Alex Tsolovikos

Hi 👋

I am an Aerospace Engineering PhD candidate at the University of Texas at Austin with experience in control and estimation theory, computational engineering, fluid flow physics, robotics, and machine learning. My PhD research is in ML for model reduction and control of high-dimensional, nonlinear systems, with an emphasis on turbulent flows. I received my undergraduate degree (BS/MS with highest honors) in Mechanical Engineering from the National Technical University of Athens.

Some of my interests include: model predictive control • reinforcement learning • perception for autonomous driving • simultaneous localization and mapping • vision and radar object detection • transformer-based architectures for motion prediction • HD maps • computational geometry • high-performance computing • and computational physics.

Research

My research is focused on estimation, control, and model reduction for high-dimensional systems, such as turbulent flows. Here are some of the topics that I am working on:

• Reduced-order modeling
• Model predictive control for turbulent boundary layers
• Data-driven stochastic optimal control
• Transformers for trajectory prediction
• Gaussian process regression for dynamics
• High-performance computing

Industry Experience

As a three-time intern with Aptiv, I have worked extensively on:

• Deep learning for object detection from low-level radar data
• Radar-based localization and mapping (SLAM)
• Sensor fusion and nonlinear state estimation
• HD Maps

Software Skills

• Programming: Fluent in C++, Python, Matlab, Fortran
• Libraries: PyTorch, JAX, ROS, GTSAM, Eigen, CVXPY, CGAL
• Tools/Platforms: Unix, MPI, OpenMP, Git, Vim, SLURM

GitHub Highlights

Here are some of my recent projects on Github:

• Goal Transformer: Dynamics-Aware Motion Prediction using Multidimensional Transformers and Gaussian Processes
• Cautious Nonlinear Covariance Steering with Gaussian Processes
• Reinforcement Learning Control of Fluid Volumes
• Model Predictive Control of Fluid Volumes

Publications and Talks

• Separation Delay in Turbulent Boundary Layers via Model Predictive Control of Large-Scale Motions, APS DFD 2022
• Control of Large-Scale Motions in Turbulent Boundary Layers, Talk at the California Institute of Technology, August 2nd, 2022
• Multiple Model Dynamic Mode Decomposition for Flowfield and Model Parameter Estimation, AIAA SciTech 2022
• Control of Large-Scale Motions in Boundary Layers, APS DFD 2021
• Cautious Nonlinear Covariance Steering using Variational Gaussian Process Predictive Models, MECC 2021
• Distributed Covariance Steering with Consensus ADMM for Stochastic Multi-Agent Systems, RSS 2021
• Model Predictive Control of Material Volumes with Application to Vortical Structures, AIAA Journal 2021
• Estimation and Control of Fluid Flows Using Sparsity-Promoting Dynamic Mode Decomposition, IEEE L-CSS
• Greedy Finite-Horizon Covariance Steering for Discrete-Time Stochastic Nonlinear Systems Based on the Unscented Transform, ACC 2020
• Toward Model-Based Control of Near-Wall Turbulent Coherent Structures, AIAA SciTech 2020