Papers
8
Total Citations
56
H-Index
4
About
Kazuki Shibata is a rising researcher at the forefront of multi-robot coordination and human-robot interaction, with a focus on reinforcement learning for cooperative transport and navigation. His most influential work, "Deep reinforcement learning of event-triggered communication and consensus-based control for distributed cooperative transport" (2022, 18 citations), introduces a novel framework that enables robots to communicate only when necessary, significantly reducing bandwidth while maintaining task efficiency. This is complemented by his 2023 study on multi-robot task allocation for cooperative transport (15 citations), which addresses the challenge of distributing robots among multiple objects of unknown weight—a critical step toward real-world deployment. Shibata has also pioneered methods for integrating human guidance into robot navigation, including geometric and pointing instructions, and has developed a "Language to Map" approach that generates topological maps from natural language path descriptions (2024). His work on cooperative grasping using ternary force representation (2025) further extends multi-agent reinforcement learning to force-based coordination without explicit communication. With a growing citation record and contributions spanning distributed coverage control and resilient monitoring systems, Shibata is establishing himself as a key innovator in scalable, human-aware multi-robot systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4Enhanced Robot Navigation with Human Geometric Instruction4 citations · 2023
- 5
- 6Enhancement of Robot Navigation Systems via Human Pointing Instruction2 citations · 2023
- 7
- 8Coverage Control for Resilient Monitoring System2 citations · 2019