Tsukasa Kusakabe

Papers

Total Citations

H-Index

About

Tsukasa Kusakabe is a roboticist advancing the frontier of contact-rich manipulation, with a focus on precision assembly and compliance control. His research centers on designing and estimating stiffness matrices—particularly non-diagonal forms—to enable robots to handle complex, unstructured environments. Kusakabe’s most impactful work, “Reinforcement Learning for Robotic Assembly Using Non-Diagonal Stiffness Matrix” (2021, 45 citations), pioneers a learning-based approach to automate tasks requiring multiple contact transitions, such as high-precision assembly, by leveraging the high design freedom of stiffness control. He further contributes to online environmental adaptation through “Probabilistic Approach to Online Stiffness Estimation for Robotic Tasks” (2022) and deepens the theoretical foundation of compliance in “Design of non-diagonal stiffness matrix for assembly task” (2022). By integrating reinforcement learning with advanced mechanical design, Kusakabe addresses critical challenges in robotic dexterity and adaptability, offering practical solutions for industrial automation. His work is essential reading for researchers in robotic manipulation, control theory, and manufacturing, demonstrating how intelligent stiffness design can unlock new capabilities in contact-rich task execution.