Model Scenarios
A dynamic modeling tool is a computer-based model that helps the analyst assess the impact of various decisions regarding the configuration or coordination of the supply chain. Two general classes of dynamic modeling tools are "What-if" models and "Optimization" models. A "What-if" model allows the analyst to test the impact of various decisions on one or more important performance measures. The tool does not attempt to make a value judgment regarding the decision, but rather presents the information to help the human manager make a more informed decision. An "Optimization" model focuses on a single performance measure. The tool determines the optimal value of one or more parameters that are under the control of the decision maker, based on an objective of optimizing performance along the single performance measure. A key advantage of an optimization model is that it will help determine the optimal value of key decision variables; a key advantage of a what-if model is that it can be used for a more complex set of interactions and multiple performance measures. A wide variety of software applications can be used to create a dynamic modeling tool, ranging from Microsoft Excel to highly specialized software for performing non-linear optimization or conducting Monte Carlo or discrete event simulation.Strategic Issues in Modeling Scenarios
Prior to developing the dynamic model, it is important to be clear about what purpose the model is intended to serve. What are the key decisions that the model will try to help make? How broad is the scope of the model (depth of the supply chain, number of entities, length of time, etc)? Will it be used by a single organization or will it be used to help drive a collective decision across the supply chain? Will it be used only by the analyst creating the model or will it be a tool that is to be used by a wide variety of personnel? Clarity around these and other related questions will help ensure that the dynamic modeling tool is developed at an appropriate level of detail to be useful; too little detail and the results will not be taken seriously, too little detail and the insight will likely be lost in the minutiae of the model. No model will ever be able to fully reflect all the nuances of a complex supply chain. Deciding which issues are the most important, and therefore necessary to include in the model, should be driven by an understanding of what the model is attempting to accomplish, and therefore which components of the supply chain are most important to include in the model to help allow the model to achieve its objectives.
When developing a dynamic model, three important issues must be determined:
i) Inputs - What are the items that will take on different values as the dynamic modeling tool is being used? Typically these either represent decisions that the decision maker has control over (e.g., the location of a warehouse) and hence wants to determine their impact on performance so that a good choice for these variables can be made, or they represent items that are uncertain (e.g., demand for a product) and hence the decision maker wants to test the robustness of those variables that he/she can control to changes in these other important items.
ii) Outputs - What performance measures should be used to assess the performance of the supply chain under different scenarios. There might be a single output (e.g., cost) that is much more important than any other result, or there might be several outputs (e.g., cost, responsiveness, quality) that are all important. The model must be designed so that these key performance measures are assessed in whatever level of detail is deemed necessary.
iii) Translation of Inputs to Outputs - The guts of the dynamic modeling tool is the conversion of inputs to outputs. This is where the logic is built that will dictate how the multiple decisions and events that are included as inputs will interact to impact supply chain performance. Any "decision making" that takes place within the model (e.g., orders are only released when sufficient funding is available) must be coded into the model in order for it to provide a reasonably accurate reflection of reality.
When developing a dynamic modeling tool, there must be steady communication between the model developer and the client. Initially, the communication will be around a requirements document that describes the purpose of the model, the model inputs, and the output that the model will provide. Once agreement is reached on these design principles, a prototype tool should be developed and then shared. Looking at the working prototype, modifications to the inputs and outputs (as well as discussion regarding the internal workings of the tool) can be agreed upon between the model developer and the client. Additional prototype iterations can be developed until the tool is able to help guide decision making within the supply chain.