Pull Planning

Like a Popeye saying, "You are what you are and you gots what you gots." MRPII is doing the (qualified) best with what you have today. JIT is a goal. MRPII embodies pull planning, making the product just before it is needed. The logic is cascaded through every planning step, saying, if this is when the material is needed, then the precursor needs to be started this much earlier - very linear logic.

Mapping what the lead-times and rates are, is very valuable. A MRPII program is a fancy calculator that applies planning constants - rates and lead-times to offset, in time, every planning step and determine the necessary amounts, based on lot size logic. Basically the programmer must capture this type of information and program it into the calculator. Note that the lead-times may be long and lot sizes large. Today the programmer must develop their program based on the current situation. But at the same time they run the risk of automating a problem. However, this was the second problem. The first was overcompensating for not even knowing what the planning constants were exactly. Overcompensating translates to making more than is necessary and doing it earlier than necessary. An integrated MRPII program solves overcompensation.

JIT seeks to eliminate the non valued-added waste in the supply chain. JIT is the endeavor to reduce to a minimum the planning constants - lot sizes, setup times, lead-times, scrap, unnecessary steps, etc. The results of the JIT effort would allow you to go back and reprogram the calculator to use the new improved planning constants. In this way, the goal would simply be, having planning constants that are at their physical (theoretical) limits and having perfect knowledge & confidence in these numbers, all programmed into the MRP planning calculator.

Push Planning

Still, the pull planning of MRP and JIT is neither feasible nor optimal. What is this, JIT and MRPII wont solve all my problems? It certainly wont help you raise your kids. Something else is needed. Once this is accepted, then we can move on.

To illustrate this point consider these situations - seasonality causing demand to exceed capacity during the seasonal peak; demand exceeding the overall capacity; the JIT goal is a long ways off due to budgetary or resource constraints; or the business is very capital intensive as in the process industry. This is where scheduling comes in.

Making product earlier than is needed is push planning. Ignoring the need to make push planning part of the planning process is like driving a car that is hot and fast (we'll call it the JIT model). Its not going to go anywhere until the light changes. The light is a physical constraint to the car's speed. The linear logic described does not recognize that, when two things require the same resource at the same time (a physical constraint), which one should get it. Just wanting it, isn't enough. Soloman was willing to split the baby but the mother chose. Important point. Who chooses and based on what? Production doesn't know what to do until this choice is made.

You see MRPII and JIT narrows down the choices and establishes the need date, but did not answer when it could be realistically produced, or should be.

Part of the Planning Process

Strategic/Tactical/Operational Planning

Answering this question should have started a long time ago. Planning typically is done in phases. Long range aggregate supply demand balances are used to strategically determine the utilization of existing capacity and whether new capacity needs to be authorized. Once a year, capital plans are reviewed for new products and adjustments to capacity for existing products. The decisions made during this planning step provides the available capacity and the basic structure of the supply chain.

Tactical planning deals in the medium range to make decisions about incremental adjustments to the capacity and/or customer service levels. Sales and Operational Planning is performed once per month to monitor and adjust plans where incremental changes can be made. Changes to rail fleets, storage, contract capacity and major swings in raw material purchases will require three to five months. The decisions during this planning step provides narrower boundaries and guidance necessary to do the scheduling exercise.

Operational planning is another name for scheduling. In the context of the previous planning steps, operational planning characterizes the conversion of a plan that can change, without penalty, to a schedule in which money will start to be spent.

Converting a Plan to Committing Resources

Defining the planning constants was not very important prior to this step. However, the objective must stay in clear focus. That objective is - a Quality schedule, so that everything comes together just at the right time. There is a real tradeoff between detail and accuracy. By definition the more detail, the less accurate. If a raw material needs to be ordered months ahead then the level of detail required may just be by week. All calculation could then be done on a weekly basis. The exact delivery date could be firmed up later. The point is to reserve a given quantity of material within a specific period.

