iSixSigma

Re-locating the fishbone diagram

Six Sigma – iSixSigma Forums Old Forums General Re-locating the fishbone diagram

Viewing 14 posts - 1 through 14 (of 14 total)
  • Author
    Posts
  • #33235

    Ranjan
    Participant

    I am just wondering whether it would be a good idea to move up the fishbone diagram analysis to the beginning of the measure phase or even later in the define phase. The reason for this is the time it takes to get data (if not already present) in the required subgroup form. Any ideas on this?
    I understand that getting it into the define phase could cause some discomfort to the proponents of the system, but I feel it is worth thinking about.

    0
    #89651

    Hemanth
    Participant

    Hi Rahul,Nice to read ur post. Its a good idea to prepone the C&E analysis, parallel to measure phase. Although what we do is instead of full C&E we plan our data collection so that we can use it in analyse phase. One tool which comes in handy is IPO diagram or SIPOC diagram. Its like the beginning of C&E analysis. One problem which I have faced is it becomes difficult to control team once you have the analysis done then they start doing things which changes the process and conflicts with the baseline study.
    But in all I agree that once you have the measurement system in place and the data collection is on, team can start the C&E analysis in conjunction. But anything before this stage…I have my own reservations..

    0
    #89667

    Ranjan
    Participant

    Hemanth,
                  While the SIPOC is a fantastic tool to understand behaviour of your system and the underlying factors that influence the “Y”, it is only “a view from 30,000 ft”. The fishbone on the other hand is a more detailed process analysis tool. In our company, we did a pareto for different stages of pareto and found that the analyse phase takes the longest time to complete (more than the define and measure put together). It is understandable since this is a pretty important phase. However, most of the time is spent collecting baseline data. For this reason I propose that we move the fishbone to at least the beginning of the measure phase and see whether we have data to collect, else start collecting basic level data.
     
    Rahul

    0
    #89675

    Mikel
    Member

    The fishbone is in the beginning of the Measure Phase – where have you been taught it goes?

    0
    #89688

    Hemanth
    Participant

    Hi Rahul
    I understand a lot of time is consumed by analyse phase. But, couldnt get when you said “most of the time is spent collecting baseline data”. We consider this activity as part of measure phase.
    If you are looking for cause verification from your baseline study, then you might run into trouble. The reason why verification must not be part of baseline study is the data which you would have is observational and not collected in a very controlled environment. Thus, you run a risk of confounding causes.
    To sum up, as long as the status quo is maintained for the process its a good idea to do C&E while baseline study is on. But the first and foremost activity for any project must be to establish a measurement system, once you start getting quality data and you have resources avbl, I agree with you start doing C&E analysis.

    0
    #89916

    Edwards
    Participant

    Moving the fishbone exercise from the Analyze phase to the Measure phase makes sense to me.  It provides an overview of all of the potential causes for the problem from which you can determine the critical few to focus on.  In fact, I have used this diagram in the Define phase to show the scope of the project to Senior Management since they always want you to fix everything at once.  

    0
    #89917

    Mikel
    Member

    You guys need to figure out who gave the bozo advice that said this is an Analyze tool. It has been in Measure from the beginning. It is the tool immediately following process flow (get a common base of knowledge on how the process works, then get a common base of knowledge on how the process goes wrong).
     

    0
    #89920

    Bob M.
    Participant

    Indeed, the fishbone diagram is a measure tool used to BRAINSTORM and scope the process at hand.

    0
    #89924

    Mike Carnell
    Participant

    Rahul,
    Just an observation that may or may not help you. Your question is indicative of the way SS has degenerated. The SS process is a method to move through the Y=f(x) problem solving equation. Going from a dependent variable to independent variables, leveraging the x’s to gain a predictable Y. If using a tool some place other than where it was taught (on this point I agree with Stan – whoever put it in Analyze was off track from where it is most commonly used) makes sense then you use it. There are some self appointed SS police that pontificate on this stuff but at the end of the day you don’t end up in Six Sigma jail. What is the consequence to trying to do it and it doesn’t work? Nothing maybe some time but it isn’t wasted if you learned something.
    If your SS program has become such a mindless process that you are just executing tools because someone told you it had to be done that way you are already in trouble. You are trying to use data to drive a solution that is coming from outside the box. When you are so ridgid in the structure of the tools and application that people stop thinking and understanding you just found yourself a different box to live in.
    Good luck.

