Finished and ongoing projects
Projects at DSME
| Title | Type | Year | Name | Supervisor |
|---|---|---|---|---|
| Achieving Compliant Motion Control for Continuous-Contact Task with Reinforcement Learning | Master thesis | 2023 | Aditya Pradhan | Emma Cramer |
| Data Transfer in Robust Direct Data-Driven Control | Master thesis | 2023 | Dmitrii Likhachev | Alexander von Rohr |
| High-dimensional Convex Bayesian Optimization for Controller Tuning | Bachelor thesis | 2023 | Felix Kappel | Paul Brunzema |
| Neural Processes for Event-Triggered Learning and Control | Master thesis | 2023 | Paul Kruse | Paul Brunzema |
| Unsupervised State Representation Learning from Video for Reinforcement Learning Applications | Master thesis | 2023 | Jonas Reiher | Emma Cramer |
| Safe Bayesian optimization of quadrocopters in time-varying environments | Master thesis | 2023 | Antonia Holzapfel | Paul Brunzema |
| Safe Bayesian optimization for learning automotive tracking control | Master thesis | 2022 | Johanna Menn | Sebastian Trimpe |
| Utilization of model uncertainty for the algorithmic optimization of model-based deep reinforcement learning approaches | Bachelor thesis | 2022 | Artur Eisele | Bernd Frauenknecht |
| Detection and Isolation of Anomalies in Process Data Through Observer | Bachelor thesis | 2022 | Fabian Feucht | Paul Brunzema |
| Balancing a Two Degrees of Freedom Inverted Pendulum with a Robotic Arm | Bachelor thesis | 2023 | Kyrylo Sovailo | Alexander von Rohr & Shiming He |
| Cooperative Behaviors in Sequential Social Dilemmas using Multi-agent Reinforcement Learning | Master thesis | 2022 | Julia Freytag | Dominik Baumann |
| Kernel-based machine learning methods and modern robust model predictive control | Master thesis | 2022 | Victoria Hankemeier | Christian Fiedler |
| Learning the stage cost in model predictive control | Master thesis | 2022 | Kenneth Goveas | Christian Fiedler |
| Learning and Forgetting in Time-Varying Bayesian Optimization | Master thesis | 2021 | Paul Brunzema | Alexander von Rohr |
| Learning Policies on a Rotatory Pendulum Using Gradient Descent Based Bayesian Optimization | Bachelor thesis | 2021 | Reis Baltaoğlu | Alexander von Rohr |
| Globally Optimal Safe Robot Learning in Higher Dimensions | Master thesis | 2021 | Bhavya Sukhija | Dominik Baumann, together with Matteo Turchetta and David Lindner (both ETH Zurich) |
| Toward Safety Guarantees for Deep Neural Network Controllers: Stability Analysis of Reinforcement Learning Policies | Master thesis | 2021 | Hannah Markgraf | Emma Pabich and Dominik Baumann |
| Learning the stage cost function for model predictive control without terminal conditions | Internship | 2021 | Kenneth Goveas | Christian Fiedler |
Projects with external partners
| Title | Type | Year | Name | External institution | DSME Supervisor |
|---|---|---|---|---|---|
| Analysis of Observation Spaces for Reinforcement Learning in Robotic Manipulation | Master thesis | 2023 | Kartik Sachdev | Siemens AG | Emma Cramer |
| Data-Driven Process Optimization for Injection Molding | Master thesis | 2023 | Christian Fiedler | Institute for Plastics Processing, RWTH Aachen | Alexander von Rohr |
| Benchmarking Learning-Based Control and Reinforcement Learning | Master thesis | 2023 | Tsung Yuan Tseng | TU Munich | Alexander von Rohr |
| Learning-based approximate model predictive control | Master thesis | 2022 | Abdullah Tokmak | ETH Zurich | Christian Fiedler |
| Curriculum Adversarial Reinforcement Learning | Master thesis | 2022 | Maximilian Tölle | Intelligent Autonomous Systems Group, TU Darmstadt | Bernd Frauenknecht |
| Physics-informed neural network regression of a robot arm's inverse dynamics | Master thesis | 2022 | Josefine Monnet | Werkzeugmaschinenlabor WZL der RWTH Aachen | Andreas René Geist |
| Multi-Task Active Learning for Computer Vision | Master thesis | 2022 | Nik Dorndorf | École Polytechnique Fédérale de Lausanne | Alexander von Rohr |
| Deep Learning for Nonlinear Observer Design | Master thesis | 2022 | Lukas Bahr | MINES ParisTech | Pierre-François Massiani |
| Data Science in press shop - Using machine learning algorithms to predict downtime | Master thesis | 2022 | Tom Jericho | Volkswagen AG | Dominik Baumann |
| Towards Industrial Reinforcement Learning: Adapting Modular GPS to Industrial Applications | Master thesis | 2022 | Robin Kupper | AZO GmbH Osterburken | Friedrich Solowjow |
| An approach to instance segmentation and pose estimation for automated dismantling of lithium-ion batteries for high-quality recycling | Master thesis | 2022 | Anna-Maria Meer | Institute for Business Cybernetics and Knowledge-Based Systems Group, RWTH Aachen University | Alexander von Rohr |
| Learning Dynamics Equations of a Novel Rotational Pendulum | Bachelor thesis | 2021 | Yannick Streicher | Universität Tübingen, Max Planck Institute for Intelligent Systems | Dominik Baumann |
| Sim-to-Real Transfer of Deep Reinforcement Learning Agents in Gearshift Optimization of Automatic Transmissions | Master thesis | 2021 | Bernd Frauenknecht | Mercedes-Benz AG | Sebastian Trimpe |
| Event Localization with Deep-Learning-based Time Series Classification for Process Monitoring in Robotics | Bachelor thesis | 2021 | Jonas Reiher | KUKA Deutschland GmbH | Alexander von Rohr |
| Data-enabled Predictive Control of Robotic Systems | Master thesis | 2021 | Felix Wegner | ETH Zürich | Sebastian Trimpe |
| Reinforcement Learning for Wendelstein 7-X Divertor heat load control | Master thesis | 2021 | Timo Thun | Max Planck Institute for Plasma Physics | Sebastian Trimpe |
| Data-driven State Estimator for Quadrupedal Robot ANYmal | Master thesis | 2021 | Qingxu Zhu | ETH Zürich | Sebastian Trimpe |