Jana Pavlasek
Papers
6
Total Citations
22
H-Index
3
About
Jana Pavlasek is a robotics researcher whose work bridges scalable education, articulated object perception, and multi-robot coordination. Her most recognized contribution is the **MBot platform**, a low-cost modular robot ecosystem that has trained over 1,400 students in autonomous navigation since 2014 at the University of Michigan and partner colleges—a landmark achievement in robotics education. In perception, Pavlasek has advanced the understanding of **articulated objects** (e.g., tools, cabinets) in cluttered environments. Her work on **NARF22** introduced neural articulated radiance fields for configuration-aware rendering, while her parts-based belief propagation method tackles the high-dimensional pose estimation problem under occlusion. She also pioneered **affordance coordinate frames** for manipulation-oriented object perception, enabling robots to generalize actions like pouring to novel containers. For multi-robot systems, her **Stein Variational Belief Propagation** framework addresses decentralized coordination in high-dimensional, uncertain spaces. With over 20 citations across her top papers and a strong focus on real-world deployment, Pavlasek’s research is shaping how robots perceive, manipulate, and learn—both in the classroom and in the wild.
Research Focus
Key Achievements
Top Papers
- 1MBot: A Modular Ecosystem for Scalable Robotics Education6 citations · 2024
- 2Sketching Affordances for Human-in-the-loop Robotic Manipulation Tasks5 citations · 2019
- 3
- 4
- 5Stein Variational Belief Propagation for Multi-Robot Coordination2 citations · 2024
- 6