas for improvement? TQ

Josh,

I suggest we do this offline.

V.Narayan.