Informational Nonconformances and Defects - Operator Error, System Error, or both?

[Mike S.]

After thinking about the "certified shipment" thread a bit I decided to try my first poll. I hope I don't screw it up! So here goes...

Some very learned folks here believe, in the event of a nonconformance or defects, that “the system is always to blame” and never the operator. Deming proposed the system was at fault about 96% of the time (someone correct me here if my percentage is wrong – I’m going by a very fallible memory), but that “operator error” did exist. What do YOU think? Feel free to expound on your vote by posting as well.

 

[noboxwine]

Good First Poll Mike !!!!

Operator Error does exist from time to time. We are mammals, not machines and special causes occur. But, in my experience I'll have to agree that nonconformances occur 95% of the time, due to a feeble system----especially repeats. In our business culture, however, it's much easier----- and we feel better by placing the blame on an individual, single them out as incapable and shift the attention away from ourselves-----the creators of the system-------rather than accept the need for change. It would admit that we are human and some of us can not accept that. Plus, it's much easier to complete an 8-D by saying "oversight"-------"operator retrained". Gimme a flippin' break. Signing off my soapbox----noboxwine.

 

[db]

To err is human....I mean system

Just doesn't sound the same does it? If we find there is a system error, could that be traced back to a human one (management)?

Consider: An employee inserts a part incorrectly in a machine and nonconforming part is formed. Human error, right? Well, perhaps it is a system error for allowing the employee to insert the part incorrectly (error proof the operation). However, should someone not have considered this possibility during planning? Oh another human error. But what in the system allowed the potential insertion error to be overlooked? We could go on and on.

The bottom line is even system errors will eventually be traced to human errors.


How do I come up with this stuff?

 

[Sam]

I'll stick with the old 80-20 rule. However when the mistake has been made percentages really don't count and it doesn't matter who/what made the mistake. What matters is lessons learned and actions implemented.
As long as there is an operator there will always be an operator error.

 

[Craig H.]

Sorry, but I can't vote on this one - the system and those within it are not seperate, but overlapping, and, to make it more interesting, fluid.

As with human/machine interfaces, human/system interaction is sometimes hard to delineate, especially when you consider the thought processes (training?) that come into play.

Add to this the fact that human beings can change their outlook from minute to minute, depending on their environment, some of which is hard to control. Is the fact that someone's spouse yelled at them before work part of the system?

This is an interesting thread, but I will be surprised if a definitive answer is found.

It doesn't matter what business you are in, it is a people business.

 

[Michael T]

Here we go again...

(Steps up onto the soapbox that was recently left by noboxwine)

Ahem.... (taps microphone)

Okay... we've been down this road before and I'll reiterate my stance.... People can be the cause of a non-conformance. It is foolish to rule them out. It is also foolish to say they are always at fault. Start with the system - if you eliminate the system as the root cause of the non-conformance - then you are only left with the operator (provided the raw material has been removed from the equation).

Take, for example, Joseph Worker. Good guy (or gal...we can say Josephine if you want... ) Jo has been on the job for 10 years and (contrary to Chickenlips' problem from another thread) always uses the work instructions to do the job. Lunch time roles around and Jo goes to lunch. During lunch he receives a phone call from his mother saying his favorite uncle has just passed away unexpectedly. Mom and Uncle Fred are on the left coast and Jo works on the right coast - so - funeral arrangements will be made in a few days - time enough for the family to get their tickets on Hotline.com and fly out.

Jo goes back to work - clearly not thinking about work - but rather thinking about Uncle Fred and his poor distraught mom on the left coast. Jo mispacks 20 of the next 100 cartons he is packing.

What's the root cause? (No - eliminating telephone calls at work isn't it... nice try... )

Food for thought for the weekend.

Cheers!!!!

Mike

 

[Mike S.]

MT,

Nice speech; good analogy. According to some, in your example of Joseph Worker, the system was at fault, period. The system is ALWAYS at fault. I have honestly tried to convince myself that this could be possible, but alas I was unable to buy into that philosophy and doubt that I ever can. I kinda went on a rant in the "certified shipment" thread explaining (in part) why I feel this way.

Craig -- I agree "the system" includes and is created by people, but in this case I'm using definitions such as Deming used, separating the "operator" (Joseph in MT's example) from the "system" which is created by higher-ups (people who get paid more!).

 

[M Greenaway]

But isnt the reason we have systems because we employee humans, and they are prone to errors ?

An interesting question that just crossed my mind (having voted 96%) is - at what point do we stop blaming the system and start blaming the operator ?

Can we say we have a good system if it allows operators to error ?

In the analogy given, should we have a system prone to the whims of the emotional state of our employees ?

 

[Mike S.]

MG,

I'm suprized by your vote, but thanks for voting.

As for "should we have a system prone to the whims of the emotional state of our employees " -- how can you possibly do otherwise in all cases?

How could you truly error-proof Joseph's job in all industries/situations? Have someone double check him? No, 100% test won't do it as you mentioned as being only 80% accurate in another thread. In the realm of practicality, how do you do it?

What if Joe is a driver delivering the boxes and he has a wreck -- crashes the truck. How do you error-proof that, make it "not prone to the whims of the emotional state of employees" as you say? Figure that one out and the world will beat a path to your door and you will be VERY rich, famous, loved by millions, a Nobel Prize winner, and solely responsible for saving thousands of lives each year.

 

[Kevin Mader]

Mike,


I wonder what the poll results would be if this were given to non-quality types.

Regards,

Kevin