Supply Chain Frontiers issue #33. Read all articles in this issue.
The sheer complexity of an international supply network makes it vulnerable to a wide range of disruption risks. But this weakness can be turned into a competitive advantage if managers know how to marshal the supply chain’s considerable resources against the impact of a service breakdown. A simulation model being developed here at the MIT Center for Transportation & Logistics (MIT CTL) - and based on a real supply chain - might offer a way to teach managers how to combat large-scale disruptions by simulating the outcomes of different risk management responses and strategies.
The model was originally built in collaboration with a large consumer packaged goods company that wanted to quantify the effects of supply chain disruption risks. The supply chain includes a primary manufacturing plant and an offshore plant that are served by suppliers in multiple countries. These two facilities send components to a packaging plant where the parts are assembled. The packaging plant sends finished product to two customer-facing distribution centers (DCs) and to a third DC that also carries out labeling operations.
“Our goal was to create a model that shows how hypothetical disruptions ripple through the supply chain and impact customer service,” says Dr. Amanda Schmitt, the MIT CTL Postdoctoral Associate who built the model. “By being able to input different operations strategies and mitigation plans, managers can see the effects in real time and learn how to allocate resources to minimize the disruption and quickly restore network flow.”
In effect, supply chain and business continuity managers from the sponsor company were able to test drive a disrupted supply chain using the model. The types of disruption mitigation strategies that can be introduced are legion, but here are some examples:
- Inventory reserves — storing extra inventory in the network to protect against upstream disruptions.
- Where are the most effective storage location(s)?
- How much inventory is sufficient?
- Back-up replacements — using other plants or distribution centers in the network as a back-up if one is disrupted.
- How fast does the response need to be?
- How much back-up capacity would be needed?
- Supplier redundancy — investing in dual sourcing options so that a disruption at one supplier doesn’t cripple the network.
- Which suppliers are the most important to back up?
- Determining the appropriate balance between these techniques.
The model brought to light some important risk management issues, such as the fact that cross-network mitigation strategies were not well documented at the sponsor company. While internal business continuity actions were thoroughly performed at every node in the supply chain, little or no documentation was available on how one node of the network could back up another. Yet, management’s perception was that multiple back-up options existed in the network and could be quickly implemented.
The simulations also demonstrated that fill rate and recovery after a disruption are very sensitive to the operational state of the network when a disruption occurs. “Operational state” refers to the inventory levels and production capacities throughout the system. If inventory downstream from a disruption is low before the incident, the impact will be greater and the recovery time longer than if the inventory levels were relatively high. Since inventory levels may not always be at their intended targets (due to demand surges, production batches, etc.), this is a realistic concern.
The model also has yielded some important secondary benefits. As part of the initial development process, for example, business continuity and operations personnel talked about possible back-up and mitigation plans; these discussions in turn spawned new and creative ways to reduce risk in the network.
Work is now under way to develop the simulator model as a powerful teaching tool for business continuity and supply chain personnel from any company. As Schmitt pointed out, although other supply chains will respond differently to the disruptions simulated by the model, users can still learn a great deal about how various mitigation measures work. They also can learn how to deploy resources in a crisis to meet their operational and customer service goals and, most important, to sustain competitive advantage.
For more information on the simulation model for network risk management please contact Dr. Amanda Schmitt, aschmitt@mit.edu .