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

4
H-Index
8
Papers
56
Total Citations
7
Avg Citations/Paper
🏆 Most Cited Paper
Deep reinforcement learning of event-triggered communication and consensus-based control for distributed cooperative transport
18 citations · 2022
📈 Most Prolific Year: 2023 (3 Papers)
🤝 Key Collaborators: 12
🏛 Institutions: Toyota Central Research and Development Laboratories (Japan), Nara Institute of Science and Technology, Toyota Motor Corporation (Switzerland)

Top Papers

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 13 days ago