Hanna Krasowski
Papers
4
Total Citations
73
H-Index
2
About
Hanna Krasowski is a leading researcher at the intersection of reinforcement learning (RL) and autonomous vehicle safety. Her work focuses on developing provably safe RL algorithms for motion planning, a critical challenge in deploying autonomous systems in the real world. Krasowski’s major contributions include the creation of **CommonRoad-RL**, a configurable and standardized environment that allows researchers to benchmark and compare different RL-based motion planners for autonomous vehicles. This work has become a foundational tool in the field, garnering over 35 citations. She is perhaps best known for her pioneering approach to safe RL, where she integrates **reachability analysis** and **polynomial zonotopes** to create a “safety shield” that projects unsafe actions onto safe ones, providing formal safety guarantees for nonlinear continuous systems. Her 2023 paper on this topic has already received over 34 citations, highlighting its immediate impact. More recently, she has extended these guarantees to satisfy complex temporal logic specifications in continuous action spaces, pushing the boundaries of what is possible in safety-critical RL. Krasowski’s work is essential reading for anyone interested in making RL practical for autonomous driving and other high-stakes applications.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4