Johannes Ultsch
Papers
3
Total Citations
17
H-Index
3
About
Johannes Ultsch is a researcher at the DLR Institute of System Dynamics and Control, where his work sits at the intersection of autonomous driving, reinforcement learning, and robust motion control. His primary research focus is on developing learning-based controllers for safety-critical automotive applications, particularly path following control (PFC) for over-actuated robotic vehicles. Ultsch’s major contribution lies in addressing the critical challenge of parametric uncertainties in vehicle models—showing how dynamics randomization during reinforcement learning training can produce controllers that generalize robustly to real-world conditions, rather than failing when model parameters shift. His most-cited works, including a 2023 paper on this technique and a 2019 study on learning-based PFC, each have 7 citations, reflecting growing interest in bridging the simulation-to-reality gap. Ultsch also co-authored a 2020 overview of DLR’s automotive control research, highlighting his role in a group dedicated to safety, comfort, and sustainability. For students and researchers, his work offers a compelling example of how modern machine learning can be rigorously applied to autonomous driving, where reliability is non-negotiable.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3