Supply Chain Risk Response Analysis Using Utility Functions

Researchers: Zhilin (Charlie) Liu, Hanyi (Hannie) Zheng, Jingyi (Jean) Wu

Climate change is an increasingly important risk factor for society. Its impacts can have a significant impact on manufacturing and present challenges to supply chain decision makers. It is critical that appropriate responses are taken based on the probability of risk occurrence and the impact of the risk. Climate is changing in ways that can no longer be ignored by supply chain managers. For example, wildfires are occurring more frequently; they can disrupt transit routes or affect the health of factory workers on which supply chain operations rely. The interruption of the supply chain will result in the late delivery of products to customers including the supply chain partners.

Acceptance, mitigation, avoidance, and transferring are the four risk management approaches for dealing with the impact of any risk, based on several factors: probability, impacts, and controls already in place. There are a variety of risk actions available for managers to select for each risk based on unique scenarios or situations. For wildfire risk, mitigating risks may involve taking action to prevent the start of wildfire or limiting the range or speed of the wildfire. The company may work with the government to build better fire alarms or fire detection systems in the forest to identify the wildfires quickly.

The approach taking by the researchers including the construction of a decision tree to document the costs and benefits associated with each alternative. The decision-making process will be investigated using a decision aid in the form of a color-coded matrix displaying the uncertainty associated with each decision in order to calculate the utility in decision makers assigns to each climate change impact. An algorithm, coded in Python or R Shiny, will create a web application to facilitate dynamic responsiveness by inputting values and then interacting with the application. Finally, an Analytic Hierarchy Process (AHP) is developed to determine the weights of each criterion through pairwise comparisons, assign numerical values to each variable using the Eigenvector method, and determine the appropriate action.