Contributions in supply chain risk assessment and mitigation

This dissertation develops contributions in the area of supply chain risk assessment and mitigation. In each of the three main chapters, we present and analyze a risk assessment or mitigation problem for supply chains. The first problem is to assess the impact of infrastructure disruptions on supply chain performance; the second problem is to develop an operational control approach to mitigate risks posed by uncertain events that disrupt network synchronization; and the third problem is to analyze the risk posed by an adversary seeking to use a supply chain as a weapon. Chapter II presents a methodology for assessing the excess supply chain costs that arise from a failure of or an attack on a critical supply chain infrastructure component. Different from many subjective risk assessment practices, our methodology provides a systematic approach to search for the most vulnerable supply chain components and measure the economic consequences of disruption. Modeling a supply chain using network flow models, we analyze the impact of disruption by linear programming theory, and propose an efficient assessment algorithm based on the dual network simplex method. Finally, a case study on the U.S. corn export supply chain is presented. Chapter III discusses the mitigation of risks created by transit time uncertainties in less-than-truckload (LTL) line-haul operations. Transit time uncertainty may undermine the performance of the load plan, which specifies the route for each shipment and is synchronized to reduce line-haul costs. In our study, risk assessment of a load plan is performed via a dispatch simulation under randomly generated travel time scenarios. The risk consequence is measured by the average excess operational cost, including transportation cost and handling cost. Compared to existing line-haul network models embedded within integer programming approaches for load plan optimization, the dispatch simulation can evaluate the performance of a load plan more realistically. In addition, a heuristic search algorithm based on "multi-tree pivots" is provided to obtain a cost-efficient load plan that is robust to transit time uncertainties. Chapter IV presents methodology to assess the consequence of risks which arise from the intentional contamination of a food supply chain. Different from many risk management practices, the source of risk in this problem is an intelligent adversary, e.g., a terrorist group, who intends to deliver chemical or biological toxins to consumers using the supply chain. First, a general modeling scheme based on state-space models is provided to describe the dissemination of toxin across consumed products in a food supply chain. Then, a case study based on a representative liquid egg supply chain is presented. Based on the system model, a risk assessment for different supply chain designs is performed by simulation. Moreover, an in-depth analysis is conducted to determine the worst-case consequence given an intelligent attack considering the operational characteristics of the system. The worst-case consequence tool developed is designed to be embedded within any risk assessment approach.PhDCommittee Chair: Erera, Alan; Committee Member: Davis, Lauren; Committee Member: Shi, Jianjun (Jan); Committee Member: Sokol, Joel; Committee Member: White III, Chelse