offers a three-dimensional pictorial of integrated shipment size, transportation
mode, and location.
The planning curve joins the point of total least cost for each shipment size. It
does not join locational points. For example, the number of locations to support least
cost for one size of shipment may be more or less than for another. Further analysis is
required to identify the specific locations that offer the least cost alternative for each
shipment size and transport combination. Assume that the locations under consideration
consist of a network ranging from 1 to 12 warehouses. Within this range, the planning
curve will identify a smaller number of acceptable locations for detailed evaluation.
In Figure 15-7, the points of least cost, shipment size, and transportation
combinations fall within a range of four to eight locations.
Analysis is required to select the final warehouse network. Initially, the time duration
of the performance cycle and inventory availability assumptions should be held
constant. The service availability and performance cycle duration serve as parameters
to help isolate an initial least cost approximation. At a later point in strategy formulation,
these parameters can be relaxed and subjected to sensitivity analysis. The fit of
the least cost planning curve requires marginal cost analysis for each shipment
sizeltransportation mode combination for networks of four, five, six, seven, and eight
warehouse locations. Provided customer service objectives are achieved within the
four-to-eight warehouse range, a first-cut least total cost network of potential warehouses
is identified.
A final refinement involves the evaluation of specific warehouse sites or facilities.
In the case of Figure 15-7, which is also the situation in most complex modeling
approaches, the best-fit location network is limited to an array of warehouse locations
selected for analysis. The results may be managerially satisfactory but not cost superior
to a different group of locations. Each warehouse assortment selected for analysis
will have a least cost combination. The final policy may require that the analysis be
completed with several different network combinations to identify those most suitable
for a given business situation. The final warehouse selection using such a trial
and error methodology will never identify the mathematical optimal solution to minimize
total logistical cost. It will, however, help management identify a superior network
to the existing operation that may better service customers at lower total logistics
cost.
To evaluate the wide range of variables in designing a logistical system, complex
models have been developed. The assumptions required to support integrated system
design are important from the viewpoint of their impact upon strategy formulation.
The integrated total cost curve must take into consideration all relevant variables that
are critical to logistical system design.
Formulating Logistical