How to use Scenarios
Network Design allows you to make changes to the input data, run the optimization again, and save the result as Scenarios which can then be compared to other scenarios to examine the effects of different kinds of changes.
We recommend the following workflow:
Save your Base Case Scenario.
Make desired changes to the data in the Analytics pages.
Run a new optimization by clicking Optimize (AIMMS icon).
Save your results as a new Scenario.
Repeat this process of making changes and saving new Scenarios as many times as you like. You can then make side-by-side comparisons.
Example
To illustrate how this feature may be used, consider the following simple supply network, with 5 suppliers serving customers across North America with a variety of products. For this example, we focus on Product M1, whose flows look like this:
To illustrate the use of the Scenario feature, suppose we learn that our Chicago supplier will be unable to provide Product M1 next month. During this outage at Chicago, we would like to reassign customers currently served by Chicago to the “next best” suppliers. We can determine this by creating a new scenario, with Chicago no longer able to supply M1 and then rerunning the optimizer. The results will tell us the best way to reassign the Product M1 demands that were previously served by Chicago.
Create the Base Case Scenario:
Click Optimize (AIMMS icon) to solve the model for normal operations (with
ChicagosupplyingM1).Go to Settings > Create Scenario.
Enter the name and description for our new Scenario in the appropriate fields. (E.g.,
BaseCase.)Click Save.
Now we have created a Scenario called BaseCase and added it to our library of Scenarios saved in the database.
Next, we will disable Chicago as a supplier for Product M1.
Any changes made on any report screens will affect the input for the next optimization. These changes and their results may be saved as Scenarios. They may not be saved as new datasets. Only the Data Navigator can create datasets.
All the input change can be reset to match the data in a dataset by selecting a dataset to load (or reload). This has the effect of clearing any current solution results and resetting all inputs to match the contents of the dataset you just loaded.
Therefore, to disable Chicago as a Supplier for M1 we go to Supplier Report page and clear the check box that marks Chicago as an active Supplier for Product M1. We then optimize the network again, to see the optimal M1 supply reassignments when Chicago’s M1 supply is not available:
After optimizing, the Results Overview page now looks like the image below.
We can see graphically in the upper map that Chicago is no longer providing M1.
To see the BaseCase results in the second map, we select the BaseCase Scenario in the drop-down list provided between the upper and lower maps shown below.
The second map displays the results from the BaseCase scenario. The upper map and table always display the results from the most recent model solution, known as the Active Scenario. Now we can easily see how the M1 customers formerly served by Chicago have been reassigned. We can also see that our purchasing costs and transportation costs have increased.
For more detailed tabular comparisons of our new solution with the BaseCase Scenario, we will use the Case Comparison feature of AIMMS. To do this we must first save this new solution as a new Scenario. We go to the Save Scenario page action which will open Save Scenario dialog, and save this solution as a Scenario named NoChicagoM1
Now we have two Scenarios in our library of saved Scenarios, as shown on the Settings > Scenario screen:
We will now use the AIMMS Case Comparison capability. For illustration, we navigate to the Transportation Report screen, because it has a nice detailed table of all product flows that will make a good example. Initially, this table appears as shown below.
It shows the results for the current Active Scenario only.
To cause this table to display each of our saved Scenarios, we click the Data Manager icon on the top menu bar, as highlighted above. This opens the Case Selection menu on the left margin. There we move the mouse pointer over our BaseCase scenario, and then select the “Load as Active” icon, highlighted below.
We repeat this process for our NoChicagoM1 Scenario, this time selecting the Compare Case option.
This causes the table to update with new subheadings for our two Scenarios, as circled in the screenshot below.
By comparing the entries for the two Scenarios we can see exactly how the sourcing of M1 has been reassigned from Chicago for each individual Customer.
In fact, all tables on every Screen in the Analytics section now have new subheadings for each of our Scenarios. For example, here is the Cost table found on the Analytics > Costs page.
