Root cause Analysis of Equipment Breakdowns
Most of the actions after machine failure are merely band-aids, a why-why analysis can helps you to identify how to really prevent the issue from happening again.
I have seen that few companies made a common format blindly. Either of WHY-WHY or 5W2H.
5W2H is one of the best format to communicate the problem to management but daily breakdowns should be discussed and analyzed in WHY-WHY format and outcome should be a kaizen or final countermeasure ( Change in JH maintenance sheet or preventive maintenance sheet) to stop it
Just compare the both
- a) What (Happened) ?
- b) Where (Happened) ?
- c) Who (Identified /Responsible)
- d) Which (Part)
- e) Why (It Happened OR Why it is a problem)?
- f) How much (Quantify to understand the severity )
- g) How to do (Action plan to stop)
- h) When (Time line to execute the action plan )
It’s a complete informative sheet to communicate to the management but searching for root cause is absent. So, By 5W2H the objective of breakdown analysis is not getting satisfied and Management hopefully have interest only in actions to stop breakdowns rather than in detail of each breakdown until unless management involvement is require for some particular major breakdowns. Out from TPM, Just see the classic example bellow, It’s a mail communication what I got from 350.org (easily communicated the subject matter in 5W)
So, the solution is
- a) Why
- b) Why
- c) Why
…Still a actionable root cause found.
It is a method of questioning that leads to the identification of the root cause(s) of a problem and So, To find root cause and take permanent corrective action, WHY-WHY is best brain storming method. Others are FTA analysis, MFMEA (Before machine design), Parato (for priority setting).
How to conduct WHY-WHY analysis?
I usually follow the format I learn from Yamaguchi San where WHY should ask to final action and I feel it is easy to go down logically rather than to describe B/D phenomena and ask why it happed.
- a) Conduct analysis season imitate after the breakdown got over ( memory will be fresh and so it is easy and less time consuming)
- b) Ask why to the final action after which machine gets started (Generally we have several actions in way to make the machine up)
- c) Then WHY WHY WHY….
Generally causes of breakdown fall in following five categories and expected final action is-
- Poor Basic condition – Final action on preventive Maintenance to restore and honor basic condition (JH and Or PM).
- Poor Operating condition – Final action on Maintenance (JH and Or PM)
- Deterioration – PM check sheet or frequency is expected to be modify
- Week Design – Cry for a Kaizen (Brain storming) to rectify the same
- Skill – Training to enhance skill in same area.
There are few terms which create unnecessary confusion and debates which should not be. Here are logic
Debates: Root cause analysis= Outcome is One and only actionable cause.
Root Cause analysis is an effort to unearth the underlying causes. It may not necessery that it would be concluded in single point root cause. There may be many causes which is influencing the problem. Bellow Pic is self explanatory (Hats off to copy write owner, I am just sharing it)
OEE = Over all Equipment efficiency. This indicator is multiplication of productivity (man-machine), equipment quality rate, equipment reliability (Availability) rate.
For a given time, in gives the projection of use of man machine potential and the scope of improvement.
Debates: What to consider, shift time or 24 Hrs (Machine potential is 24 Hrs.). of time ?
It is management to decide how many Hrs. it need to run the shift. We just calculate what is the potential of “ready to dispatch” production for that given time.
ð So, OEE will be calculated depends upon a particular facility operating time.
Debates: For a assembly line with different product rate of individual machine, what will be total available time?
It is better to consider Whole line as a single machine and calculate the rest. If one machine affects the product output rate then aggregate OEE of individual machines will give wrong picture.
MTBF = Mean time between failure
= total machining time / no. of stops+1.
Debates : Why +1 ?
Consider the scenario that One machine Ran for 1 hour and the breakdown for 1 Hr. and then again ran for one Hr.
So, MTBF in conventional way it will be,
Machining time 2 Hrs. , Number of breakdown 1 and So, MTBF is 2 Hrs. Which indicates a wrong picture.
So, +1 with number of breakdown comes closer to realiy.t