Separating tactical and operative planning for master scheduling

Master scheduling has to deal with an increasing number of transactions, especially if a company goes through  a lean transformation. The lean transformation leads to shorter lead times and smaller batch-sizes. Having reached the ideal of batch size one may suddenly increase the number of transactions that master scheduling deals with by factor 10 to 100.

This is especially true if your configuration is set for your operative scheduling - usually covering few weeks, or in the lean ideal a few days. But if you use the same configuration to calculate your tactical plan, that usually spans over a couple of month up to a year, the volume of transactions can be considerable. At a customer review we found more than 7000 planned transactions for a category C part - a part that should be planned on minimum stock level instead.

In a meeting I had with the planning team of the customer, it became transparent, that one of the challenges is to configure the lead times as well as the positive and negative days for master scheduling. You need to start with analyzing how demand patterns of different products and materials behave. Unfortunately, the formula "longer is better" does not apply here. Having 200 planned purchase orders in the future for an item that has a purchase lead time of only few days is - in lean terms - waste.

To limit the runtime of master scheduling on one hand and the number planned orders and action messages on the other side, It is recommended to work with 2 different plans:

• An operational plan, that uses fences and other parameters from the coverage groups and item coverage, not overwritten by the plan itself. The operational plan is updated every night.

• A tactical plan that overwrites the planning and coverage fences to a longer horizon to make a full explosion in all details. The tactical plan is re-calculated once a week, usually on the weekend. Firming of planned orders should usually not   be done based on the tactical plan.

 It is recommended to structure the coverage groups by demand patterns and use item coverage only to punctually override settings for products, that require special treatment or to evaluate modified behavior.

 Negative / Positive days:

•  Positive days:
  Positive days allows master scheduling to use a planned receipt to cover a specific demand, that lies n days before the demand. In other words, it allows master scheduling to plan for excess inventory to minimize the number of planned orders. To consider actual on-hand for any demand, the positive days have to be at least as long as the coverage fence. This is valid for low turning products or products that are covered by high security stocks. For quick turning products that have a regular replenishment or receipt per day, week or month  (EPEI = n days) should use positive days that are close to the EPEI. Another way of expressing this: positive days should not be more than MAX 2 times inventory  turnover rate.

• Negative days:
  Negative days allows master scheduling to use a planned receipt to cover a specific demand, that  lies n days AFTER the demand. It basically allows Master scheduling to plan for being late. It is important to understand that in a supply  chain this setting can cumulate for every planning step with the end result of allowing to be late on almost every planned order. Negative days are needed when the inventory or purchase lead times are longer than the corresponding Sales lead times, to avoid multiple creation of planned orders for the case where a new demand that cannot be covered is placed earlier than the new planned order will be scheduled.

• In AX 2012, the new possibility of Dynamic negative days have been introduced.
  Negative days can be hard to set, as lead time can be different for products assigned to the same coverage group. That is where Dynamic negative days should be used (found under master planning/ setup / master planning parameters). Once dynamic negative days is set, the lead time defined on released product (default order settings, site specific or item coverage) will always be the minimum negative date counted from today. If a demand is calculated in the future, then  negative days will be equal to number negative days set on the coverage group.
  This feature is a significant change in defining negative days, and should be much easier to define. When calculating with dynamic negative days, the negative days settings should now be 0-5 days - depending on the demand pattern of the related product, and not as earlier recommended = lead time. The usage of dynamic negative days will decrease actions and futures. + all issues that are within lead time will always see receipts within lead time. Especially useful for long lead time items, where periodic or min max is used as requirement policy.

For the improvement of the settings of the operational plan the following guidance should be considered:

• The coverage fence of  purchased material should not be longer than double the lead time of the material, ideally only 2-3 weeks longer than the purchase lead time. The operational plan should only contain planned purchase orders that are needed to take action by the planner, not the tactical forecast. If a planner needs a longer term view he can consult the tactical plan. For purchase products the positive days should be mostly equal to the coverage fence. Purchase
  material should not have negative days = the purchase lead time or even better, use Dynamic negative days. Then purchase material should run with 0 negative days.

• Planned transfers that are used for replenishment and internal logistics between warehouses should not have negative days. The positive days should also be very small, ideally 0 to create the planned transfers at the point in time that the     transfers are really needed.

• Finished goods should only have negative days, if the sales lead times are smaller than the corresponding inventory lead times (or for finished goods that are directly purchased, the corresponding purchase lead time). The negative days would then be the delta between sales lead time (first possible new demand) and the inventory or purchase lead time (first possible new supply). And again, the alternative is to use the Dynamic negative days principle, where again the negative days are calculated automatically based on the lead times.

Master scheduling with Microsoft Dynamics AX is truly a dynamic task. There are many ways of influencing the  planning results and the performance of master scheduling by modifying the right parameters. The resulting plan can only be as good as the configuration allows - and of course, dependent on the quality of the actual and the forecasted demand. We would love to hear about your experiences and your best practices related to positive and negative days and Dynamic negative days. We are also interested in your experiences with multi-plan strategies.

Meet the Microsoft Dynamics AX manufacturing team at the AXUG Summit next week or at Microsoft Dynamics AX Technical conference 2012 in the manufacturing focus room.