It wouldn't make sense to call a vendor up every week to change the exact date of delivery. Plus or minus a week doesn't matter to them more than one month ahead of time. Operational planning covers the range of commitments - where loose plans are communicated to when specific schedules (to the day or even hour) are conveyed.

Thus far the role of scheduling has been defined. From some of this discussion there is a degree of inference made about its importance. Did you get it? MRPII will double book the schedule. Some customers are not going to get what they wanted on time unless some push planning is applied. There is much more at stake. Push planning is only one aspect of scheduling.

Impact on Asset Management

Inventory

Push planning causes inventories to build in a make-to-stock environment. A different sequence will cause the first scheduled product (A) to carry additional inventory for the length of the second scheduled product's (B) campaign. Therefore, enough A must be made to satisfy demand for the full length of B's campaign. If five products are cycled through the same equipment, now the inventory is going up geometrically for A, B, C and D. Some questions jump out. Is there enough storage for this much A, what is the value of A verses the other products, is there a shelf life problem, how good was the forecast as far out as A was made? These questions will impact the sequence picked for the schedule and the size of the campaigns for all affected products.

Capacity

The challenge of scheduling has to do with the satisfying multiple objectives simultaneously. Customer demand was the first. The second is physical capability, having enough capacity. Every time equipment is changed over from one product to another, capacity is lost. This is in conflict with the inventory scenario just discussed. Inventory consideration promotes more, smaller campaigns.

Scheduling is directly impacting fixed and working capital. Combined with it's potential affect on customer service, quality scheduling is critical to your business. Hopefully the point has been made about how important scheduling is and its strategic position with respect to MRPII and JIT. In other words, no matter how advanced those MRP programs are in your organizations, scheduling is a long term part of that picture. The last segment to discuss is how to get control over scheduling.

Satisfy Demand

Its not adequate to talk about the decision making process of what the schedule should be, without understanding what is driving the process. Understanding demand is as tricky as the sequencing question, but must be put under control to get scheduling under control.

Orders

Demand is an ambiguous word. In a make-to-order business, demand is simply customers orders that have been booked. The schedule is then driven by backlog.

The process industry is not so simple. It assumes that its business is make-to-stock. This type of business must schedule its production by forecasts.

Production must be started before all the orders are available. As the process industry goes toward more diversity and specialty markets, it can no longer afford to make all of its products in a make-to-stock mode, however, some portion of manufacturing will still need to be started before the orders are available.

Forecast

Picking orders over forecast to drive production is not a free will choice, if you will pardon my regression into an old philosophic predilection. It is dependent on the responsiveness of the supply chain and customer service expectations. Let's assume it takes one day to package and four days to make the unpackaged product. If the customer is willing to wait five days - no problem. One day, then you must forecast the generic product and the last step is order driven. This is the classic Make-to-Assemble environment.

The intent here is to give you a flavor for the considerations. This whole area is another topic in itself. Before we leave it one more point would be appropriate though.

Order/Forecast Reconciliation

If it is assumed that scheduling is forecast driven then the question is - when is the forecast changed? The temptation is to monitor the orders and react immediately to orders exceeding forecasts and slowly to orders below forecasts. There needs to be a conscience game plan on how to react to these situations - how long to stay the course and how much to change when its time.

The slower the supply chain, the less a business is able to react to changing market situations. But let's also flip this around. A car can also be over driven. Speed up - slow down - speed up - slow down. Similarly, quality can often suffer from changing rates too often. There is no point of reacting faster than the system allows you to. Compromise is the key until the JIT effort solves this problem.

Assume we have massaged all this and the manufacturing plant now trusts the demands placed on it. Basically the demand pattern is controlled and the process is demand driven. The first objective has been reached. This would be equivalent to a single in a baseball game.

Feasibilize Constraints

A few constraints that cause push planning are, production capacity, labor, limited raw material availability and storage. Only two will be discussed to make the points on how to manage constraints. With a little latitude on the English language, we will call the process of leveling constraints, feasibilizing the schedule.

