** Next:** Summary
** Up:** A Decision Support System
** Previous:** Modeling System

Our work is primarily motivated to address the problems faced during decision making. Different problems require different modeling approach. Some problems may require detailed simulation models to evaluate
the various alternatives available. For example, the problem of deciding the
optimal manufacturing strategy amongst make-to-stock, make-to-order, and
assemble-to-order. It is not possible to build tractable analytical models for
these problems with realistic assumptions. On the other
hand, there are various problems such as selection of new suppliers and scheduling
which can possibly be modeled accurately analytically. There are certain cases where the iterative use of a modeling approach may provide the best solution. For example, consider the problem of controlling inventory
at various locations while maintaining service levels for customers. This
problem can be solved analytically only at an aggregate level of detail. Therefore after getting an initial solution through an analytical
approach the system can be simulated under various scenarios to evaluate the
system against the required performance measures. Thus simulation along with analytical
modeling can often be deployed in the decision making process.
As already seen in an earlier section,
there are a wide gamut of tools and techniques deployed in supply chain
problem solving at the strategic, tactical, and operational levels.
In particular, techniques such as linear programming, integer programming,
mixed integer programming, heuristic optimization, and simulation are popularly
employed. A large number of customized algorithms have also been developed
over the years. The decision workbench includes the commonly deployed
approaches such as linear programming, mixed
integer programming, and simulation, and can also easily support customized
solutions.

Different tools in the workbench need the supply chain details to be presented
in different ways and at different levels of abstraction. Object models can
support this in a natural way. Such models once built can be used to provide customized viewpoints
required for specific supply chain issues.

Thus DESSCOM could be used by the users who are responsible for implementing
supply chain decisions. The decision workbench can be used to compute various
optimal decisions or to evaluate various alternatives available. In the next
section we describe the prototype of DESSCOM that we have implemented.

The use of DESSCOM in modeling and solving the various decision problems
has been demonstrated on a wide variety of problems. For details,
see [#!biswasthesis!#].

** Next:** Summary
** Up:** A Decision Support System
** Previous:** Modeling System
*Shantanu Biswas*

*2000-08-16*