Papers
9
Total Citations
33
H-Index
3
About
Masashi Sugimoto is a robotics researcher focused on enabling autonomous decision-making in dynamic, unpredictable environments. His core research areas include robot control, reinforcement learning, state-action prediction, and multi-agent systems. Sugimoto’s major contribution lies in developing methods for real-time sequential decision-making, where robots predict future state-action pairs to adapt their behavior on the fly—a critical capability for machines operating in complex, real-world settings. His work on using online Support Vector Regression (SVR) for state-action prediction and flexible-weight coefficients has laid groundwork for more adaptive robot controllers. More recently, he has explored deep reinforcement learning architectures, including Dueling DQN and centralized critic approaches for multi-agent coordination, as well as tracking abilities of Deep Q-Networks using LiDAR-equipped mobile robots. While his citation counts (ranging from 1 to 8 per paper) reflect a niche but dedicated audience, his sustained research output from 2013 to 2024 demonstrates a consistent commitment to advancing autonomous robotics. Sugimoto’s work is particularly relevant for students and researchers interested in bridging classical control theory with modern deep RL for real-time robot adaptation.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5
- 6FUTURE MOTION DECISIONS USING STATE-ACTION PAIR PREDICTIONS3 citations · 2015
- 7
- 8
- 9