Skip to main content
Styliana Avraamidou sits at table with laptop

With NSF CAREER award, Avraamidou is building a tool to help the Wisconsin dairy industry run smooth as butter

Mentioned:

The modern food distribution system is a wonder of logistics and engineering, with ingredients moving seamlessly from farms and fields to processors, manufacturers, distribution hubs, and, eventually, grocery store shelves. That system is even more impressive when it comes to perishable items like dairy products (or even certain pharmaceuticals, or blood) that need to stay refrigerated for their entire journey, degrade over time and have a defined lifespan.

But what happens if a major railway or interstate hub is knocked out of commission? What if one key ingredient is in short supply? What if a factory goes offline unexpectedly? Any of these scenarios can wreak havoc on the perishable goods supply chain, resulting in spoiled food, wasted medication and money down the drain.

Answering those questions is how Styliana Avraamidou, the Duane H. and Dorothy M. Bluemke Assistant Professor in chemical and biological engineering at the University of Wisconsin-Madison, will use her National Science Foundation-funded CAREER award. She’s applying her expertise in process systems engineering to develop tools that allow farmers, food processors and distributors to better understand and work around unexpected disruptions in the perishable supply chain.

Avraamidou and her students will develop an integrated computational framework for modeling, optimizing and analyzing resilience in perishable food chains. First, they will develop a graph-based model to explore the cascading effects these disruptions can have throughout the supply chain and perform game-theoretic analysis to explore the impacts of collaboration and competition among stakeholders.

“We’ll explore disruptions like increasing energy and fuel prices, limited worker power, transportation disruptions, natural disasters and changes in laws and regulations,” she says. “We will then develop a unified modeling framework to see how different disruption events propagate, how processing units or industry actors are affected and how the decisions of industry actors affect each other.”

The team will then develop optimization tools that can help industries make long-term preparations for unexpected disruptions and allow for real-time operational changes to mitigate these problems.

Avraamidou will use the Wisconsin dairy industry as her primary case study and plans to collaborate with the UW-Madison Division of Extension Dairy Program to help ground the framework in real-world scenarios. “We hope to capture what’s currently happening in Wisconsin and then see if we can redesign the supply chains of perishable products to make them more resilient to different types of disruptions,” she says.

Because decisions in the real world are often more complex and less logical than scenarios represented in a computer model, Avraamidou and her students are also developing a board game based on Wisconsin’s dairy supply chains. They’ll invite students of different ages, along with industry insiders like farmers and distributors, to play, observe how these experts make decisions and glean knowledge and nonintuitive data about how the industry works in real life. The researchers can represent those insights via game theory and add them to their framework, providing a deeper understanding of how power dynamics and information asymmetry shape system-wide outcomes. The players, through the game’s repeated decision-making process, will be able to learn about economic trade-offs, short and long-term impacts of their decisions, concepts about finite resources and the cascading effect of disruptions.

At the end of the project, Avraamidou says she plans to roll out a dairy resilience hub that will help industry stakeholders collaborate and plan so that the milk stays fresh and the ice cream remains frozen. “We hope that we will have user-friendly tools that policymakers and industry can use to figure out what changes are needed in the current supply chain,” she says. “And then they can use them to predict future disruptions and how we can optimally address them.”