KTH mathematician Jan Kronqvist is awarded the Ruth and Nils Erik Stenbäck Foundation Prize

Jan Kronqvist

Jan Kronqvist  was born in Finland and has Swedish as his mother tongue. He received his PhD with distinction from Åbo Akademi University in 2018, where he was also awarded the prize for the best doctoral thesis at the Faculty of Science and Engineering. During his doctoral studies, he spent time as a visiting researcher at Carnegie Mellon University in the United States.

In 2018, he was awarded a Newton International Fellowship from the Royal Society in the United Kingdom, which enabled two years of postdoctoral research at Imperial College London. He was subsequently appointed Associate Professor of Optimization and Systems Theory at the Department of Mathematics, Division of Numerical Analysis, Optimization and Systems Theory, KTH Royal Institute of Technology in Stockholm.

One of this year’s award recipients is Doctor of Technology Jan Kronqvist, who is being awarded the Ruth and Nils Erik Stenbäck Foundation Prize for his outstanding scientific contributions in mathematics and optimization. The prize includes €30,000 in personal funds awarded to Kronqvist.

“I am very happy and honoured to receive the award. It is an important recognition of the work behind the research. This is one of the largest prizes for early‑career researchers in mathematics, physics, and chemistry in Finland. In today’s competitive research environment, merits like this are crucial for advancing research,” says Jan Kronqvist.

Research in Mixed‑Integer Optimization

Kronqvist’s research focuses on mixed‑integer optimization, a mathematical field that combines discrete variables, such as yes/no decisions with nonlinear relationships. His work includes both the development of new mathematical theory and the design of efficient algorithms, with applications in areas such as artificial intelligence (AI) and machine learning.

“My research focuses on developing mathematical theory, methods, algorithms, and computational techniques to solve non‑convex optimization problems. I am particularly interested in optimization problems that involve discrete variables and nonlinear relationships, as they arise across a wide range of applications and have interesting mathematical properties,” says Jan Kronqvist.

Such problems arise in areas including energy systems, transportation, radiation therapy, planning and scheduling, as well as AI‑based decision‑support systems.

Global Optimality and Reliable Solutions

A central goal of Kronqvist’s research is to find globally optimal solutions while being able to verify that no better solution exists.

“Finding a global optimum and at the same time certifying that the solution is truly optimal is often computationally very demanding. However, for many applications such an optimality certificate is essential, as it means that the solution can be regarded as reliable,” he explains.

By identifying mathematical structures in complex optimization problems, Kronqvist and his colleagues have in several cases achieved dramatic improvements in computational efficiency.

“There have been problems that previously could not be solved within two weeks, but which through the right mathematical reformulations, can now be solved on a laptop in under ten seconds,” he adds.

Significant contributions to AI safety and software development

Kronqvist is internationally recognized as a leading researcher in convex and nonlinear mixed‑integer optimization. He has developed a large number of advanced algorithms and award‑winning software, including the open‑source MINLP solver SHOT, which received the COIN‑OR Cup at the INFORMS Annual Meeting 2018.

His research has been published in leading journals and conferences, including AAAI and NeurIPS, and has had significant impact on how optimization methods are used for verification, robustness, and safety in neural networks.

“In AI and machine learning, rigorous global optimization methods were long considered computationally too difficult. In recent years, however, it has become clear that they can play a central role in ensuring that AI systems are safe and do not make dangerous or unreliable decisions,” says Kronqvist.

Awards and Academic Leadership

For his scientific contributions, Jan Kronqvist has received several international distinctions, including the Howard Rosenbrock Prize for Best Paper in Optimization and Engineering, as well as the Best Paper Award at the CPAIOR Conference in Vienna. He currently leads an active research group consisting of postdoctoral researchers and PhD students at KTH.

In 2024, he was elected a member of the Young Academy of Sweden. He also serves as Chair of the Swedish Operations Research Society since 2024.

“In my research, I strive both to advance the fundamental theory and mathematical understanding of optimization problems and to design computationally efficient methods for solving difficult and societally important problems,” Kronqvist concludes.

Text: Jelina Khoo