Papers
29
Total Citations
727
H-Index
14
About
Kyohei Otsu is a robotics researcher specializing in autonomous navigation, multi-robot systems, and planetary exploration, with a particular focus on enabling robots to operate reliably in extreme and unstructured environments. His work spans off-road terrain traversability, subterranean exploration, and next-generation space robotics, making him a prominent figure at the intersection of field robotics and artificial intelligence. Otsu's most influential contributions include STEP, a probabilistic framework for risk-aware navigation in challenging off-road terrain (110 citations), and the NeBula autonomy architecture developed for TEAM CoSTAR's back-to-back success in the DARPA Subterranean Challenge (105 and 51 citations), one of the most demanding robotics competitions ever conducted. His work on PLGRIM advances hierarchical decision-making under uncertainty for large-scale exploration, while CHORD and ACHORD address the critical challenge of communication-aware coordination in multi-robot teams operating with intermittent connectivity. Beyond terrestrial applications, Otsu has pushed the frontier of space robotics through EELS, a snake-like robot designed to explore icy moon environments like Saturn's Enceladus (62 citations), and through MAARS, which brings machine learning capabilities to Mars rover systems. His research consistently bridges theoretical rigor with real-world deployment, establishing him as a leading voice in autonomous exploration robotics.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10MAARS: Machine learning-based Analytics for Automated Rover Systems29 citations · 2020