Page 174 - Orlicky's Material Requirements Planning

CHAPTER 8 Lot Sizing 153

of LUC will be larger than the carrying-cost element. The LTC technique, which seeks to
equalize these elements, therefore is biased toward larger order quantities.

Part-Period Balancing (PPB)

This technique employs the same logic as LTC, and its computation of order quantities is
3
identical except for an adjustment routine called look-ahead/look-back. This feature is
intended to prevent stock covering peak requirements from being carried for long peri-
ods of time and to avoid orders being keyed to (i.e., starting coverage with) periods with
low requirements. The adjustments are made only when the condition exists that the
look-ahead/look-back corrects. In many cases, therefore, PPB and LTC will yield identi-
cal results. This would be the case with the demand data used in the preceding examples,
and in order to demonstrate look-ahead/look-back, it is necessary to use different series
of net requirements. The look-ahead adjustment would be operative with the following
net requirements schedule:
Period: 1 2 3 4 5 6 7 8 9
Net requirement: 20 40 30 10 40 30 35 20 40
Planned orders: X X →

With an EPP of 100, the first lot of 90 would cover periods 1 through 3, and the next
lot would be keyed to period 4. But before this is firmed up, a look ahead to period 5 is
made. The 40 units in period 5 would have to be carried in inventory for one period,
which would cost 40 part-periods. If the 10 units in period 4 were added to the first lot,
they would be carried for three periods at a cost of 30 part-periods. It appears that it
would be more economical to key the second lot to period 5. The complete planned-order
schedule, adjusted for look-ahead, is shown in Figure 8-11.
The look-ahead test is repeated for successive pairs of period demands until it fails.
In our example, the second test (for periods 5 and 6) fails in that it would be more costly
to carry 40 for four periods than 30 for one period. If this were not so, the lot would be
keyed to period 6. To prevent the look-ahead feature from trying to overcome a steep
upward trend in the demand (this would create very large order quantities and defeat the
logic of LTC), an additional test is made. The part-period cost of the last period demand

FIGURE 8-11
Period 1 2 3 4 5 6 7 8 9 Total
Part-period
balancing with Net Requirements 20 40 30 10 40 30 35 20 40 265
look-ahead.
without look-ahead 90 80 95 265
Coverage
with look-ahead 100 105 60 265

3 J. J. DeMatteis, “An Economic Lot-Sizing Technique: The Part-Period Algorithms.” IBM Systems Journal. 7(1):30–38,
1968.