Information Technology Reference
In-Depth Information
Algorithm 3. Computation of investment costs
Computation of Investment Costs
compute_investments
(one DSN setup) returns InvestmentCost {
for each manufacturing node in setup do
determine total volume throughput;
determine number of manufacturing lines;
InvestmentCost = InvestmentCost + (no. of manuf. lines * investment per line);
end
return InvestmentCost;
}
nokia iMPle Ment ation and case study
Tool Implementation Method
The tool based on above formalism was developed in Java2 language, and embedded in the Oracle
environment hosting demand supply planning data repository. The tool's Web user interface lets the
logistics managers specify product structures, and costs for alternate suppliers. Common data such as
transportation costs between cities, duty and tax rates are centrally updated by few key users. The Web
UI has “analyze” button associated which invokes the Java algorithm. The Java algorithm first reads the
database tables for product and supplier information, constructs the corresponding DSNnet instance,
and runs the reachability analysis on it. Final results are stored in database tables and displayed in Web
browser. Permissions to view data are granted in such a way that a normal user may see only his work
when using the tool. Key users for each business group have view of all programs in that business unit.
Finally, the global logistics directors see all programs in every business unit. The next subsection con-
tains a real example of an analysis case for a product that started shipping in the summer of 2006.
Analysis of a new handset
The product structure for the new handset that started to ship in the summer of 2006 is shown in Figure
6.
As indicated in the diagram modules A, B and H had 2 supplier options, with the rest having a
single decided supplier. For a single customer this product structure produces 8 (2*2*2) demand supply
network setups. The analysis covered two customers, so the total number of demand supply networks
came to 64 (8^2). Interestingly, the static costs for the alternative suppliers of modules A, B and H (e.g.
production costs, investment costs) were not radically different. However, their global position was
such that the buffering and transportation costs associated with the total demand supply network were
significantly different from one solution to another. For this product, whose expected lifetime is one
year, the demonstrated cost differences from the cheapest to the most expensive demand supply network
were ca. 15 Million Euros. The analysis tool was able to compute these 64 solutions in approximately 10
Search WWH ::
Custom Search