Research

I have had the opportunity to lead a number of interesting research projects during my Ph.D. My research combines network science, data analysis, optimization, multi-criteria decision-making, econometrics, and geographic information science methods to investigate the resilience of agri-food flow networks across spatial scales and then proposes practical solutions ensuring societal well-being and environmental stability.

Agri-food supply chains are complex systems that incorporate production, distribution, processing, and consumption of perishable goods. Their safeguarding is essential to feed humanity. However, food supply chains have become subject to an array of unforeseen threats, such as climate change, pandemics, cyber intrusions, geopolitical conflicts, and economic crises. As “affordable and available food for everyone, at all times” ensures a stable economy and civil rest, my research recognizes food supply chains as non-traditional defense objects. I concentrate on their distribution step (i.e., food flow networks) which serves as the bridge between supply and demand. I lead descriptive and predictive analytical studies to ensure resilient agri-food flow networks; thus, food security. Here is a summary of some of my research efforts.

Mapping U.S. agri-food flows on multi-modal transportation infrastructure

Currently, I leverage my experience in data analytics, network science, and geographic information science to map U.S. food flows on multi-modal transportation infrastructure.

In this study, through developing a unified framework, I will evaluate the load of agri-food supply chains on transportation infrastructure by flow type. Further, I will assess the efficiency vs. adaptability against disruptions balance among inland highways, railways, and waterways within the U.S. using econometrics and complex network science. My findings will illuminate the real-world implications of cost vs. benefit in agri-food flow network resilience. This study is soon to be submitted.

Optimization of national grain imports to balance risk vs. return

Recently, I explored the interplay of unit cost, cost volatility, and supply diversity in national grain imports with modern portfolio theory.

Many nations rely on imported grain to meet their dietary requirements. Countries aim to balance import risk with the expected return of their grain supplies. I brought these dual objectives of grain trade together by employing modern portfolio theory. Nations that rely on more risky suppliers are observed to be also paying higher unit cost of import. To introduce portfolios that adeptly balance risk vs. return, I formulated a mean-variance optimization model with mass balance. I proposed both optimal and optimum portfolios to determine opportunities for importer nations to achieve more stable and diverse grain supplies, while also acknowledging economic considerations.

Mass imports of empirical, optimal, and optimum grain portfolios for case study nations.

Structural chokepoints of U.S. agri-food flow networks

I fused multi-criteria decision-making and network science techniques for a vital component-importance analysis of food flows within the U.S. to address President Biden’s Executive Order #1407.

The agricultural and food systems of the United States are critical for ensuring the stability of both domestic and global food systems. Thus, it is essential to understand their structural resilience to a suite of threats. I employed connectivity, resilience, and efficiency metrics from complex network science to identify the spatially resolved structural chokepoints in the agri-food supply chains of the United States. Urban centers emerged as key hubs for processed food transit, but diverse food production geographies led to distinct chokepoints. These chokepoints are also found to be generally consistent through time, particularly for processed food commodities.

Structural chokepoints of all agri-food aggregated network corresponds to transit hubs.

Temporal food flows between U.S. counties

I estimated food movement across U.S. counties in 2007, 2012, and 2017 by a machine learning model, and uncovered spatial trends through time-series analysis.

I built on the existing literature in fine grained spatial fluxes of food, but with the additional consideration of time. I developed an improved version of the Food Flow Model, a data-driven model, to estimate food flows (kg) between all county pairs across all food commodity groups for the years 2007, 2012, and 2017 which requires estimating 206.3 million links. Then, I determined the core counties to the U.S. food flow networks through time by leveraging a multi-criteria decision analysis technique. To share the key findings of this study with the general public, an openly available data visualization portal, foodflows.org, is built together with the College of Education.

Food in and out flows for over 3000 counties are estimated per commodity through time.

Resilience and efficiency trade-off in global food trade

I captured the logistics of agri-food trade with a network science perspective and assessed the balance between efficiency and resilience using statistical connectivity metrics.

Trade is typically optimized to promote efficiency, whereas resilience is increasingly being recognized as another important objective. However, it is not clear if prioritizing resilience comes at the expense of efficiency or if the two objectives can be promoted simultaneously. I developed a complex network framework, built upon the repeat movement of food commodities across nations, to quantify the efficiency and resilence of food trade for the last century. As unprecedented in investigating the interplay between efficiency and resilience for global food trade, I revealed that trade of food groups is more at equilibrium than individual commodities.