Thomas Beckers
About me
I am an Assistant Professor at the Department of Computer Science, Vanderbilt University. My research interests include physics-enhanced learning, nonparametric models, and safe learning-based control
Research
In modern control applications such as autonomous driving or human-robot-interaction, the control design and modeling process becomes very time-consuming or even infeasible due to the complexity of the systems. A solution to overcome these issues is provided by data-driven models which require only a minimum of expert knowledge. However, a general problem of data-driven models is the unpredictable outcome which limits their application to non-safety critical systems so far.
My research focuses on safe learning-based modeling and control with Bayesian data-driven models. In particular, I focus on Gaussian Processes (GP) as they provide not only a mean prediction, but also a variance as uncertainty measure of the model. I have developed new GP-based control laws which guarantee safety and high-performance, introduced algorithms to include physical prior knowledge for improved generalization and trustworthiness, and exploited the uncertainty measure for active online learning.
Short Bio
Thomas Beckers is an Assistant Professor of Computer Science and the Institute for Software Integrated Systems at Vanderbilt University. Before joining Vanderbilt, he was a postdoctoral researcher at the Department of Electrical and Systems Engineering, University of Pennsylvania, where he was member of the GRASP Lab, PRECISE Center and ASSET Center. In 2020, he earned his doctorate in Electrical Engineering at the Technical University of Munich (TUM), Germany. He received the B.Sc. and M.Sc. degree in Electrical Engineering in 2010 and 2013, respectively, from the Technical University of Braunschweig, Germany. In 2018, he was a visiting researcher at the University of California, Berkeley. He is a DAAD AInet fellow and was awarded with the Rhode & Schwarz Outstanding Dissertation price. His research interests include physics-enhanced learning, nonparametric models, and safe learning-based control.
News
03/2023 Our paper "Physics-enhanced Gaussian Process Variational Autoencoder" has been accepted at L4DC 2023.
03/2023 Two accepted paper at IFAC WC. In addition, we will organize a workshop on Gaussian process based identification and Control
03/2023 I gave a seminar on Gaussian Process Port-Hamiltonian systems at University of Wuppertal
01/2023 I offer a project at the Vanderbilt School of Engineering Summer Research Program for undergraduates.
01/2023 I've started my new position as Assistant Professor of Computer Science at Vanderbilt University.
12/2022 I presented our three papers on Gaussian process based learning and control at the IEEE Conference on Decision and Control
10/2022 It was my pleasure to give a keynote at the IFAC Symposium on Robot Control, Oct 17-20, 2022
09/2022 We will organize a workshop on Gaussian Process Learning-based Control at the IEEE Conference on Decision and Control (CDC) 2022
07/2022 All of our three papers have been accepted at CDC 2022. See you there!
06/2022 Our article "Safe Trajectory Tracking for Underactuated Vehicles with Partially Unknown Dynamics” has been accepted for publication in the Journal of Geometric Mechanics.
06/2022 I am happy to share that I will be joining the Department of Computer Science at Vanderbilt University as Assistant Professor in January 2023.
05/2022 I gave a seminar at Technical University of Darmstadt, German Aerospace Center (DLR), and Leibniz University Hannover
04/2022 I gave a seminar at Vanderbilt University
02/2022 I gave a seminar at Columbia University
02/2022 I have been selected as a DAAD AInet fellow for the Postdoc-NeT-AI 3/2022 – AI and Robotics