If demand were only constant, managing productive systems sure would be a lot easier. The sage who made this statement prefers to remain anonymous because he knows that, although it is so true, only a dreamer dares to think about an ideal world. And the sage is not a dreamer but a vice-president of manufacturing who works in the real world of random demand variations and seasonality. He must be concerned with the utilization of facilities and work force and the level of inventories.
When inventories seem too large, the president complains about the investment tied up and the costs of carrying them. When inventories are low, the marketing vice president complains about poor service to customers. When top management decides to lay off workers during a sales slump, the Union President complains and sometimes makes threats and even manage to have visits from influential political officials pleading for whatever employment stabilization measures that are possible.
This manufacturing vice president has the responsibility of producing high quality products at low cost and timed to be available when the market wants them. In addition, he must consider trade-offs between cost and employment stabilization, cost and market timing, and market timing and employment stabilization. When product demand is seasonal, the problem of achieving the needed balance among these factors is much more difficult.
Most managers want to plan and control operations at the broadcast level through some kind of aggregate planning that bypass the details of individual products and the detailed scheduling of facilities and personnel. This fact is a good illustration of how managerial behavior actually employs system concepts by starting with the whole. Management would rather deal with the basic relevant decision of programming the use of resources. This is accomplished by reviewing projected employment levels and by setting activity rates that can be varied within a given employment level by varying the hours worked (working overtime or under-time). Once these basic decisions have been made for the upcoming period, detailed scheduling can proceed at a lower level within the constraints of the broad plan.
What is needed first for aggregate plans is the development of some logical overall unit for measuring output, for example, gallons of paint in the paint industry, cases of beer in the beer industry, perhaps equivalent machine hours in mechanical industries, beds occupied in hospitals, or pieces of mail in a post office.
Management must also be able to forecast for some reasonable planning period, perhaps up to a year, in these aggregate terms. Finally, management must be able to isolate and measure the relevant costs. These costs may be reconstructed in the form of a model that will permit near optimal decisions for the sequence of planning periods in the planning horizon.
Aggregate planning increases the range of alternatives for capacity use that can be considered by management. The concepts raise such broad basic question as the following: To what extent should inventory be used to absorb the fluctuations in demand that will occur over the next 6 to 12 months? Why not absorb these fluctuations by simply varying the size of the work force? Why not maintain a fairly stable work force size and absorb the fluctuations by changing activity rates by varying work hours? Why not maintain a fairly stable work force and let subcontractors wrestle with the problem of fluctuating order rates? Should the firm purposely not meet all demands? Each of the preceding policies may be termed a pure strategy because only one policy is used to smooth production activity.
In most instances, it is probably true that any one of these pure strategies would not be as effective as a balance among them. Each strategy has associated costs. We seek an astute combination of the pure strategies called a mixed strategy.