Jakub Bednarek
Papers
7
Total Citations
141
H-Index
6
About
Jakub Bednarek is a robotics researcher whose work sits at the intersection of haptic sensing, legged robot locomotion, and terrain perception. His research has made significant contributions to how robots understand and interact with their physical environments through touch — an area often overshadowed by vision-based approaches yet critical for real-world deployment. Bednarek's most influential work focuses on haptic terrain classification, demonstrating that mobile and legged robots can reliably identify surface types using force, torque, and proprioceptive data alone. His 2019 paper on classifying terrain via haptic sensing (40 citations) established a strong foundation in this niche, while subsequent work employing Transformer architectures (2021) pushed the field toward faster, more practical solutions. His research on proprioceptive Monte Carlo localization (22 citations) showed that robots can navigate reliably even when cameras and LiDAR fail — a breakthrough for extreme environments. Beyond locomotion, Bednarek has explored object stiffness estimation using soft grippers and neural networks (21 citations), bridging manipulation and tactile intelligence. His contributions to foothold selection for quadrupedal robots further round out a cohesive research identity centered on robust, touch-aware autonomy. With over 140 cumulative citations, his work is shaping the next generation of physically intelligent robots.
Research Focus
Key Achievements
Top Papers
- 1What am I touching? Learning to classify terrain via haptic sensing40 citations · 2019
- 2Robotic Touch: Classification of Materials for Manipulation and Walking24 citations · 2019
- 3
- 4
- 5Fast Haptic Terrain Classification for Legged Robots Using Transformer16 citations · 2021
- 6
- 7CNN-based Foothold Selection for Mechanically Adaptive Soft Foot3 citations · 2020