iSixSigma

QFD Example/Help

Six Sigma – iSixSigma Forums Old Forums General QFD Example/Help

Viewing 11 posts - 1 through 11 (of 11 total)
  • Author
    Posts
  • #34580

    Coffee
    Participant

    I am working on a QFD to create the ideal sales system. The idea is that we can should be able to create the ideal system on paper then compare that to what we have and create a gap analysis and plan to close it.
    While I understand the QFD and can explain it I am having difficulty putting together a detailed plan to execute.
    Does anyone have any advice / direction or examples??
    [email protected]

    0
    #95477

    rwatters
    Member

    First Step=Voice of the Customer
    Usually done by building a “house of quality”

    0
    #95501

    Coffee
    Participant

    I got that part.
    What I really want is to talk to someone who has actually completed one of these things (with real results). It seems to me that everyone that I talk to Starts them and never Finishes them. My guess here is that they (like me) know all the book stuff and get the concept but don’t know how this much work can have the desired ROI.
    I am going to do one of these for a major initiative and don’t need to be winging the results. If I can’t see the goal clearly I am not going to be able to teach that to the team.

    0
    #95502

    Anonymous
    Guest

    QFD is a great tool .. but some people place too much emphasis on the Voice of the Customer. Good products still need great technologists and great design engineers.
    Since I am not a salesperson, let us first consider a transportation system. Image a bus travelling along a route and stopping at a number of stops. If we ask passengers what they would like, they will usually mention that they would like the bus to arrive on time.
    Now to achieve this, the bus must arrive at the bus stop by the scheduled arrival time, but as we all know a bus system is subjected to delays due to loading and unloading passengers, selling tickets, and rush hour traffic. These factors are noise and the only defense against noise that can’t be elimianted is scaling.
    To use scaling, the ideal bus system should be capable of arrriving at the bus stop much earlier than the scheduled time. In other words, the bus should be capable of travelling very fast and unimpeded. It could then wait at the bus stop and until the scheduled arrival time; presumeably this is why some bus routes have special traffic free lanes, although few have safety belts suitable for fast travel.
    Your case is similar … in the ideal sales system, the time spent on potentially large orders should be much larger than the time spent on small orders. However, salesman are usually measured by the number of calls they make and amount of orders taken, which means that some sales people might not be willing to spend all their resouces on getting one big order in case they can’t get it before the end of the month, and of course spending too much time on small orders is inefficient. So my recommendation is use the VOC by all means, but also ask the design engineers and the ‘next operation as customer’ to consider the ‘ideal system.’
    Now this approach may appear contradictory,  but when we applied it to the design of an image setter, the design engineers designed one of the most reliable film handling systems in the world – Sumo.
    Good luck,
    Andy

    0
    #95557

    Coffee
    Participant

    Andy, I absolutly agree with you. I plan on conducting the VOC and a VOE (Voice Of Engineer).
    That being said, do you combine these in the first house?
    Or do you do these in the second house? (which makes a little more sense I think).  
    Then what is the output from this house?

    0
    #95576

    Anonymous
    Guest

    Cofee:
    We used the Phase 1 Whats to capture both the non-technical and technical voice of ‘the end-user.’ Using this scheme the Hows became a system of metrics and not a design concept.
    In Phase 2, as well as the progressed Hows from Phase1, we also added the voice of the engineers as  ‘the next operation as customer’, and this included the R&D deputy director, technologists, design engineers, technical servicing, quality, and manufacturing, but we broke it down into metric categories, such as cost, part variety, MTBF, etc. The Phase 2 Hows then became a system design proper and we structured it by sub-assembly, right down to the component level. While developing the system concepts we performed a number of cardboard modelling experiments and other parameter design studies around the ‘Ideal function’ of each sub-assembly.
    One of the most significant contribtions of this approach was the realisation that each component designer had assumed that they had the entire system tolerance range for themselves, which would have spelt disaster for the entire project! Fortunately, tolerancing did not become a major issue because our analysis of ‘the ideal’ function led to the discovery of how to desensize a film handling system; something several Japanese design engineers originally believed only applied to chemical processing, which just goes to show how important it is to challenge engineers constructively and see the joy on their faces when they make a breakthrough!
    Good luck,
    Andy

