Many companies are
producing to customer order to a larger extent than some years ago.
Lot or batch size reductions and competition have contributed to
this change. Good customer service has to include informing tne
customer, at the earliest point in time, if the full commitment may
not be met. Therefore, linkages to both production order status and
projected completion (from individual customer orders) is needed to
support exception reporting, and communication with the customer.
This typical "job shop" facility is now a widespread need.
production schedule planning functions of the classical body of
knowledge are not incorrect in terms of role or significance. They
still have a role to play, particularly as the interface with sales
and operations planning. But traditional master scheduling systems
are batch processed, only periodically (e.g., weekly), with frozen
zone restraints and forecast accuracy dependencies.
Under the three preceding headings, we have already discussed the
increase in customer order driven production, and associated
resource analysis aspects. But the overall need is for more dynamic
and interactive functions that can be more sensitive to the customer
order driven environment of Quick Response. Demand management must
be more focused toward order management. The MPSP functions must
follow this trend, and provide the functional assistance that we
have just discussed.
planning of requirements is still key for effective utilization of
resources while maintaining service level. Both capacity and
materials availability need to be analyzed in this way.
Requirements Planning (MRP) and Capacity Requirements Planning
(CRP) are the generally accepted tools, but suffer from severe
constraints in the make to order environment. In some cases,
particularly textile mills (weaving, knitting, and finishing) the
usual MRP/ CRP program calculations are unusable. There are
substantive reasons for the opinion that "MRP packages doesn't work
in our business."
planning system approach (i.e., MRPII), is based on the master
schedule. It calculates material schedules, allows decision inputs,
and then separately performs detailed capacity (requirements)
planning. But why do we have to separate these two functions? If you
are a manufacturer with many products/variants and few materials
(the inverted triangle) you are usually going to first decide on
when you have capacity to produce, then deal with materials. Within
the fiber/textile/apparel chain (and process oriented businesses)
the earlier stages of the production chain (i.e.,
fiber/yarn/fabric/finishing) are usually planned in this way (i.e.,
capacity first, material second).
Also, the lead time
offset used in (MRP) scheduling of materials and the lot sizing
techniques are inappropriate to manufacturers who have many
relatively slow machines making similar products, such as in
spinning and weaving. The planner decision process must look at
requested quantities and dates, and then consider such variables as
the number of machines to use, and loom processing and occupancy (or
current/planned warp setup) in calculations. Instead of the
classical materials planning multilevel processing and
recommendations, the practitioner really needs to perform level by
level simulations, by identifying opportunities to add to
existingproduction, change number of machines, and so on. Order
release becomes a different task. It has to be supported by the
ability to change these parameters, make backward or forward
scheduling calculations, change dates/quantities and review load
against capacity at each level of material or production. All this
has to be interactive, cumulative and order based, not batch
processing with results next morning.
In many situations
the lack of these facilities in planning has contributed to the
popularity of the so-called finite planning systems. If you could
interactively perform scheduling simulations such as we have
described, without reliance on batch MRP/CRP techniques, the need
for finite planning software would probably be reduced. A further
symptom of these MRP inadequacies are systems where
planning/execution integration is basic or non-existent. The current
popularity of separate manufacturing execution systems is also due,
in some part, to the cumbersome order release facilities of typical
To Be Continued