Finding the Root Cause - Which technique is better?

Marc

Fully vaccinated are you?
Leader
What I am also seeing here is why defining the problem is so important in a problem solving process. A wrong perception or definition can throw one far away from the actual or root cause.
I fully agree.

This problem is being made far too complicated! Rust occurs through inadequate protection or inadequate storage. Determine where in the process the rust was detected then work upstream looking at where the process went wrong. Is the protection adequate at each stage? Is the storage area and packaging suitable? Is there good stock rotation?
The causes and countermeasures should be apparent!
The key word is *should*. I remember one scenario where rusted parts showed up in the customer's assembly line. There was a lot of in-house investigation and it took a couple months before we got the customer let us send observers to their assembly plant. All the time they had been screaming at us, it was their own receiving and storage method which was the cause.

I went through a couple of similar situations working with a company that was supplying parts for the V-Rod some years back. Until we got in and watched their processes. Our team several times ended up showing Harley where their own internal processes were at fault.
 

Jim Wynne

Leader
Admin
I fully agree.

The key word is *should*. I remember one scenario where rusted parts showed up in the customer's assembly line. There was a lot of in-house investigation and it took a couple months before we got the customer let us send observers to their assembly plant. All the time they had been screaming at us, it was their own receiving and storage method which was the cause.

In this post (Re: How much third-party sorting is unnecessary) I told a story of a nightmare situation where a lot of effort and running around was wasted because of poor RCA which was occasioned by not having samples of the allegedly NC parts and by the customer jumping to conclusions about the source of the "problem."

I went through a couple of similar situations working with a company that was supplying parts for the V-Rod some years back. Until we got in and watched their processes. Our team several times ended up showing Harley where their own internal processes were at fault.
I could tell a few stories about V-Rod development that I can't tell because I would probably be hunted down and killed. :tg:
 

Marc

Fully vaccinated are you?
Leader
Harley-Davidson

I assume Harley gave up worrying about me, having worked both sides (first for them directly as a consultant, and in later years for some of their suppliers). Obviously I can't get into too many details, but I have come pretty close a number of times in some threads here. I've never had a 'love' of/for Harley. Its been too long ago for me to still be a thorn in their behind. None the less I admit there's a lot I could tell that probably could still get me into trouble.
 

charanjit singh

Involved In Discussions
You are absolutely right Marc. A good understanding of various analytical techniques plus the knowledge of the processes involved throughout the chain is essential.

I remember a case where a customer reported open-circuit failure of a small percentage of pieces in a batch of wire-wound resistors even before the components could be used in the circuit. And the application was critical. Though we supplied batch after batch of resistors regularly, this phenomenon occured at random and in some batches only. We did utilise the Ishikawa Diagram & 5M analysis and brain-storming sessions but they were really used as guides to the actual investigation. Utilising our knowedge base concerning various materials, storage, handling, processing conditions (coupled with common sense of course) the trail led us right from customer's premises to the manufacturing operations of supplier of a critical raw material. It took us good 2 years to hunt down the cause that was proving to be a mysetry in the initial stages.
 

Terry

Involved In Discussions
Not being a fan of fishbone diagrams or root cause brainstorming, I would suggest an alternate approach.

I typically begin with some Problem definition questions (such as those utilized by the Kepner-Tregoe 'Is, Is Not' approach: starting from the very good point that the actual root cause of rust is a well known reaction, I might ask if the parts have some type of rust proofing that has been compromised. Does the customer require that these parts not exhibit rust by contract/drawing requriement? Is this a chronic problem or are these 22 parts with rust something that hasn't been seen before? Are these 22 parts all from one lot? what is a PDI report? were these defects found internally? or by the customer? how many customers?

asnwers to these questions can help direct us and will begin to provide clues as to the next steps...
Thanks! PDI is pre-delivery inspection within our company. Every month the inspector will send PDI report to relevant personnel. our goal is zero defect before delivery.
Terry
 

Terry

Involved In Discussions
IMO, choice of Problem solving tools depends on the needs, I mean the nature and depth of analysis required
(Sometimes the Customer needs, of course :D)
it is our needs, our goal is zero defect.
Thanks!
 
Last edited by a moderator:
As examples go, I have a beauty: It was my very first day as an inspector, and I was told to have a careful look at a fan assembly for the cabin AC in our construction vehicles.

After one glance I said: -The wires are crossed...
The answer was: -Huh?

I repeated my statement and as it turned out, the wires really were crossed when compared to the drawing (It was in fact one of the few things on the drawing that could be checked at a glance.

Further investigation revealed that this had been the case for months, and as a consequence.... the fans had been running backwards! I also learned that there had been a problem with poor performance for a long time, and tests had been run without turning up the cause. Oops... :notme:

Needless to say I got off to a pretty good start there, in my very first quality related job... :D

/Claes
 

Bev D

Heretical Statistician
Leader
Super Moderator
Terry - what about the other questions? Or do you have some good ideas to move forward from the other posts?
 

Terry

Involved In Discussions
To the OP, Terry's question, yes use a Fishbone Diagram and list all causes. For the example you mentioned use a 5M method - Machine, Method, Man, Materials, Measurement and Environment for your causes.

Once identified all causes, go on to use the 5 Whys to come to root causes.

Pl let us know your progress of investigation.
Thanks for your reply.
I will take your ideas.

Thanks!
 
R

Ron Tollett

Would anyone have workable Fish diagram I could use? Is there such a thing?
 
Top Bottom