Zhilan Feng
$249,999
Brigham Young University
Utah
Mathematical and Physical Sciences (MPS)
This award is funded in whole or in part under the American Rescue Plan Act of 2021 (Public Law 117-2). The need for more robust and resilient global supply chains is highlighted by numerous recent events such as COVID, the Suez Canal obstruction, the Texas power crisis, and the Colonial Pipeline attack. Unfortunately, the lack of global, firm-level supply chain data has led to a proliferation of models relying on data that is either simulated, incomplete, or overly aggregated. Without good models of supply chain behavior, it is difficult to predict the outcome of disruptions and plan appropriate interventions to minimize future impacts. The principal investigator has acquired complete data on the significant supply chain dependencies of all publicly traded firms worldwide, which opens the door to validated and predictive models. The principal investigator will use this data to advance supply chain modeling in three ways. The first is by creating a more realistic model of supply chain disruption, including deriving more accurate robustness estimates. The second is by modeling cascading failures and dependencies between different layers of the global value chain, which goes beyond supply chain to include relationships such as financing, intellectual property sharing, and strategic alliances. And finally, by assessing the effect of the many unobserved private firms on supply chain robustness. These three advances will greatly augment the ability to predict and protect against future supply chain disruptions. This project will enable the establishment of an interdisciplinary lab at the principal investigator's institution that incorporates participation of under-represented minorities and undergraduates, including building relationships with the Lavassani lab at North Carolina Central University, an HBCU. In preliminary work, the principal investigator and coauthors assessed the robustness of a large supply chain using novel (and more realistic) metrics, but much more work is needed to make these assessments realistic. The three goals of the project are: (1) Incorporate firm metadata (such as industry and revenue) to get tighter bounds on robustness compared to existing graph-theoretical techniques. This includes creating a generative model of the supply chain using the data and stochastic block modeling tools. (2) Derive rules for cascading failures and dependencies between different layers of the global value chain, thus extending existing cascading failure ideas to this important type of data. (3) Assess boundary/missing data effects on the model outputs. This will take inspiration from the boundary conditions techniques in dynamical systems and partial differential equations, as well as generative techniques in statistical physics. The development of these techniques has great significance in the broader field of network science, where boundary effects are common but not frequently analyzed due to lack of accepted frameworks.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.