While Carla Michini and her family took their usual summer trip to her native Italy in 2025, this was no relaxing getaway. Michini, an assistant professor of industrial and systems engineering at the University of Wisconsin-Madison, spent much of the time in her seaside hometown of Roseto degli Abruzzi working on her proposal for the National Science Foundation’s prestigious Faculty Early Career Development Program—better known as the NSF CAREER Award.
But in between exhaustively outlining the aims of her project, developing the research plan and tracking down references to cite, she spotted a pair of gold-and-white-enamel hoop earrings at her sister Lucia’s jewelry shop. She decided she’d buy them to celebrate if she got the CAREER Award.
Instead, Michini’s sisters surprised her with the earrings before the end of her trip. And, a few months later, Michini got the good news: She had, in fact, earned the CAREER Award, which provides nearly $560,000 over five years in support of her research in combinatorial optimization.
“They brought me luck,” she says.
In many ways, Michini’s work is about removing luck’s influence, by developing mathematical models and methods to solve sticky problems in complex systems. She exploits the structures of combinatorial optimization problems—problems in which you have a finite but huge set of options to choose from, such as the “traveling salesman problem” or the “knapsack problem”—to derive efficient algorithms for computing optimal or nearly optimal solutions.
For her CAREER Award project, Michini will focus on developing algorithms and models to enhance decision-making in decentralized systems. In systems such as supply chains, transportation or logistics, many individual decisions can have broader, system-wide effects. But often those decisions are made independently, without coordination.
“In a decentralized system, the final outcome might not be aligned with what would be a system optimum,” says Michini. “Due to the lack of coordination, there is inefficiency, which we would like to decrease. But to do that, first of all, we need models that capture these interdependencies and are able to predict what could be an equilibrium of the system. And then we want to analyze the gap between the system optimum and what we can instead achieve with an equilibrium in a decentralized system. My goal is to rethink, adapt and apply combinatorial optimization tools in the framework of
game theory, pushing their impact from centralized to decentralized decision-making.”
Specifically, Michini plans to test her toolset on post-disaster aid operations, which involve scattered supply chains and complex transportation networks.
“Humanitarian organizations have a common goal to help people that are in need,” she says. “They act in a decentralized way, so there might be issues coming from congestion of scarce resources, and also the lack of coordination means, for example, cost-sharing mechanisms are not exploited. We can enhance coordination hopefully through better models.”
Michini will also use the CAREER Award funding to develop a new graduate-level course on game theory, which intersects with her work in combinatorial optimization. She also plans to organize visits from researchers in the field to meet with students to discuss both their work and their career paths.
“I want to break that barrier that sometimes exists between professors and students and try to have the students think, ‘You know, I can also maybe achieve that,’” she says. “Hopefully we can engage them.”