A case for "supermodeling"

No one can predict the future. But by using "supermodeling," or a computer replica of a real or planned supply chain system, companies can get close, write Knud Erik Wichmann and Tim Lawrence in "Painting a bigger picture" from the Quarter 4/2009 issue of CSCMP's Supply Chain Quarterly.

Supermodeling helps companies plan ahead by defining potential scenarios, risks, and options, and then assessing the likely outcomes of each. Here's how the process worked for one example company.

The company wanted to move from a decentralized distribution network in Europe to a more centralized one. With 16 distribution centers (DCs) in the "as is" situation, this was not a simple optimization exercise. Complicating matters was the fact that all of the country directors were against closing "their" warehouses. It was important to help them objectively evaluate and compare alternatives. To accomplish this, the company's European logistics organization followed four steps.

Step 1. Develop a baseline. The company created a baseline scenario by using a computer model to simulate one year of operation as the network was currently structured. The baseline gave the company confidence in the model because it generated total annual logistics costs, inventory levels, and delivery performance that were approximately the same as in real life.

Step 2. Develop future scenarios.The team changed inputs to the model, producing several simulated scenarios and showing the associated effects on costs and service. This allowed the team to understand the problems in the network and draw up a short list of realistic potential solutions.

Step 3. Evaluate options. The model proposed closing down various groups of DCs and simulated the outcomes, testing for all kinds of "what ifs." The solution that proved best —cutting back from 16 DCs to three — would allow the company to cover all of its markets in Europe while enjoying a 20-percent saving in logistics costs.

4. Evaluate results. Members of the company's wider European logistics organization jointly selected the best solution. The modeling approach helped to establish consensus among the stakeholders, avoiding the dangerous route of making decisions based on political and emotional resistance to change.

Subscribers to CSCMP's Supply Chain Quarterly can read the full text of the article here. To subscribe to the Quarterly and gain full access to all current and archived articles, click here.