Haruhisa Motoyama
Papers
3
Total Citations
21
H-Index
3
About
Haruhisa Motoyama is a researcher in the field of robotics and reinforcement learning, with a particular focus on how robots can autonomously acquire motion forms through trial-and-error processes. His work centers on the application of Q-Learning—a model-free reinforcement learning algorithm—to enable real-world robots to develop locomotive behaviors without explicit programming. In his most cited paper (2006, 10 citations), Motoyama analyzed the motion forms generated by a caterpillar robot with two actuators as it learned to perform looping and advance actions, using reward databases to guide the learning process. A key contribution of his research is the investigation of dynamic reward structures: in a subsequent study (2006, 8 citations), he explored how changing rewards during training—mimicking human learning processes—can improve motion acquisition. His third notable work (2006, 3 citations) further examined how biological evaluation principles can inform the design of learning algorithms for actual robots. Though his citation counts are modest, Motoyama’s early experiments with physical robots and adaptive reward systems offer valuable insights into the intersection of reinforcement learning and bio-inspired robotics, making his work a foundational reference for researchers interested in embodied AI and autonomous skill acquisition.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3