Production Capacity

The first constraint is production capacity. Leveling this constraint means sequencing the order of products made, changing the start date of the campaigns and its size, effectively changing the number of required changeovers.

Raw Material Availability

Raw materials may only be available at certain times and/or only at a fixed maximum rate. This will cause the scheduling to sequence the order of products made, change the start date of the campaigns and their size, and cause adjustments to the rate in which product can be made.

These two constraints illustrate that different scheduling parameters are used to feasibilize the schedule depending on the cause of the constraint. Other constraints will similarly required their own set of parameters. Feasibilizing the short term is a double. Feasibilizing the short and long term is a triple.

Optimize Costs / Profitability

The scheduling job could stop after satisfying demand and feasibilizing the schedule, against the constraints, and often does. Just doing these two steps does as much to improve customer service as carrying more inventory. But neither insures profitability. Scheduling has a direct affect on costs.

Setup costs

Block operating equipment means changing it over when switching from one product to another. Lost capacity is a hidden cost especially if more is needed and it is contracted or additional capital is spent to achieve it. Besides loosing capacity, every setup incrementally adds costs. Clean-out and scrap are two big contributors plus retooling (or repiping). The fastest growing cost area is the disposal cost of the solvents used for cleaning the equipment.

These are significant factors to the process industry. Since the process industry is very capital intensive, there is a strong desire to fully utilize available capacity. This means block operating several products on the same equipment. The point is, it does not come free. The "extra" cost is the setup costs. The combination of lost capacity and setup costs is a real motivation to minimize the number of campaigns. The results are high inventories.

Inventory Carrying Costs

The value of inventory can be argued but I will assume it to be the investment value plus the costs to maintain that investment. Thought of that way, the value is 25 to 35% of manufactured costs per year. The cash flow implications of this magnitude dramatically affects profitability.

The trade off between set up costs (including lost capacity) and inventory carrying costs is determined by the timing of manufacturing relative to demand and the number of campaigns. Finding the best cost position between these two is the optimizing process. That's a homer.

Many things can be optimized in the scheduling exercise - capacity, customer service, costs and profitability. These optimization objectives are often in conflict with each other. I would preferentially pick either costs or profits to optimize.

Capacity only needs to be optimized when there is not enough. Even then improving capacity should only be done relative to profitability. In other words, the extra inventory carrying costs could override the savings from setup costs and extra generated revenues from more product.

Customer service insures the long term. The net demand on manufacturing combines the sales forecast with the change in inventory caused by a stocking strategy, which should have been done with a fixed customer service level objective in mind. Therefore, customer service was already taken into account in the "Satisfying Demand" step.

Quality

Satisfying demand, feasibilizing and optimizing the schedule together are the characteristics of a quality schedule. These characteristics symbolize the traction between the tires and the road. The best MRP system and JIT program can not operate without a quality schedule. This is the qualification on MRP doing the best with what you have.

Stability

Committing delivery to the customer from inventory is fairly straight forward. It's either there or not. However, committing delivery based on the schedule assumes that the schedule will not change. The schedule therefore requires stability. Only a high quality schedule can maintain stability.

A Longer Plan

There are three dimensions to the job that a scheduler must be sensitive to since their time to do the job is limited. The first is the quality of the schedule as detailed above. The second is the number of products they are responsible for. The last is how far out in time they try to schedule.

There is a requirement for how far out the schedule should be developed. It has to do with the lead times. Commitments are made at various stages in the supply chain. In the ideal world, the schedule would be developed and frozen for the entire lead time. That is usually not practical. Optimally, then, the schedule should go out as far as the majority of costs have been incurred.

Control of Costs and Assets

In plain language, let's not lose sight of what is the real objective for scheduling. It is not merely insuring the delivery of product on time. As illustrated, inventory carrying costs and set up costs are major contributors to the total delivered cost for the process industry. Scheduling is a critical process in controlling costs. Bottom line, you better schedule better than your competitor.

Return to Home Page