iSixSigma

Creating a Pull System – Lean Six Sigma

Six Sigma – iSixSigma Forums Old Forums General Creating a Pull System – Lean Six Sigma

Viewing 7 posts - 1 through 7 (of 7 total)
  • Author
    Posts
  • #48639

    Mr IAM
    Participant

    All,  I’m having trouble with a concept.  In Michael George’s Lean Six Sigma Book – page 34.  He lays out the process for creating a pull system.
    One of these steps is to “cap the work in process (WIP)” based upon what lead time your customers need.  He suggests doing this through Little’s Law.
    Suggestion is.. if my customers need a lead time of 8 days (current lead time is 10 days with WIP of 100) then I should take 8*completion rate (10) and then cap the WIP in process at 80 and put the remaining WIP (20) into an incoming triage system.  Going forward I would only realease work into the process when work is completed.
    I don’t see how this reduces lead time.  If we are putting work into a buffer up-front how can we just now exclude this work from the lead time calculation?  I don’t see how this method reduces Lead Time.  It only reduces it if you don’t count the WIP in the triage system as WIP.
    Thoughts anyone?  What am I missing??? Thanks a lot.  M

    0
    #164638

    EdG
    Participant

    Mr IAM,
    If you originally had 100 units in WIP and experienced a lead time of 10 days but now established your process with 80 units in WIP and an additional 20 units in your “incoming triage” the reality is you have done nothing to benefit from chaning your process.  If you look at the wait time that the 20 units in your triage experience and add that to your 8 day lead time I would expect that the total time is still 10 days.  End result: NO BENEFIT.
    If your process can operate without that “incoming triage” then your process is really operating at 8 days lead time.  Until you do, you are only fooling yourself.  Eliminate the need for that triage and gain the benefits…
    Good luck, EdG

    0
    #164640

    Dr. Scott
    Participant

    Ditto what EdG said.
    Regards,
    Dr. Scott

    0
    #164645

    GrayR
    Participant

    One of the fundamental lean concepts is that the ratio between value added time vs. non-value added time in most multi-step processes is very low, less than 5%.  Your thinking  & conclusions are correct only if all of the units or orders in your WIP have extremely high value added/non-value added time ratios, e.g., >40 – 50%; and the orders are pretty much processed as first in-first out. 
    Other than that, capping the WIP does reduce OVERALL LEAD TIMES (= release buffer + WIP).  The WIP times decreases significantly by capping WIP, and the decrease is more than enough to offset any longer time in the release buffer.
    A very good reference for this sort of thing is Hopp & Spearman, “Factory Physics”.  They developed a similar approach, called CONWIP (= Constant WIP.)  It all goes back to reducing the variability in the process.  Their book is pricey, but their website, “factoryphysics.com” has some downloadable papers that go into some detail on managing release points and the meaning behind pull . . .
    Interestingly, a few years ago, I worked with a machine shop having many multi-step grinding, machining and heat treating processes.  The normal practice was to get any firm order out on the shop floor so that the “they had plenty of time to work on it and meet the ship date”.  However, they always had late orders (40% on-time), and and lead times were weeks out and getting longer.  The first cut at controlling inventory (capping WIP) was to reduce the number of orders on the floor by 1 week; and we did this by NOT RELEASING ANY new orders (except for a few expedites) for a week.  Then, after the week, we set up a process where the next order was not released until an order was finished.  Results were seen in one month.  And after about six months, on-time deliveries were near 80%; and total lead times (release buffer + WIP time) dropped well over a week to the final customer.  At the same time, the controller saw better cash flow & profitability in the business. (The next iteration was to take another week out of the WIP; but the company stopped at that point because they saw that some of the machines & operators were idle at times.  In other words, everything business-wise (on-time, lead times, costs, & cash flow) were going in the right direction, but they ran into some of the driver issues, such as traditional production metrics & cost accounting, that have a tendency to kill any lean program.)
    You should go ahead with capping the WIP.  It would be interesting if you could post your results.

    0
    #164646

    GrayR
    Participant

    Just looked back at the factoryphysics website.  The papers are available, but you do have to register (e-mail & password).  The two papers that apply are:
    “To Pull or Not to Pull”
    “Principles for Managers – #4”
    (you probably will need to review Principles for Managers, #1 -#3, to get through the concepts and relationships between time, capacity, inventory & variability.)

    0
    #164647

    Mr IAM
    Participant

    Awesome, thanks for the response.
    So… by capping WIP and level loading to help create flow and reduce/eliminate wait stages between process steps is where the lead time reduction comes from?  That makes sense…  Thanks for you note.  I’ll check out Factory Physics.
    Cheers, M

    0
    #164651

    GrayR
    Participant

    Yes that’s how it works.  As far as it making sense, that’s where it becomes difficult . . . I’ve heard it a dozen times that lean is mostly “common-sense”, but it isn’t intuitive . . .
    Also, just remembered that Levinson came out with a book that explains it similar to Hopp & Spearman (even has a paragraph on CONWIP): “Beyond the Theory of Constraints”.  But it is a little easier read, and less costly ($36).  Good luck.

    0
Viewing 7 posts - 1 through 7 (of 7 total)

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