Correct Me If I’m Wrong
I once heard a gentleman offer a group of quality
professionals the following explanation of corrective and
preventive action: “If you sent your customers the
wrong product, the corrective action would be to send them
the right product. Then you would do a root cause analysis
on that and take a preventive action to keep it from happening
again.”
This example is misleading. It exemplifies a common misconception
rather than dispelling it.
If you sent your customer the wrong product, you might
correct the error by replacing it with the right product.
This action constitutes mere correction (i.e., eliminating
a detected nonconformity). If, however, you act to eliminate
the cause of the error, you’re taking a corrective
action. Before any corrective action can be taken, one
must first determine the cause of the error.
After determining a cause, any actions taken to eliminate
that cause and its effects are corrective actions.
Effective corrective actions eliminate the cause’s
underlying problems, thereby preventing such problems from
recurring. Therefore, preventing a known problem from recurring
is part of corrective action. In other words, a corrective
action fixes the problem and all like it so it doesn’t
happen again. There’s a preventive nature to corrective
action.
Should the problem recur because of the same error, then
the corrective action wasn’t effective for any of
a variety of reasons (e.g., the root cause was misdiagnosed,
the chosen action didn’t fully eliminate the cause
or the action wasn’t properly
implemented).
A preventive action, on the other hand, actually precludes
potential problems from arising in the first place. Preventive
actions operate only upon potential problems. They don’t
prevent existing problems from recurring; corrective action
does that.
For the hair-splitters, what follows is an example that
contrasts corrective and preventive action by illustrating
how a preventive action might arise in
connec-tion with a corrective action.
In a manufacturing company, nonconforming product was
found at a machining center. An investigating team determined
that the parts were mistakenly loaded into the machine
backward, which resulted in nonconforming product.
The team first reasoned that the problem might be caused
because the parts were loaded into the machine by hand.
However, further investigation into this hypothesis revealed
that no problems had ever been detected with the other
20 part numbers that were also hand-loaded. The team decided
that any action to control all hand-loaded parts merely
because they were hand-loaded would add no value and prove
counterproductive. The team concluded that the problem
wasn’t specifically caused from hand-loading
parts, but that these particular parts were prone to orientation
errors.
Thus, the root cause was confirmed to involve “orientation
errors.” To appropriately address the problem, any
corrective actions would have to focus only on those parts
that were prone to such errors.
The nonconforming parts were corrected (i.e., reworked).
A jig was developed to prevent them from being loaded backward,
thereby mistake-proofing this step in the process. This
proved to be an effective corrective action because it
eliminated the problem’s cause and was verified to
prevent the problem from recurring.
Only one other part among the 20 was also subject to
orientation errors, although as stated earlier, no problems
had ever been detected with these parts. However, because
the same root cause was recognized to exist in that circumstance,
and with a similar probability of error, a jig was introduced
for loading this part type as well. This action was still
part of the original corrective action because the same
cause--an orientation error--was being eliminated in another
circumstance where it existed. Although this action might
seem preventive because no problems had been encountered
with this type of part, it was--strictly speaking--corrective
because it eliminated another instance of the same existing
cause.
Another of the 20 hand-loaded parts was very small and
required particular attention to load into the machine.
This part wasn’t subject to orientation errors, however,
so the earlier corrective action, as defined, didn’t
apply. Although no problems had ever been encountered with
this part, the lessons learned from the previous corrective
action suggested that a jig could mistake-proof the loading
process for this part as well. Accordingly, a third jig
was introduced to the process.
Introducing this jig into the loading process was a preventive
action. Why? Because it prevented the occurrence of a potential
problem whose cause was different from that addressed by
the earlier corrective action. Because the third jig prevented
potential problems from this new source, the similarity
to the earlier problem doesn’t qualify this action
as corrective. Rather, introducing the third jig was a
bona fide preventive action.
Dan Nelson has worked with ISO 9000 for 10 years.
Nelson is a QMS 2000 Principal Auditor (IRCA) and earned
a master’s
degree in business administration from the University
of Iowa.
He consults under the business name Quality Crossroads.
|