Yuni Fuchioka
Papers
3
Total Citations
46
H-Index
2
About
Yuni Fuchioka is a robotics researcher whose work bridges reinforcement learning, imitation learning, and soft robotics to create more agile and adaptable robot behaviors. Their most impactful contribution is **OPT-Mimic** (2023, 39 citations), a framework that enables dynamic quadruped locomotion by imitating optimized reference trajectories rather than relying on motion capture data. This approach simplifies reward design while producing robust, natural-looking gaits—a significant advance for legged robot control. In parallel, Fuchioka’s 2024 work on **robotic object insertion with a soft wrist** (5 citations) tackles contact-rich manipulation under uncertainty. By combining a compliant wrist with sim-to-real privileged training, the system safely handles variations in object grasp and hole pose, demonstrating how soft hardware can ease the sim-to-real gap for precise assembly tasks. Together, these contributions showcase a talent for extracting simple, powerful priors from complex optimization—whether for bounding quadrupeds or delicate insertions. Fuchioka’s research is particularly notable for its practical focus: both projects address real-world deployment challenges, from unstructured environments to safe physical interaction. As a rising voice in robot learning, their work offers clear blueprints for students seeking to combine simulation, optimization, and hardware-aware design.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3