    0
    #95579

    John J. McDonough
    Participant

    Coffee
    QFD is a wonderful tool, but like a hammer, it is a tool.  You can use it in a hundred different ways.
    In a DFSS context, you would typically begin by doing your information inventory.  From that, you can design your VOC.  Generally, from your VOC you synthesize a set of requirements. 
    Depending on the project, you then distill these requirements through some sort of affinitizing.  How far you take this is something of a judgement call.  The QFD process can be breathtakingly expensive, and the cost is driven by the number of requirements.  You need to use some discretion in getting sufficiently detailed requirements while not exploding the cost of the QFD.
    Now, you need to score the requirements, usually using AHP.
    The requirements go on the left in House 1.
    On the right, you put your competitive benchmarking.  You score your competitors, usually 1,2,3,4,5, on how well they meet the requirements.  You also score your target product.
    Along the top, you generate measures for each of your requirements, and you put the target specifications for each measure along the bottom.  In the roof, you identify conflicts and synergies among the measures.
    Above the specs at the bottom you put the weight of each measure, calculated from the requirements weight and the stregth of the association between the requirement and the measure.
    Beneath the specs, you put your competitive ranking against the measures, and again, you draw on that your intended position.
    From this, you can identify those key measures you will carry forward to help select your design.  Note that it isn’t just automatically based on the weight.  You want to look at the weight in the context of where you want to position yourself in the market.
    This is only house 1.  Once you have the key measures, you can now make the decision as to whether you will use QFD or some other tool to help deploy the functions into a design.
    This is not something that is easy to do the first time, but after you have done it enough times it becomes very comfortable.
    Ditto with examples … of course, examples are always a problem because they are porprietary.  I have a public service example online, but you need to read through all of the material to understand the context in which the QFD is taking place:
    http://www.qsl.net/w8kea/ATVproject.html
    Even though this is a simple project, it can be pretty challenging just to gather an understanding of what the QFD is trying to accomplish.  And as a public service project, it lacks some of the dimensions of a commercial project, particularly, there is no competitive benchmarking.  But, it is one example.
    If you would like to talk through some specifics of your project, I would be happy to chat via email.
    –McD
    [email protected]
     

    0
    #112078

    O’Connell
    Participant

    The problem with this thought process is, of course, that you assume there is a bus!  You have a solution in mind, now you need to design the process and environment around it. QFD and VOC really help you to define the needs at a higher level and consider all alternatives. E.g. the need is to get from point A to point B on time, comfortably, safely, etc.  The Bus is part of the SOLUTION, not the need…

    0
    #112125

    Anonymous
    Guest

    Brian,
    My comments were directed to phase 2.
    I can just imagine facilitating another group of Fujifilm engineers and telling them … before we can proceed and design this system, we have to go back and decide if we really need an image setter after all.
    By the way, if you think the function of a bus is to go from A to B, think again. Buses have to go from A to B consecutively, in other words from bus stop to bus stop, arriving at each stage on target and with minimum variability.
    But thanks for your input and confirming that I am wasting my time visiting this forum.
    Andy

    0
    #112138

    O’Connell
    Participant

    I’m not sure you read my post. I didn’t say that the purpose of a bus is to go from A to B, I said the NEED of the CUSTOMER is to get from A to B. You’re trying to design a need (A to B consecutively) around a solution you already have (a bus).  If all you ever thought about was getting the bus to be more efficient, you might never have thought of a high-speed train, etc.  That’s all.

    0
    #112221

    Anonymous
    Guest

    Brian,
    The two products I worked on are considered world-class. The first is called Pulsar and the second is called Sumo.
    The failure rate of Sumo is so low that it could not be measured within the product development time, which is why we had to develop other means.
    What QFD project have you worked on?
    Andy
     

    0
Viewing 11 posts - 1 through 11 (of 11 total)

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