I’m currently a PhD student at the Chair of Automatic Control Engieering at the Technical University of Munich (TUM).
My research focuses on safe and optimal control of uncertain, time-varying dynamical systems, such as robots with wearing joints, degrading batteries, or aircraft whose dynamics change with flight conditions. Controlling such systems involves a trade-off between exploration and exploitation, as actions should perform well under current knowledge while also shaping what is learned about the system. I am particularly interested in dual control, which explicitly accounts for how actions reduce model uncertainty, and its connection to Bayesian reinforcement learning. Since optimal dual policies are generally intractable, I develop efficient methods that capture their essential exploration behavior and study how active learning affects closed-loop stability and safety. To this end, I combine learning-based and adaptive control with probabilistic modeling, approximate dynamic programming, and optimization-based control.
I’ve obtained my master’s degree in Robotics, Cognition, Intelligence from TUM. During my masters, I had the opportunity to be an exchange student at the Johou Systems Kougaku (JSK) Laboratory at the University of Tokyo under the supervision of Prof. Kei Okada.
Prior to that, I’ve been a corporate student at Robert Bosch GmbH, where I worked in different departments in Abstatt, Curitiba (Brazil), and Renningen.
Social Media: GitHub, LinkedIn, Google Scholar