Thomas Bergener
Papers
5
Total Citations
73
H-Index
4
About
Thomas Bergener is a pioneering researcher in cognitive robotics and dynamical systems, whose work has fundamentally shaped how anthropomorphic robots perceive, decide, and interact with humans. His landmark 1999 paper, "Complex behavior by means of dynamical systems for an anthropomorphic robot" (42 citations), introduced a groundbreaking generative approach to behavioral organization—using continuous dynamical systems to seamlessly switch between multiple robot actions, replacing brittle, rule-based architectures with fluid, mathematically grounded control. This framework, detailed further in his 1998 foundational work (15 citations), codes logical action contexts into matrices of differential equations, enabling robots to exhibit emergent, lifelike behavior. Bergener extended this paradigm to human-robot interaction in his 2002 study (8 citations), demonstrating on his multi-degree-of-freedom robot "Arnold" how dynamical systems can robustly interpret fuzzy, rapid human gestures for intuitive man-machine interfaces. He also advanced robot perception through evolutionary optimization, showing in two papers (2000, 2001; 3 and 5 citations respectively) that derandomized evolution strategies can automatically tune visual obstacle detection parameters (e.g., Inverse Perspective Mapping) to outperform expert-tuned settings. Bergener’s integrated vision—where continuous dynamics govern both action selection and interaction—remains a cornerstone for designing autonomous robots that move, decide, and collaborate with natural, human-like fluidity.
Research Focus
Key Achievements
Top Papers
- 1Complex behavior by means of dynamical systems for an anthropomorphic robot42 citations · 1999
- 2
- 3
- 4
- 5Evolutionary Parameter Optimization for Visual Obstacle Detection3 citations · 2000