Papers
12
Total Citations
127
H-Index
8
About
Christian Pek is a robotics and autonomous systems researcher whose work sits at the intersection of safe motion planning, human-robot interaction, and reinforcement learning. His research addresses one of the most pressing challenges in modern robotics: ensuring that autonomous systems behave safely and predictably in real-world environments shared with humans. Pek has made significant contributions to perceived safety in human-drone interaction, demonstrating that physical safety alone is insufficient — robots must also *feel* safe to nearby humans. His work on shield synthesis and human-feedback integration into deep reinforcement learning (RL) advances the frontier of policy safety without overly constraining robot behavior. Notably, he has pioneered approaches that leverage non-expert human feedback to repair and align RL policies, reducing the burden of reward engineering while keeping systems robust to real-world variability. His research extends into data-driven model predictive control for complex dynamics, belief-space planning under uncertainty, and spatio-temporal logic for task specification — reflecting a remarkably broad technical range. With papers accumulating citations across safety, control theory, and human-in-the-loop learning, Pek's body of work is shaping how the next generation of autonomous robots will reason about risk, uncertainty, and human expectations in dynamic environments.
Research Focus
Key Achievements
Top Papers
- 1Increasing Perceived Safety in Motion Planning for Human-Drone Interaction22 citations · 2023
- 2
- 3Safe Data-Driven Model Predictive Control of Systems With Complex Dynamics14 citations · 2023
- 4
- 5
- 6
- 7Belief Control Barrier Functions for Risk-Aware Control11 citations · 2023
- 8
- 9Risk-aware Spatio-temporal Logic Planning in Gaussian Belief Spaces4 citations · 2023
- 10