Leveling Production to Demand

Six Sigma – iSixSigma Forums General Forums New to Lean Six Sigma Leveling Production to Demand

Viewing 9 posts - 1 through 9 (of 9 total)
  • Author
  • #55973

    Aaron Olson

    This is my first real question/post so I’m sorry if I don’t provide the right information or ask a stupid question.

    I’m working on a certification project to deal with leveling production vs demand.

    There are 3 machines producing 6 different part numbers. We suspect that to level production with demand we only need 2 machines and occasionally part of the 3rd when demand increases but we need to prove it so that we don’t fail our customer. With the extra capacity, we could bring in some fabrication that we have been outsourcing (costing about $250k in the last year).

    Part of the objectives would be to:
    -Control the finished goods inventory (it has been piling up very high recently, and sitting for months on our shelves; space is an issue in our shop)
    -Stop Over-Producing
    -Utilize excess capacity to increase profitability

    So here are my questions:
    -Does this seem like a possible six sigma project as opposed to just lean implementation? This is hard for me because we don’t really have a good system for accurately determining when and how much to produce for these parts.

    -What measures might be used to help control the output (production) to the input (demand)?
    My thoughts were to use:
    Sales Qty
    Production Qty
    Inventory Qty

    Thanks for helping, I just want to know if I’m on the right track.



    It’s more of a lean project than SS since it doesn’t involve finding and reducing the causes of defects, defined as variation outside of customer specifications. I’d recommend that you use Theory of Constraints to reduce bottlenecks and “pull” to optimize production within constraints in order to match demand. Sure, there’s lean six sigma which strives to blend SS and lean, and 6-ToC which blends SS and ToC. And failure to meet demand is a sort of defect from the customer perspective. But I wouldn’t go there to make it an SS project.


    Aaron Olson

    @Straydog Thanks for your response.

    If the ability to meet demand whenever they needed it were one of the top factors in keeping their business, would that change your thoughts?

    Our motto that we have printed on our trucks and our boxes is essentially ‘If you order today, you’ll get it tomorrow.’ That’s what we sold them on, so an inability to meet that is very rough on our business relationship. It doesn’t help that they account for an obscenely large chunk of our business and therefore are always a focus of management.



    @aaronolson No, I wouldn’t change my thoughts. In project management we teach that there is a difference between functional and non-functional requirements. Functional requirements concern what the product, service, or whatever the project is developing actually does. Non-functional requirements are things such as costs, deadlines… Six sigma deals with identifying the causes and reducing the rate of defects — variations outside of functional specifications. Admittedly the distinction gets conflated if your business is, for example, delivery. Then failure to deliver on time is a functional defect. But from what you say, your business is production. Matching production to demand is much more about lean than six sigma.


    What you want to do is optimize production and costs given several constraints (i.e., demand, machine capacities, labor availability, etc.) which is a classic linear programming problem that can be solved using Excel’s Solver function. You can first try to meet demand using only 2 machines and then try three. You’ll need to forecast demand using past sales and also any information your Sales Department can provide.


    James Heidt

    I agree with all comments, and agree this is a linear programming problem. Check out a YouTube video on using Excel to solve it with linear programming.


    Mike Carnell

    @aaronolson I wouldn’t worry about whether it is SS of Lean. At the end of the day if to tell someone it is SS and it isn’t nothing changes. If you tell someone it isn’t and you are wrong nothing changes. This is one of those things people will make an issue of if they don’t know what else to ask you.

    I do believe you need to take a different view of this. Rather than sit back and wonder how to make this work best for production you need to start at the end of your line and understand what has to happen at the end of the line to satisfy your customer. From that starting point work backwards into your process and figure out how to make it happen. Once you have that process defined you can make modifications to Level load your line but understand that when you move from the plan that satisfies your customer to a plan that is level loaded for you then you have introduced risk in terms of the waste of over production.

    You could look at the model from working backwards (the goal of that model is to deliver the right material at the right time i.e. efficient) to how to make it work. Once your goal shifts from your customer to level loading your factory so you are no longer focused on the customer.

    Just my opinion



    This can be a good six sigma project because it requires some data analysis not found in the Lean toolkit.
    Question for you to answer:
    1)Is your business a cyclical type business that has seasonality? Find out
    2)What is the acceptance rate at each machine? What is the primary causes of defects
    3) Is this a multi shift operation? Do you have shift to shift variation details?
    4) Operator to operator variation?
    5)Are the product being produced similar? If yes can you stock partially completed inventory to reduce leadtime?
    6)Have you documented the entire process from order placement to shipping to determine the flow and times involved in each?

    Let me know how it goes.


    Aaron Olson

    @Ziggymarlow Thanks for your answer and your questions.

    So far throughout the project, we are beginning to measure and validate our capacity, historical demand, and our current inventory levels.

    To answer your questions:
    1) It is a little bit cyclical. These parts have only been through 1 inconsistent year so it’s hard to tell.
    2) Not sure what the acceptance rate for these parts is. I would need to do some digging.
    3) Multi shift with moderate variability for output and quality
    4) Lots of Operator to Operator variability
    5) The parts are different 4 different lengths with 2 different raw material types (different extrusions) but we do stock the raw machined part and have them finished according to the order’s color specification
    6) We are finalizing our Value Stream Map in the next couple of weeks while we figure out some times associated with each step

    We also recently learned that we have a tool that is meant to indicate when to produce and when to hold, based on having an 8 week supply of inventory on-hand for our customer. That tool hadn’t been followed and led to building up about twice the suggested levels. That was an interesting find.

    And to clarify, our goal right now is to determine our capacity and lead time, use that to determine proper inventory levels so that we can ramp up and down without overproducing beyond customer’s demand. With this information, we want to be able to use the extra capacity to machine parts that are being outsourced.

    So I think what we need to do is figure out what process (or processes) is the source of variation in scheduling and producing for orders; also what needs to be done to shorten lead time and reduce variation in the output of each machine/shift/operator.

    Does that sound like a good direction?

Viewing 9 posts - 1 through 9 (of 9 total)

You must be logged in to reply to this topic.