Page 212 - Orlicky's Material Requirements Planning

CHAPTER 10 A New Way of Looking at Things 191

to and output from the shop. It is obvious that when input exceeds output, work-in-
process (and lead time, with the exception noted earlier) goes up, and vice versa.
It follows, therefore, that in order to reduce work-in-process and lead times, output
must exceed input temporarily. In order to control work-in-process so as to keep it from
exceeding a given level, input must be held to the existing rate of output. In an MRP envi-
ronment, input to the shop is represented by the work (and cost) contents of the shop
orders being released (actually, being recommended for release) by the system. The flow
of these orders, however, must not be held to the rate of current production.
The flow of orders being generated by an MRP system actually cannot be controlled
or regulated directly by anyone. It will not be possible to release an order prematurely if
the system has planned the component materials to become available only at the time of
need. To hold back on orders that the system is trying to release may create an increasing
backlog of orders past due for release. All such orders later will have to be released with
less than planned lead time, which is likely to cause difficulties in the shop and missed
order due dates.
With MRP, unimpeded order release is required. The system is releasing orders with
correct priorities, in correct sequence, and at what is, by definition, the right time.
Priority, that is, need, rather than the work-in-process level or any other consideration
should govern the releasing of orders. When the priority strata of work in process are
taken into consideration (work with deferred or nonexistent requirements is, at any given
time, “immovable” regardless of current input and output), the level of work-in-process
assumes secondary importance. What are important are priorities, order completions,
shipments, and customer delivery service.
When there is a significant change in requirements as a result of revising the MPS,
the MRP system may pump a large number of new orders into the shop on top of an
already high work-in-process. As long as the MPS is valid in terms of both marketing
needs and capacity, let work-in-process be what it will, and let the MRP system do its job
without interference at the order-release point.
When input to the factory exceeds output, it is an indication of the fact that the MPS
is overstated in terms of capacity and should be changed unless capacity can be increased
promptly. An updated view of the relationship between input and output is presented in
Figure 10-9. The desired balance between the rates of input and output should exist not
at the point of the MRP system’s output but at the point of its input. What should be reg-
ulated in relation to the factory’s output is the amount of future work implicit in the MPS.
This measure of capacity has relevance only at the bottleneck for the plant. Having the
other resources temporarily overloaded or underloaded—provided that it does not affect
the schedule at the bottleneck—is not cause for concern.

Living with Bad Forecasts
Forecasting of independent demand, a classic problem of inventory management,
appears in a decidedly different light in view of the new capability to update for change.