Preparing a Root Cause Report - Possible Root Causes

[sebastr23]

Was already searching a bit forum, but cannot find...

I am preparing a root cause report (kinda 8D and process audit report in one sheet) and on front page after summary would like to put some choices to choose what kind of root cause was it; system, process, etc. To have on the end some kind of statistics and see what root causes are per each supplier.

So I need your help. If we draw a line, how would you divide root causes?

Thanks in advance

 

[somashekar]

Re: Possible Root Causes

I recollect we were doing this some years back, but then we got rid of it, as its not really of any great use.
We had three :
1. No process in place
2. Process is in place, but is not followed
3. Process is in place and followed, but is not effective

 

[Mike S.]

Re: Possible Root Causes

How about including the "bones" from a fishbone diagram?

 

[Mike S.]

Re: Possible Root Causes

We also run into "inadequate training" a good bit.

Of course, it is all in how far you wanna dig. If you have ability to affect change on it, the true "root" is often best described as inadequate management.

 

[Kronos147]

I recently introduced the following nonconformance classifications to our system:
-Dimensional
-Paperwork (F.A.I.R., C of C)
-Outside Process/Specification
-Finish
-Packaging\Labeling
-Internal Requirement
-AS9100 Requirement
-Process not Completed per Policy
-Risk not identified
-Other

And I have it on my ToDo list to add "Human Factor" as per AS9100 D as one of my classifications.

I'm not happy about the re-introduction of the Human Factor either, as we are now poised to going back to 'blamestorming.' I worked hard at my organization reminding people while determining causes that it is processes, not people.

 

[normzone]

[Mike S.], all too often it comes down to " the guy who owns the company gets what he wants ". He may not have known that his desires would result in these issues, but that's how we get there.

[Somashekar] is correct - you can track it but it's difficult to find ROI for the data.

 

[ChuckHughes]

I have a fixed opinion that all of the "locations" of problems offered by Kronos147 have a root in human interaction. Wrong dimensions are created by and reviewed by humans for the vast majority of machines I visit annually. Documentation is completely dominated by human intervention from creation, review, publication, distribution to reading its content. I don't think any of the items on the list are devoid of human influence at a minimum.

So Human influence on design, manufacture, delivery and service must be acknowledged and managed. And when it is managed effectively good things emerge. Employee morale is key to providing an environment in which failure is not a sin, but an opportunity to learn more about variation.

In our lean and mean manufacturing environment, have we exceeded the capacity of employees to be attentive to doing tasks or are we staffing at the minimum level and, with one absence, the line goes into hyperdrive. Remember Lucile Ball putting chocolates on the cake line? For those of you who remember black and white TV.

From my experience in watching a lot of companies make stuff, the distinction between those that are under water, complaint wise and those that enjoy single digit PPM year after year is the work environment the employee is presented. Positive morale goes hand in hand with low/no defects. Some may call this a chicken / egg situation but I think morale creates defect free environments.

WOW. What a weight to get off my chest.

 

[Bev D]

I always differentiate between specific actionable physical root causes (or I prefer the the phrase 'causal mechanism') of any specific defect/failure and the systemic cause of the physical cause...if I take action on the specific physical cause I can prevent the recurrence of the specific defect/failure. If I take action on the systemic cause I can prevent future similar defects/failures.

An example: an increase in the insertion failure rate of a specific type of chemistry slide (into a chemistry analyzer) was traced back to a change in the chemical makeup of the ink on the slide (used to create the bar code and other markings). The chemical change increased 'softness' of the ink and as the slides are stacked together with other slides the soft ink stuck the slides together. The ink change was the specific physical cause and when corrected the failure went away. The systemic cause was a lack of validation of the unintended consequences of the change. This systemic cause had occurred several times for failures resulting from deliberate, intended, process/product changes. Our systemic change was to improve our change assessment and V&V process. (Not particularly popular with some of our engineers but we did experience a marked decrease in the number of failures resulting from changes that we were notified of).

Anyway, we came to this conclusion regarding systemic improvements not by categorizing systemic failures but by analyzing the failures and performing lessons learned separate from (although including) the team that worked on the specific failures...the management team determined the systemic causes through a deep understanding of what actually happened.

In my experience determining categories ahead of time and allowing the team to 'pick one' leads to over-aggregation of the information, simplistic analysis of the systemic cause and just another 'check the box' mentality. The qualification of the categories then becomes meaningless and in-actionable.

My advice is that the value comes from the analysis and not the ease of 'canned' categories...if your organization is not willing to invest the time and effort into the he systemic cause investigation, no amount of categorization will help you...

Just my 5 cents...

 

[ChuckHughes]

Well put Bev

 

[Mike S.]

Every single tool or method has both advantages and disadvantages that the user should be aware of. But without categorization Pareto charts and Ishikawa diagrams do not exist.