    0
    #89926

    Ranjan
    Participant

    Mike,
    I appreciate your observations. I would like to think that different projects demand different schedules of application of the tools of 6s. I believe that there is no fixed rule as to when to use specific tools. The more effeciently you get to your end result, the better off you will be. I was taught that the fishbone was done in the measure phase, after the gauge R&R. While I undestand this reason, I question this format in projects that have a good R&R (established earlier by experience?). I suggested that the fishbone be moved up to the Define phase if necessary, so that the top hitters can be recorded and time is not lost in the future phases of the project. What I have typially seen is that projects get bogged down in the Analyze-Improve phase due to lack of data (which in most cases could have been collected earlier).
    Rahul

    0
    #89929

    Mike Carnell
    Participant

    Rahul,
    Like I said – I don’t think it matters where you do it as long as you understand why you are doing it. Personally I don’t use it that much. At the end of the day it is primarily an opinion driven tool.
    I prefer the structured approach like an FMEA. If you are dealing with a soft company culture it is a good tool to use because they always feel better when you give them an opportunity to write on post it notes and walk around (CEDAC). When they get done you end up pulling together data to either accept or reject whatever idea was placed on the chart.
    Just my opinion.
    Good luck.

    0
    #89932

    Ranjan
    Participant

    Mike,
    I agree with you.
    I have found the fishbone to be a very educative tool,especially to people not used to analyzing root cause. I allow group members to freely put on the chart any contributing X and then let the group vote on what we should concentrate on.
    Taking off on the FMEA approach, I prefer to use a short-form FMEA to implement my control plan. Again, this is project specific. I’ve had projects that looked at how a particular feature on a part was being machined. In such specific places, where the oeprator is the main owner of the process (maybe an engineer too) I prefer using a vastly reduced FMEA map to simplify control.
    I’d like to know your thoughts on this.
    Rahul

    0
    #89958

    Mike Carnell
    Participant

    Rahul,
    As I said in an earlier post I am not a fan of C&E diagrams. I primarily use them to placate people in a soft culture who have this burning desire to demonstrate their insight. Group voting is just about as useless. If you have a bad measurement system in place and everyone associated with the process has based their “knowledge” on the data from the measurement system then they are collectively misinformed. Allowing them to vote as a group doesn’t fix that – you will end up with a collectively misinformed opinion. Concentrating on collectively misformed opinions is a waste of time and reinforces the concept that it is OK to run around doing thing with absolutely no data behind your actions.
    If you use a tool like an FMEA you can start them on a path to using data. Frequency of Occurrance is an opportunity to get them working on capability studies. They can base detection on MSA studies. Severity can come from field data, failure analysis, and defect insertion. All are data driven. You can pareto the RPN’s and it sets up the priority for hypothesis testing. What doesn’t get resolved in hypothesis testing or is believed to have an interaction gets pushed to the DOE.
    It is about the flow. The output of one tool is the input of another. Not very complicated.
    We spend a lot of time worried about all these nuances in analysis tools. Part of the reason there has been so much “success” by these people selling their non-math approach to Six Sigma is because the processes as a whole are so screwed up it doesn’t really take much in the way of sophistication to fix something.
    FMEA’s can be used for a lot of things. If your reduced FMEA is working – that is great (so why are you willing to reduce the FMEA and then seek absolution for moving the C&E?) I am assuming you are paid for results – if you are getting them (in the long term as well as the short term) then you have done what you agreed to do when you were hired. If your supervisor sits you down at your next review and says “we were giving you a 5% raise but since you reduced the FMEA we are only giving you 4%” then you have a different deal. That won’t happen – they want results. That is what their review is about and we know that stuff flows down hill.
    We have done FMEA’s, in the past, on a machine. When it goes down the operator can look up the failure mode and find those things that they can do before they call maintenance. It worked well to off load some of the work for maintenance and the operators had more ownership because that document empowered them to do more than just mindlessly feed material into the machine. We have done FMEA’s on DOE’s when the parts were expensive or time was critical. It just gave us a structured approach to assessing where the risk of the DOE going wrong was before we were in the middle of the DOE.
    How you use these tools is up to you. As long as you think you understand what information you need to know and can match it with a tool that will supply it then it is the right thing to do – so you do it. Don’t ask permission – just do what you believe to be right. Anyone who fires you for that – you were probably working in the wrong place to begin with.
    Just my opinion. Good luck.

    0
    #89963

    Spiderman
    Member

    Rahul…
    In my training….Fishbone is taught to be used in a couple of places in the  phases.  In Define, when you are trying to isolate the CTQ….early in Measure to help identify X factors to base your data collection.  And again in Analyze…along with a Impact matrix. 
    But to your point…I tell my BBs and GBs ….to pull from the toolbox the tool that will best help meet the objective of that particular meeting…or step. 
     

    0
Viewing 14 posts - 1 through 14 (of 14 total)

The forum ‘General’ is closed to new topics and replies.