Pneumatic Soft Rotary Actuator for Humanoid
The pneumatic rotary actuator for soft humanoid robots is developed based on a bidirectional soft actuator made of hyperelastic materials, aiming to realize smooth and safe forearm-like rotational motion in a human-like manner. This actuator operates solely through internal air pressure regulation and enables stable and flexible rotation without the need for rigid mechanical components. By incorporating both the compressibility of air and the nonlinear mechanical behavior of soft materials, the actuator is modeled as a nonlinear dynamic system, and advanced control strategies are applied for precise actuation. A key focus is on the real-time control stiffness, allowing the actuator to adapt to varying external conditions and mimic the contraction-relaxation characteristics of biological muscles. This research paves the way toward integrating such actuators into a comprehensive soft robotic forearm system, contributing to enhanced safety and adaptability in human-robot interaction.
Multi-curvature and Variable Stiffness Soft Gripper
This research aims to develop a soft gripper robotic system capable of multi-curvature and variable stiffness. Conventional grippers struggle to handle objects with low surface stiffness, such as baked goods and agricultural produce. To overcome this limitation, we have developed a new soft gripper. To address the low stiffness inherent in soft grippers, we implemented a variable stiffness mechanism, enabling the gripper to handle objects weighing up to 1.8 kg. Additionally, incorporating a multi-curvature mechanism allows for realizing seven distinct curvatures. This innovation is expected to find broad applications in sectors where conventional grippers have been inadequate, including bakeries, agriculture, and the service industry.
Configurable Soft Robot
We research a design method that can drive the deformation of a soft robot to a precise target without any control. In previous research, designing soft actuators to attain a predetermined actuation stroke has been challenging due to the complexities of deformation analysis. This issue often made it hard to achieve user-specified kinematics in soft robotics. However, ensuring the actuator converges to the target upon inflation can be achieved by the surface geometry-based method that determines the initial shape of the actuator based on its desired deformed shape. A notable advantage of our methodology is its ability to rapidly compute soft robot design without advanced complex continuum mechanics theories, making it adaptable to various scenarios.
Variable Stiffness Mechanism using particle jamming phenomenon
This research aims to optimize variable stiffness mechanisms through theoretical modeling. Despite the widespread use of particle jamming in variable stiffness mechanisms, few theoretical models exist, primarily due to the system's complexity. Existing models also present limitations when applied to the field of robotics. In this study, we derive an energy-based theoretical model for particle jamming and detect force chains, which are critical for stiffness augmentation. In conclusion, the optimized particle jamming mechanism will be employed across various fields.
