Finding the Root Cause - Which technique is better?
- Thread starter Terry
- Start date
This is the key question. Your first cut of problem solving tools should be based on whether the problem is common or special cause.
5-Why is typically used for special causes. The Fishbone may be used for either, but is not effective in identifying the root cause by itself. It identifies a lot of potential causes that must then be ruled out or confirmed by other means.
You may also want to consider
Fault tree analysis or the Statistical Engineering Diagnostic Tree approach (see page 119). Both tools may be used in such a manner that you rule out specific branches using data in order to funnel your investigation (Diagnostic Tree approach).All methods have their strengths and weaknesses. You should become competent is a number of approaches, and use the correct approach for the problem.
Last edited:
While corrosion is a natural chemical reaction, it can be eliminated by preventing the oxygen to react with the metal.
Paint or any other coating comes to mind.
Stijloor.
I ran into this some years ago and while the OP didn't specify the entire scenario, in the case I was involved in a change in shipping methods was the cause. Shipping was changed to rail. I was involved in (gasp) moving a manufacturing plant in the US to a plant in Mexico. We were building up an inventory (cast iron brake cylinders) by building ahead and inventory was shipped by rail instead of the usual trucks. So - Shipping time (cars sat in Loredo, TX awaiting customs, etc.) and differences in weather conditions (temperature, humidity, salt in the air, etc.) were the culprit. Solution was packaging changes (now THAT was expensive!).
D
DPeter67
I would start simple and use the 5 why first. Be careful when documenting your "why's" and allways read them backwards using "Threrfore". This will help keep you from listing symptoms instead of causes. If 5 why does not get you there, try the fishbone. Use Man, Method, Material, Mother Nature, Measurement, and Machines as your spines.
The 8D method is even more robust and should include the "Is/Is not, 5 Why, Fishbone. With any of these methods, you should verify the root cause by being able to turn the problem on or off.
The 8D method is even more robust and should include the "Is/Is not, 5 Why, Fishbone. With any of these methods, you should verify the root cause by being able to turn the problem on or off.
The true culprit was really Oxygen. Only secondary was the failure of your ogranization to adequately protect the product from the oxygen. As you mentioned -- expensive packaging was required.
Since the root cause is oxygen their work is almost complete. All the original poster needs to do is focus on those areas of his process in which the metal is exposed and whether there is adequate rust protection. My educated guess is to look at 2 possibilities first -- incomming raw material that has been sitting exposed or improperly packaged finished goods. In the sceme of problem solving these are the easy ones.
It's simply not that simple
My response was to your post:
To quote you:
So - The *real* Root Cause of that non-conformance wasn't a simple "...oxygen reaction...". The Root Cause was insufficient planning (analysis of changing shipping methods/transportation suppliers) during a manufacturing plant relocation project. "Changing Shipping Methods and Transportation Suppliers without Analysis" was the assigned Root Cause.
Attempting to assign cause to basic aspects such as a chemical reaction simply are not necessarily applicable as a Root Cause. Aspects such as a chemical reaction are often a symptom of a Root Cause. You are assigning a symptom as a Root Cause. That doesn't fly.
My response was to your post:
I thought the reply was a bit simplistic.
Well - Yes. Any rust is an oxygen reaction with the alloy. Differences come into play including the alloy composition. But the root cause wasn't a reaction with oxygen considering the entire scenario. It wasn't a process issue. That was well established as the process was many years mature. That also addresses receiving inspection and other typical aspects. The complaint was rust on parts upon receipt of the product by the customer.
To quote you:
You have to do an analysis which covers the entire process from receiving inspection to arrival at the customer's facility. Part of a good, effective, cost efficient root cause analysis is an evaluation of the scenario as a whole. Theoretically there is a single root cause for a non-conformance, but we all know that's theory (as opposed to reality) because of the variables of every individual process.
Yup - Like I said. We were moving a line to Mexico and during the planning stages a different shipping method and route was the cause.
So - The *real* Root Cause of that non-conformance wasn't a simple "...oxygen reaction...". The Root Cause was insufficient planning (analysis of changing shipping methods/transportation suppliers) during a manufacturing plant relocation project. "Changing Shipping Methods and Transportation Suppliers without Analysis" was the assigned Root Cause.
Attempting to assign cause to basic aspects such as a chemical reaction simply are not necessarily applicable as a Root Cause. Aspects such as a chemical reaction are often a symptom of a Root Cause. You are assigning a symptom as a Root Cause. That doesn't fly.
Claes Gefvenberg
Admin
This is too simplistic. Take an example of a fire started from oily rags. Say that oxygen caused that. Or the oil caused it, or the heat caused it. In reality, a fire requires all three (heat, oxygen and fuel). Take away any one of those and there is no fire. So which is the root cause.
Apollo Root Cause Analysis teaches one philosophy that is worthwhile in this type of situation. That is that there is no such thing as a simple root cause in isolation. Causes are a combination of an Action and a Condition. Another RCA approach added that an Inaction could also qualify as the Action.
In this situation, steel and oxygen are conditions. There may have been an inaction such as failure to apply a rust preventative, or an action such as touching the parts with bare hands.
The 8D approach requires that you look for three causes: Cause of problem, cause for escape and systemic cause. I have found that the Action or inaction is usually the problem cause, Inaction the escape cause and conditions the systemic cause.
Remember - the most correct answer to anything is: "It Depends."
Guys, there is no reason to complicate it. The root cause for rust is, in fact very simple -- the reaction of oxygen and iron. Nothing more, nothing less. Here on earth, you can only manage/reduce the occurence of rust.
Once you know the root cause, you know where to look for OTHER causes -- ie; failure to coat or package properly.
"Changing Shipping Methods and Transportation Suppliers without Analysis" is only the reason the problem appeared, not the cause of the problem per se. To my thinking, it doesn't go far enough. If the person doing the analysis didn't know why rust forms, they may not have even considered packaging issues. Only after you know the root cause -- oxygen and iron -- can you zero in on packaging issues and take action.
Once you know the root cause, you know where to look for OTHER causes -- ie; failure to coat or package properly.
"Changing Shipping Methods and Transportation Suppliers without Analysis" is only the reason the problem appeared, not the cause of the problem per se. To my thinking, it doesn't go far enough. If the person doing the analysis didn't know why rust forms, they may not have even considered packaging issues. Only after you know the root cause -- oxygen and iron -- can you zero in on packaging issues and take action.
Similar threads
