Getting Started With Intelligent Greenfield Analysis
Greenfield Analysis Overview
Greenfield analysis (GF) is a method for determining the optimal location of facilities in a supply chain network.
The name Triad comes from the oldest known species of frogs – Triadobatrachus. You can think of it as the starting point for the evolution of all frogs, and it serves as a great starting point for modeling projects too! We also can use Triad to identify 3 key parameters:
- The optimal number of distributions centers
- The location of each distribution center
- Which customers should be served by which distribution center
GF is a great starting point for network design—it solves quickly and can reduce the number of candidate site locations in complicated design problems. However, a standard GF requires some assumptions to solve (e.g. single time period, single product). As a result, the output of a Triad model is best suited as initial information for a building more robust Cosmic Frog optimization (Neo) or simulation (Throg) model.
Running Your First Greenfield Analysis
You can run GF in any Cosmic Frog model. Running a GF (Triad) model only requires two data tables:
- Customers
- CustomerDemand
A greenfield analysis starts with clicking the “Run” button, just like a Neo or Throg model.
In the “Run” parameters menu, select “Triad (Intelligent Greenfield)” as your engine.
We call our GF approach “Intelligent Greenfield” because of the different parameter options available in our engine. These options can be configured in the Greenfield Settings table.
Greenfield Scenarios
Like a Neo or Throg model, Triad scenarios allow you to simultaneously run multiple Triad models with different parameters.
Triad scenarios make edits to the “GreenfieldSettings” table by changing the value of an engine parameter.
Customer Clustering
To improve the solve speed of a Triad model, we can use customer clustering. Customer clustering reduces the size of the supply chain by grouping customers within a given geometric range into a single customer. We can set the clustering radius (in miles) in the Greenfield Settings table under the Customer Cluster Radius column.
Clustering is optional, and leaving this column blank is the same as turning off clustering.
While grouping customers can significantly improve the run time of the model, clustering may result in a loss of optimality. However, Greenfield is typically used as a starting point for a future Neo optimization model, so small losses in optimality at this phase are typically manageable.